From sequoiamobil at gmx.net Mon Jun 1 06:19:33 2015 From: sequoiamobil at gmx.net (Sebastian Gabler) Date: Mon, 01 Jun 2015 08:19:33 +0200 Subject: [OmniOS-discuss] Failsafe boot In-Reply-To: References: Message-ID: <556BF975.1090408@gmx.net> Hi, on Friday I ran into a "Error 16: Inconsistent filesystem structure". To recover from this, I needed a failsafe environment. I ended up in using Firefly 0215, which worked fine for rebuilding all BEs like described here http://honglinsu.blogspot.co.at/2010/08/fix-solaris-10-zfs-error-16.html, but there are several concerns beyond taking a bit too much time to sort out this problem. 1. I understood from some previous request on omniti.com that there is no failsafe environment for Omnios. 2. The situation made it clear to me that a failsafe environment is a required tool to maintain an Illumos installation. With the above finding assumed correct, we depend on a separate distro project to keep pace - outdated Firefly images may be useless in the best case. 3. Are there alternatives, i.e. to make a failsafe environment for the specific machine, in case of doubt including additional drivers, on current ZFS versions, etc, as a sysadmin? BR, Sebastian From jimklimov at cos.ru Mon Jun 1 09:54:09 2015 From: jimklimov at cos.ru (Jim Klimov) Date: Mon, 01 Jun 2015 11:54:09 +0200 Subject: [OmniOS-discuss] Failsafe boot In-Reply-To: <556BF975.1090408@gmx.net> References: <556BF975.1090408@gmx.net> Message-ID: <88AFFDC3-1AE2-44FB-BC46-4E2523FE1489@cos.ru> 1 ???? 2015??. 8:19:33 CEST, Sebastian Gabler ?????: >Hi, > >on Friday I ran into a "Error 16: Inconsistent filesystem structure". >To >recover from this, I needed a failsafe environment. I ended up in using > >Firefly 0215, which worked fine for rebuilding all BEs like described >here >http://honglinsu.blogspot.co.at/2010/08/fix-solaris-10-zfs-error-16.html, >but >there are several concerns beyond taking a bit too much time to sort >out >this problem. > >1. I understood from some previous request on omniti.com that there is >no failsafe environment for Omnios. >2. The situation made it clear to me that a failsafe environment is a >required tool to maintain an Illumos installation. With the above >finding assumed correct, we depend on a separate distro project to keep > >pace - outdated Firefly images may be useless in the best case. >3. Are there alternatives, i.e. to make a failsafe environment for the >specific machine, in case of doubt including additional drivers, on >current ZFS versions, etc, as a sysadmin? > >BR, > >Sebastian >_______________________________________________ >OmniOS-discuss mailing list >OmniOS-discuss at lists.omniti.com >http://lists.omniti.com/mailman/listinfo/omnios-discuss The best I came up with, was copying the failsafe image, mounting as an FS, putting in new versions of kernel and driver files from the running system, and packaging that back. Details in my comments on Alex's blog where he announced Firefly. Yes, it is a rather perverse way, and a more straightforward, maintained (e.g. what if other drivers become needed?) and automated one would be welcome ;) Jim -- Typos courtesy of K-9 Mail on my Samsung Android From danmcd at omniti.com Mon Jun 1 14:39:23 2015 From: danmcd at omniti.com (Dan McDonald) Date: Mon, 1 Jun 2015 10:39:23 -0400 Subject: [OmniOS-discuss] zpool list -p In-Reply-To: <1E73E845-FD72-496D-9E7C-650580B7E305@licenser.net> References: <1E73E845-FD72-496D-9E7C-650580B7E305@licenser.net> Message-ID: <547B88E8-4C42-4FF7-875E-6A8E2BA4CE2E@omniti.com> > On May 30, 2015, at 11:58 AM, Heinz Nikolaus Gies wrote: > > If there is a reason behind that decision it?s fine and I?ll gladly work around it, but it feels like an oversight. At first glance, I don't see where in the code this gets implemented. I can't, therefore, tell you quickly whether or not it's something easy to upstream. Dan From danmcd at omniti.com Mon Jun 1 14:45:11 2015 From: danmcd at omniti.com (Dan McDonald) Date: Mon, 1 Jun 2015 10:45:11 -0400 Subject: [OmniOS-discuss] I think I broke part of the network stack In-Reply-To: <098DC8FA-29CD-4DA0-A87F-58DF4629A092@licenser.net> References: <098DC8FA-29CD-4DA0-A87F-58DF4629A092@licenser.net> Message-ID: > On May 30, 2015, at 2:55 PM, Heinz Nikolaus Gies wrote: > > 408 dladm create-vnic -l bge0 net0 -p zone=2398fe7c-032f-11e5-abb0-b33f9f953915 That one may be a SmartOS-ism. From the OmniOS dladm(1M) man page: zone Specifies the zone to which the link belongs. This property can be modified only temporarily through dladm, and thus the -t option must be specified. To modify the zone assignment such that it persists across reboots, please use zonecfg(1M). Possible values consist of any exclusive-IP zone currently running on the system. By default, the zone binding is as per zonecfg(1M). At the point you had dladm stop working, did you utter "svcs -xv". I suspect maybe the "network-datalink-management" service may have been reported as faulty at that point. Dan From danmcd at omniti.com Tue Jun 2 03:02:07 2015 From: danmcd at omniti.com (Dan McDonald) Date: Mon, 1 Jun 2015 23:02:07 -0400 Subject: [OmniOS-discuss] WARNING --> bloody repo is offline Message-ID: <8A59163B-16FD-4743-8271-53DDFF68B2EC@omniti.com> An accidental build-right-to-publisher occurred earlier this evening. This build is HIGHLY experimental, and not even ready for the bloody servers yet. I've taken the bloody repo server offline until I can clean things up. What may result is a full-world update push to the bloody repo server. Sorry, Dan From danmcd at omniti.com Tue Jun 2 03:08:57 2015 From: danmcd at omniti.com (Dan McDonald) Date: Mon, 1 Jun 2015 23:08:57 -0400 Subject: [OmniOS-discuss] Back online (was Re: WARNING --> bloody repo is offline) In-Reply-To: <8A59163B-16FD-4743-8271-53DDFF68B2EC@omniti.com> References: <8A59163B-16FD-4743-8271-53DDFF68B2EC@omniti.com> Message-ID: <1AC68B79-4D14-4601-AC34-FB52C58A5AFF@omniti.com> > On Jun 1, 2015, at 11:02 PM, Dan McDonald wrote: > > An accidental build-right-to-publisher occurred earlier this evening. This build is HIGHLY experimental, and not even ready for the bloody servers yet. > > I've taken the bloody repo server offline until I can clean things up. What may result is a full-world update push to the bloody repo server. Well, that was a quicker cleanup than I thought. A fortunate side-effect, I suppose, of gcc51 being hashed early in "buildctl build all"'s list. Yes, you read that right. I'm attempting to bringup gcc51. It won't be for the illumos-omnios bits, but it WILL be for all of the userland bits. I'm a bit concerned about breaking omniti-ms packages, among other things, which is why it's HIGHLY experimental still. :) Thanks! Dan From paladinemishakal at gmail.com Tue Jun 2 09:41:51 2015 From: paladinemishakal at gmail.com (Lawrence Giam) Date: Tue, 2 Jun 2015 17:41:51 +0800 Subject: [OmniOS-discuss] Problem with Omni R151014 upgrade Message-ID: Hi All, I am trying to test upgrade of Omni R151014 from R150008 but I am getting this error message: root at sgbk02:~# pkg update --be-name=omnios-r151014 entire at 11,5.11-0.151014 Packages to remove: 3 Packages to install: 9 Packages to update: 394 Mediators to change: 1 Create boot environment: Yes Create backup boot environment: No DOWNLOAD PKGS FILES XFER (MB) text/gnu-gettext 62/406 1546/18458 15.2/336.5 Errors were encountered while attempting to retrieve package or file data for the requested operation. Details follow: 1: Framework error: code: 18 reason: transfer closed with 101471 bytes remaining to read URL: ' http://pkg.omniti.com/omnios/r151014/omnios/file/1/135914f1bfb857941cb5b16641fb523bb22b0326'. (happened 4 times) 2: Framework error: code: 18 reason: transfer closed with 14577 bytes remaining to read URL: ' http://pkg.omniti.com/omnios/r151014/omnios/file/1/1f8d529160acc1839550884b4cec080fea55fb8b '. 3: Framework error: code: 18 reason: transfer closed with 275217 bytes remaining to read URL: ' http://pkg.omniti.com/omnios/r151014/omnios/file/1/1f8d529160acc1839550884b4cec080fea55fb8b'. (happened 3 times) root at sgbk02:~# traceroute pkg.omniti.com traceroute: Warning: Multiple interfaces found; using 10.25.1.201 @ igb1 traceroute to pkg.omniti.com (199.15.226.221), 30 hops max, 40 byte packets 1 202.55.95.113 (202.55.95.113) 0.863 ms 0.449 ms 0.423 ms 2 202.55.93.229 (202.55.93.229) 1.304 ms 1.246 ms 1.152 ms 3 singapore1.vqbn.com (132.147.112.65) 1.150 ms 1.205 ms 1.148 ms 4 203.208.190.241 (203.208.190.241) 1.586 ms 1.632 ms 1.457 ms 5 203.208.158.5 (203.208.158.5) 1.352 ms 203.208.158.25 (203.208.158.25) 1.564 ms 203.208.158.17 (203.208.158.17) 1.546 ms 6 203.208.171.86 (203.208.171.86) 180.558 ms 203.208.154.46 (203.208.154.46) 173.444 ms 203.208.171.138 (203.208.171.138) 185.412 ms 7 203.208.172.146 (203.208.172.146) 185.745 ms 203.208.158.118 (203.208.158.118) 177.875 ms 203.208.171.118 (203.208.171.118) 179.536 ms 8 ae12.lax21.ip4.gtt.net (46.33.80.33) 250.320 ms 240.556 ms 240.547 ms 9 et-5-1-0.chi11.ip4.gtt.net (141.136.110.18) 264.002 ms 249.863 ms 262.364 ms 10 as23352.chi11.ip4.gtt.net (199.229.229.210) 246.427 ms 246.627 ms 246.499 ms 11 ae4.cr2.ord6.us.scnet.net (204.93.204.87) 250.055 ms ae12.cr2.ord6.us.scnet.net (204.93.204.83) 255.018 ms 248.397 ms 12 ae2.ar3.ord6.us.scnet.net (204.93.204.111) 248.383 ms 262.558 ms 251.149 ms 13 as36597.ge-11-2-8.ar3.ord6.us.scnet.net (75.102.1.66) 240.511 ms 245.123 ms 245.365 ms 14 pkg.omniti.com (199.15.226.221) 250.099 ms 250.100 ms 240.571 ms root at sgbk02:~# Is there a problem with repo? What is this error that I am seeing? Thanks & Regards. -------------- next part -------------- An HTML attachment was scrubbed... URL: From danmcd at omniti.com Tue Jun 2 13:40:30 2015 From: danmcd at omniti.com (Dan McDonald) Date: Tue, 2 Jun 2015 09:40:30 -0400 Subject: [OmniOS-discuss] Problem with Omni R151014 upgrade In-Reply-To: References: Message-ID: <864D9706-FB7B-4BA3-A042-1EFA21EC7934@omniti.com> > On Jun 2, 2015, at 5:41 AM, Lawrence Giam wrote: > > 1: Framework error: code: 18 reason: transfer closed with 101471 bytes remaining to read > URL: 'http://pkg.omniti.com/omnios/r151014/omnios/file/1/135914f1bfb857941cb5b16641fb523bb22b0326'. (happened 4 times) > 2: Framework error: code: 18 reason: transfer closed with 14577 bytes remaining to read > URL: 'http://pkg.omniti.com/omnios/r151014/omnios/file/1/1f8d529160acc1839550884b4cec080fea55fb8b'. > 3: Framework error: code: 18 reason: transfer closed with 275217 bytes remaining to read > URL: 'http://pkg.omniti.com/omnios/r151014/omnios/file/1/1f8d529160acc1839550884b4cec080fea55fb8b'. (happened 3 times) I performed 'wget' on all of these individually without an issue. Could you try the same at your end (wget) to make sure no meddling middlebox might be detecting these as virus payloads? Thanks, Dan From graham at thestephensdomain.com Wed Jun 3 09:55:31 2015 From: graham at thestephensdomain.com (Graham Stephens) Date: Wed, 03 Jun 2015 10:55:31 +0100 Subject: [OmniOS-discuss] Wiki change Message-ID: <556ECF13.5040608@thestephensdomain.com> Guys, Where do I go to submit a change to the Wiki? I know the KVM instructions page says it is not finished (at http://omnios.omniti.com/wiki.php/VirtualMachinesKVM), but the instructions as they are will cause a system crash. Ask me how I know! The add device section should read: set match="/dev/zvol/dsk/rpool/zvol" (no '*', and 'dsk' not 'rdsk') The system crash has been reported to the illumOS group. Cheers! From hasslerd at gmx.li Wed Jun 3 10:49:31 2015 From: hasslerd at gmx.li (Dominik Hassler) Date: Wed, 3 Jun 2015 12:49:31 +0200 Subject: [OmniOS-discuss] Wiki change In-Reply-To: <556ECF13.5040608@thestephensdomain.com> References: <556ECF13.5040608@thestephensdomain.com> Message-ID: Could you please explain why to use 'dsk' instead of 'rdsk'? I never had any issues using 'rdsk'... Btw. joyent uses 'rdsk' as well and since they ported KVM to Illumos I suppose they know what they are doing. > Gesendet: Mittwoch, 03. Juni 2015 um 11:55 Uhr > Von: "Graham Stephens" > An: omnios-discuss at lists.omniti.com > Betreff: [OmniOS-discuss] Wiki change > > Guys, > > Where do I go to submit a change to the Wiki? > > I know the KVM instructions page says it is not finished (at > http://omnios.omniti.com/wiki.php/VirtualMachinesKVM), but the > instructions as they are will cause a system crash. > Ask me how I know! > > The add device section should read: > > set match="/dev/zvol/dsk/rpool/zvol" > > (no '*', and 'dsk' not 'rdsk') > > The system crash has been reported to the illumOS group. > > Cheers! > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss > From sjorge+ml at blackdot.be Wed Jun 3 11:01:11 2015 From: sjorge+ml at blackdot.be (Jorge Schrauwen) Date: Wed, 03 Jun 2015 13:01:11 +0200 Subject: [OmniOS-discuss] Wiki change In-Reply-To: References: <556ECF13.5040608@thestephensdomain.com> Message-ID: <5b8c175f768e27d74257f45f33453c83@blackdot.be> rdsk makes more sense to me over dsk. My memory is a bit vague but IIRC rdsk has less os overhead than dsk but I could be wrong. (Probably am) On 2015-06-03 12:49, Dominik Hassler wrote: > Could you please explain why to use 'dsk' instead of 'rdsk'? I never > had any issues using 'rdsk'... > > Btw. joyent uses 'rdsk' as well and since they ported KVM to Illumos I > suppose they know what they are doing. > >> Gesendet: Mittwoch, 03. Juni 2015 um 11:55 Uhr >> Von: "Graham Stephens" >> An: omnios-discuss at lists.omniti.com >> Betreff: [OmniOS-discuss] Wiki change >> >> Guys, >> >> Where do I go to submit a change to the Wiki? >> >> I know the KVM instructions page says it is not finished (at >> http://omnios.omniti.com/wiki.php/VirtualMachinesKVM), but the >> instructions as they are will cause a system crash. >> Ask me how I know! >> >> The add device section should read: >> >> set match="/dev/zvol/dsk/rpool/zvol" >> >> (no '*', and 'dsk' not 'rdsk') >> >> The system crash has been reported to the illumOS group. >> >> Cheers! >> _______________________________________________ >> OmniOS-discuss mailing list >> OmniOS-discuss at lists.omniti.com >> http://lists.omniti.com/mailman/listinfo/omnios-discuss >> > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss From graham at thestephensdomain.com Wed Jun 3 11:17:50 2015 From: graham at thestephensdomain.com (Graham Stephens) Date: Wed, 03 Jun 2015 12:17:50 +0100 Subject: [OmniOS-discuss] Wiki change In-Reply-To: References: <556ECF13.5040608@thestephensdomain.com> Message-ID: <556EE25E.8010404@thestephensdomain.com> I am going off the Oracle docs here: http://docs.oracle.com/cd/E19253-01/819-5461/gbebi/index.html which state: "Note that adding a raw volume to a zone has implicit security risks, even if the volume doesn't correspond to a physical device. In particular, the zone administrator could create malformed file systems that would panic the system when a mount is attempted." It may be overkill, but considering my initial problem was one of a system crash, it stuck in my head. ------------------------------------------- On 03/06/2015 11:49, Dominik Hassler wrote: > Could you please explain why to use 'dsk' instead of 'rdsk'? I never had any issues using 'rdsk'... > > Btw. joyent uses 'rdsk' as well and since they ported KVM to Illumos I suppose they know what they are doing. > >> Gesendet: Mittwoch, 03. Juni 2015 um 11:55 Uhr >> Von: "Graham Stephens" >> An: omnios-discuss at lists.omniti.com >> Betreff: [OmniOS-discuss] Wiki change >> >> Guys, >> >> Where do I go to submit a change to the Wiki? >> >> I know the KVM instructions page says it is not finished (at >> http://omnios.omniti.com/wiki.php/VirtualMachinesKVM), but the >> instructions as they are will cause a system crash. >> Ask me how I know! >> >> The add device section should read: >> >> set match="/dev/zvol/dsk/rpool/zvol" >> >> (no '*', and 'dsk' not 'rdsk') >> >> The system crash has been reported to the illumOS group. >> >> Cheers! >> _______________________________________________ >> OmniOS-discuss mailing list >> OmniOS-discuss at lists.omniti.com >> http://lists.omniti.com/mailman/listinfo/omnios-discuss >> From danmcd at omniti.com Wed Jun 3 14:45:53 2015 From: danmcd at omniti.com (Dan McDonald) Date: Wed, 3 Jun 2015 10:45:53 -0400 Subject: [OmniOS-discuss] Wiki change In-Reply-To: <556ECF13.5040608@thestephensdomain.com> References: <556ECF13.5040608@thestephensdomain.com> Message-ID: <66EFF20E-C7C8-4408-A806-035ED8E1CF81@omniti.com> > On Jun 3, 2015, at 5:55 AM, Graham Stephens wrote: > > Guys, > > Where do I go to submit a change to the Wiki? > > I know the KVM instructions page says it is not finished (at http://omnios.omniti.com/wiki.php/VirtualMachinesKVM), but the instructions as they are will cause a system crash. > Ask me how I know! > > The add device section should read: > > set match="/dev/zvol/dsk/rpool/zvol" > > (no '*', and 'dsk' not 'rdsk') dsk vs. rdsk is a fair suggestion and I can fix it. OTOH, the '*' works for me just fine *because* I have filesystems under zvol: r151014(~)[0]% zfs list | grep zvol rpool/zvol 41.3G 50.3G 96K /rpool/zvol rpool/zvol/oi-disk 41.3G 60.5G 31.1G - r151014(~)[0]% Perhaps if you do not, AND you specify '*' a crash happens. Please put the vmdump.N file somewhere for downloading and analysis. > The system crash has been reported to the illumOS group. 1.) It's "illumos" 2.) I've not yet seen this report. Where did you send it? Dan From graham at thestephensdomain.com Wed Jun 3 18:44:43 2015 From: graham at thestephensdomain.com (Graham Stephens) Date: Wed, 03 Jun 2015 19:44:43 +0100 Subject: [OmniOS-discuss] Wiki change In-Reply-To: References: <556ECF13.5040608@thestephensdomain.com> <66EFF20E-C7C8-4408-A806-035ED8E1CF81@omniti.com> <556F38CE.7090706@thestephensdomain.com> <1412CD62-B4D9-4C33-88AF-5A23573FEB71@omniti.com> <817C9D49-E84F-4375-A52E-939B80E6B8AD@omniti.com> Message-ID: <556F4B1B.7010108@thestephensdomain.com> I'd better look into doing an upgrade as early as possible, before I get this little project completed. That way I'll have less to redo if I make a hash of it :) Thanks for your help, Graham. ---------------------------------------- On 03/06/2015 18:46, Dan McDonald wrote: > >> On Jun 3, 2015, at 1:42 PM, Dan McDonald wrote: >> >> I'd recommend updating to r151014 to make sure you have the latest fixes. I'll analyze this dump on an 006 box, in case the bug still exists later, but I'd recommend using the latest release (which is also the new LTS release). > > You were bitten by this: > > https://www.illumos.org/issues/5421 > > Fixed in r151014, BTW. :) > > Dan > From heinz at licenser.net Thu Jun 4 00:35:33 2015 From: heinz at licenser.net (Heinz Nikolaus Gies) Date: Thu, 4 Jun 2015 02:35:33 +0200 Subject: [OmniOS-discuss] FiFo on OmniOS Message-ID: <3D2F3906-CA3E-493E-A05B-A0D8FB1DC51C@licenser.net> Hi everyone, some of you might already noticed by some tweets flying around but Dan and us have started working making OmniOS a possible hypervisor for Project-FiFo, or if you want to put it the other way around make Project-FiFo an possible management system for OmniOS. We?ve had some great success so far and the feedback to the tweets was excitingly positive so it is time to get the community involved! That means the lot of you! I want to give a rough idea of the plan and current state and the pitfalls. Given I know FiFo better then OmniOS this is from the point of view of a FiFo user, if a point is not clear please let me know and I?ll elaborate - (nearly) full manageability of zones, create, start, stop, reboot, delete - support for snapshots - support for the zone console - SMF support (starting, stopping, viewing services of a zone) - ?FiFo native feel? (read: it should make little difference weather one deployed a ipkg zone or a joyent zone from the workflow) - metrics and performance graphing There are some things that are on the ?nice to have? list but might not land in a first version: - resizing (changing memory/disk/etc) - adding/removing nics And some things that pretty sure will not happen any time soon (due to too much difference from the current way it?s done on SmartOS): - KVM There are some things that are a bit problematic and will probably work but not be exactly the same as they are on SmartOS - NIC management, SmartOS seems to have some custom code to bring up nics and how to handle them so we?ve to be creative here. - SSH key management in zones, Dan has build libzdoor (used for SSH key management in SmartOS) but so far it?s not tested - packaing chunter (the service running in the GZ) and tachyon-meter (the service collecting performance metrics) run flawlessly in the GZ, they are packaged as tarballs and can simply be extracted however there are only pkgin packages for the other components. They will pretty guaranteed just compile on OmniOS given a new enough Erlang version (R17+) exists but there are currently no packages provided for that. (read: need to install from source of have a SmartOS zone around) - it does not look we can get the exact same amount of lock in for zones (mostly for nic management) so a first release will probably required trusted tenants. What is working so far: - creating (with resource control but no networking configured, ipadm acts a bit freaky for me, probably a local issue) - starting / shutting down / halting (commands are all issued but zones tend to hang in shutting_down for me, probably a local issue) - metering (at least the hypervisor itself) - SMF services should be working - the Zone console should be working What can you do? If anyone wants to help, any kind of discussion is great, there are surely things we simply missed and having a more diverse group of people pondering the problem will make for a better result in the end. There will probably be some questions of how to best do things on OmniOS that I would greatly appreciate help with figuring out (see problems with ipadm and shutting_down zones). For brave soles who want to give the current status a spin be warned it is part of the development branch of fifo so things are rather experimental, but, time permits, I?ll try to walk anyone interested though setting it up. --- Cheers, Heinz Nikolaus Gies heinz at licenser.net -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 801 bytes Desc: Message signed with OpenPGP using GPGMail URL: From guorong.koh at gmail.com Thu Jun 4 01:39:57 2015 From: guorong.koh at gmail.com (Guo-Rong Koh) Date: Thu, 04 Jun 2015 11:09:57 +0930 Subject: [OmniOS-discuss] Crashplan alternatives? In-Reply-To: <1432257123.13727.16.camel@gmail.com> References: <1432257123.13727.16.camel@gmail.com> Message-ID: <1433381997.17917.17.camel@gmail.com> To anyone who comes across this post in future. All my Crashplan backup clients (Linux & Windows) eventually failed whe n the 4.2.0 update rolled through. I ended going with option 1, creating a Linux KVM instance and installi ng Crashplan there. Usefully, Crashplan allows you to adopt a computer and I was able to co ntinue backups where they stopped by making the existing Solaris archives visible to the Linux Crashplan KVM. regards, Guo-Rong On Fri, 2015-05-22 at 10:42 +0930, Guo-Rong Koh wrote: > Hello everyone, > > I know others here are running Crashplan for Solaris on OmniOS. > However, given the (not so recent) retirement announcement: > https://helpdesk.code42.com/entries/53070937-Solaris-Platform > -Retirement-Announcement > I'm seeking some discussion and advice on possibilities. > > My original strategy for supporting Linux and Windows clients in a > home server > environment is starting to disintegrate. > Due to this issue: > http://support.code42.com/CrashPlan/Latest/Troubleshooting/Computer > -To > -Computer_Backups_Between_CrashPlan_App_4.2.0_And_Earlier_Versions_Co > ntinuously_Synchronize > my Linux client is now no longer backing up the way it used to > (OmniOS server is on Crashplan 3.7, Linux client automatically > upgraded to 4.2.0). > Thankfully, the Windows clients seem to be OK for now. > > Eventually however, I expect the whole solution to fail when Code42 > EOL all Solaris support. > > My current options are: > 1. Migrate Crashplan to a Linux KVM instance > - this seems like the least effort for now > 2. Find an alternate, multiplatform solution > - thus far I have found nothing suitable > > Do others here have a migration plan? > > regards, > Guo-Rong -------------- next part -------------- An HTML attachment was scrubbed... URL: From tobi at oetiker.ch Thu Jun 4 05:30:41 2015 From: tobi at oetiker.ch (Tobi Oetiker) Date: Thu, 4 Jun 2015 07:30:41 +0200 Subject: [OmniOS-discuss] Crashplan alternatives? In-Reply-To: <1433381997.17917.17.camel@gmail.com> References: <1432257123.13727.16.camel@gmail.com> <1433381997.17917.17.camel@gmail.com> Message-ID: I have not gotten round to trying this, but sinc crashplan is written in java, it might still be possible to run it on omnios. at least the old crashplan was sucessfully used in unsuported environments. cheers tobi Tobias Oetiker tobi at oetiker.ch 062 775 9902 > On 04.06.2015, at 03:39, Guo-Rong Koh wrote: > > To anyone who comes across this post in future. > > All my Crashplan backup clients (Linux & Windows) eventually failed > when the 4.2.0 update rolled through. > > I ended going with option 1, creating a Linux KVM instance and > installing Crashplan there. > Usefully, Crashplan allows you to adopt a computer and I was able to > continue backups where they stopped by making the existing Solaris > archives visible to the Linux Crashplan KVM. > > regards, > Guo-Rong > > >> On Fri, 2015-05-22 at 10:42 +0930, Guo-Rong Koh wrote: >> Hello everyone, >> >> I know others here are running Crashplan for Solaris on OmniOS. >> However, given the (not so recent) retirement announcement: >> https://helpdesk.code42.com/entries/53070937-Solaris-Platform-Retirement-Announcement >> I'm seeking some discussion and advice on possibilities. >> >> My original strategy for supporting Linux and Windows clients in a home server >> environment is starting to disintegrate. >> Due to this issue: >> http://support.code42.com/CrashPlan/Latest/Troubleshooting/Computer-To-Computer_Backups_Between_CrashPlan_App_4.2.0_And_Earlier_Versions_Continuously_Synchronize >> my Linux client is now no longer backing up the way it used to (OmniOS server is on Crashplan 3.7, Linux client automatically upgraded to 4.2.0). >> Thankfully, the Windows clients seem to be OK for now. >> >> Eventually however, I expect the whole solution to fail when Code42 EOL all Solaris support. >> >> My current options are: >> 1. Migrate Crashplan to a Linux KVM instance >> - this seems like the least effort for now >> 2. Find an alternate, multiplatform solution >> - thus far I have found nothing suitable >> >> Do others here have a migration plan? >> >> regards, >> Guo-Rong > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss -------------- next part -------------- An HTML attachment was scrubbed... URL: From paladinemishakal at gmail.com Fri Jun 5 13:37:27 2015 From: paladinemishakal at gmail.com (Lawrence Giam) Date: Fri, 5 Jun 2015 21:37:27 +0800 Subject: [OmniOS-discuss] Problem with Omni R151014 upgrade In-Reply-To: References: <864D9706-FB7B-4BA3-A042-1EFA21EC7934@omniti.com> <53A7B19C-6ED3-41F1-BEE1-F7ADF1D01602@omniti.com> <0C1A3514-C569-4768-8F1A-7FCC7140768C@omniti.com> Message-ID: Hi Dan, How can I get all but exclude the kayak stuff from the pkgrecv? Regards. On Fri, Jun 5, 2015 at 5:39 PM, Lawrence Giam wrote: > Hi Dan, > > What if I setup a local Wholesale Repo Copy with the following: > # zfs create tankNodeR/NFS/omniosR151014repo > # pkgrepo create /tankNodeR/NFS/omniosR151014repo > # pkgrepo set -s /tankNodeR/NFS/omniosR151014repo > publisher/prefix=OmniOSr151014.abc.internal > # pkgrepo info -s /tankNodeR/NFS/omniosR151014repo > PUBLISHER PACKAGES STATUS UPDATED > OmniOSr151014.abc.internal 0 online > 2015-06-05T08:08:24.798197Z > > # svcadm disable pkg/server > # svccfg -s pkg/server setprop pkg/inst_root = > /tankNodeR/NFS/omniosR151014repo > # svccfg -s pkg/server setprop pkg/port = 10000 > # svcadm refresh pkg/server > # svcadm enable pkg/server > > # pkgrecv -s http://pkg.omniti.com/omnios/r151014 -d > file:///tankNodeR/NFS/omniosR151014repo/ 'pkg:/*' > > What other steps do I need to do to get the other internal OmniOS servers > to upgrade the OS from this local repository? > > Is this correct? > root at sgdevOmniOS:~# pkg publisher > PUBLISHER TYPE STATUS URI > omnios origin online > http://pkg.omniti.com/omnios/r151010/ > root at sgdevOmniOS:~# pkg unset-publisher omnios > root at sgdevOmniOS:~# pkg set-publisher -P --set-property > signature-policy=require-signatures -g http://OmniOSr151014.abc.internal/ > omnios > root at sgdevOmniOS:~# > > root at sgdevOmniOS:~# pkg update --be-name=omnios-r151014 entire at 11 > ,5.11-0.151014 > > Thanks and Regards. > > On Thu, Jun 4, 2015 at 11:12 PM, Dan McDonald wrote: > >> I'd actually recommend re-upgrading from scratch with a new BE. You've >> kept incrementally upgrading, and I suspect (thanks to the proxy that's >> been tripping you up) corrupted the BE in question. >> >> Since you're in your active BE still, there's no loss of data or >> important state. >> >> Dan >> >> > -------------- next part -------------- An HTML attachment was scrubbed... URL: From vab at bb-c.de Fri Jun 5 14:18:39 2015 From: vab at bb-c.de (Volker A. Brandt) Date: Fri, 5 Jun 2015 16:18:39 +0200 Subject: [OmniOS-discuss] Problem with Omni R151014 upgrade In-Reply-To: References: <864D9706-FB7B-4BA3-A042-1EFA21EC7934@omniti.com> <53A7B19C-6ED3-41F1-BEE1-F7ADF1D01602@omniti.com> <0C1A3514-C569-4768-8F1A-7FCC7140768C@omniti.com> Message-ID: <21873.44991.757078.936479@glaurung.bb-c.de> > How can I get all but exclude the kayak stuff from the pkgrecv? You would have to create a list (using e.g. "pkgrepo list -s " or "pkg list -avf -g "), remove the kayak packages from that list, and pass the remaining pkg FMRIs to pkgrecv on the command line instead of the wildcard pattern at the end. I do that all the time using grep and cut and friends because I am too lazy to write a proper Perl script. :-) Regards -- Volker -- ------------------------------------------------------------------------ Volker A. Brandt Consulting and Support for Oracle Solaris Brandt & Brandt Computer GmbH WWW: http://www.bb-c.de/ Am Wiesenpfad 6, 53340 Meckenheim, GERMANY Email: vab at bb-c.de Handelsregister: Amtsgericht Bonn, HRB 10513 Schuhgr??e: 46 Gesch?ftsf?hrer: Rainer J.H. Brandt und Volker A. Brandt "When logic and proportion have fallen sloppy dead" From paladinemishakal at gmail.com Fri Jun 5 14:44:52 2015 From: paladinemishakal at gmail.com (Lawrence Giam) Date: Fri, 5 Jun 2015 22:44:52 +0800 Subject: [OmniOS-discuss] Problem with Omni R151014 upgrade In-Reply-To: <21873.44991.757078.936479@glaurung.bb-c.de> References: <864D9706-FB7B-4BA3-A042-1EFA21EC7934@omniti.com> <53A7B19C-6ED3-41F1-BEE1-F7ADF1D01602@omniti.com> <0C1A3514-C569-4768-8F1A-7FCC7140768C@omniti.com> <21873.44991.757078.936479@glaurung.bb-c.de> Message-ID: Hi All, I am trying to upgrade OmniOS but I keep getting network issue so I am think of setting up a local repo so that my internal servers can grab the upgrade file from the local repo, can someone teach me how to do? I followed the wiki but it is very confusing and I need to do this tomorrow. Thanks and Regards. On Fri, Jun 5, 2015 at 10:18 PM, Volker A. Brandt wrote: > > How can I get all but exclude the kayak stuff from the pkgrecv? > > You would have to create a list (using e.g. "pkgrepo list -s " > or "pkg list -avf -g "), remove the kayak packages from that > list, and pass the remaining pkg FMRIs to pkgrecv on the command line > instead of the wildcard pattern at the end. > > I do that all the time using grep and cut and friends because I am too > lazy to write a proper Perl script. :-) > > > Regards -- Volker > -- > ------------------------------------------------------------------------ > Volker A. Brandt Consulting and Support for Oracle Solaris > Brandt & Brandt Computer GmbH WWW: http://www.bb-c.de/ > Am Wiesenpfad 6, 53340 Meckenheim, GERMANY Email: vab at bb-c.de > Handelsregister: Amtsgericht Bonn, HRB 10513 Schuhgr??e: 46 > Gesch?ftsf?hrer: Rainer J.H. Brandt und Volker A. Brandt > > "When logic and proportion have fallen sloppy dead" > -------------- next part -------------- An HTML attachment was scrubbed... URL: From vab at bb-c.de Fri Jun 5 16:00:18 2015 From: vab at bb-c.de (Volker A. Brandt) Date: Fri, 5 Jun 2015 18:00:18 +0200 Subject: [OmniOS-discuss] Problem with Omni R151014 upgrade In-Reply-To: References: <864D9706-FB7B-4BA3-A042-1EFA21EC7934@omniti.com> <53A7B19C-6ED3-41F1-BEE1-F7ADF1D01602@omniti.com> <0C1A3514-C569-4768-8F1A-7FCC7140768C@omniti.com> <21873.44991.757078.936479@glaurung.bb-c.de> Message-ID: <21873.51090.373145.14761@glaurung.bb-c.de> Hi Lawrence! > I am trying to upgrade OmniOS but I keep getting network issue so I > am think of setting up a local repo so that my internal servers can > grab the upgrade file from the local repo, can someone teach me how > to do? I followed the wiki but it is very confusing and I need to do > this tomorrow. I think you were almost there. Quoting from your initial mail: > # zfs create tankNodeR/NFS/omniosR151014repo > # pkgrepo create /tankNodeR/NFS/omniosR151014repo Yes. > # pkgrepo set -s /tankNodeR/NFS/omniosR151014repo > publisher/prefix=OmniOSr151014.abc.internal No. You should use the same publisher as the repo you are copying from: /usr/bin/pkgrepo add-publisher -s /tankNodeR/NFS/omniosR151014repo omnios Then you copy the repo from the source: > # pkgrecv -s http://pkg.omniti.com/omnios/r151014 -d > file:///tankNodeR/NFS/omniosR151014repo/ 'pkg:/*' Yes, except that '*' will be enough. Why did you want to exclude the two kayak packages here? No need really, just get everything. Or is it the kayak stuff that triggers the network issue? Anyway, then you set up the pkg server: > # svccfg -s pkg/server setprop pkg/inst_root = > /tankNodeR/NFS/omniosR151014repo > # svccfg -s pkg/server setprop pkg/port = 10000 > # svcadm refresh pkg/server > # svcadm enable pkg/server Yes. Note that you set the port to be 10000 here. Verify that your server works by connecting to http://OmniOSr151014.abc.internal:10000/ with your browser (note the port). Then, on the client do: > root at sgdevOmniOS:~# pkg unset-publisher omnios > root at sgdevOmniOS:~# pkg set-publisher -P --set-property signature-policy=require-signatures -g http://OmniOSr151014.abc.internal/ omnios Almost. You forgot the port. Also, you can do it in one step: # pkg set-publisher -P --set-property signature-policy=require-signatures -G '*' -g http://OmniOSr151014.abc.internal:10000/ omnios Before you update the client, refresh the local catalog: # pkg refresh --full omnios Then update the client: # pkg update -v --be-name=omnios-r151014 entire I usually do a dryrun with -n before the actual update. Hope this helps -- Volker -- ------------------------------------------------------------------------ Volker A. Brandt Consulting and Support for Oracle Solaris Brandt & Brandt Computer GmbH WWW: http://www.bb-c.de/ Am Wiesenpfad 6, 53340 Meckenheim, GERMANY Email: vab at bb-c.de Handelsregister: Amtsgericht Bonn, HRB 10513 Schuhgr??e: 46 Gesch?ftsf?hrer: Rainer J.H. Brandt und Volker A. Brandt "When logic and proportion have fallen sloppy dead" From paladinemishakal at gmail.com Fri Jun 5 16:12:12 2015 From: paladinemishakal at gmail.com (Lawrence Giam) Date: Sat, 6 Jun 2015 00:12:12 +0800 Subject: [OmniOS-discuss] Problem with Omni R151014 upgrade In-Reply-To: <21873.51090.373145.14761@glaurung.bb-c.de> References: <864D9706-FB7B-4BA3-A042-1EFA21EC7934@omniti.com> <53A7B19C-6ED3-41F1-BEE1-F7ADF1D01602@omniti.com> <0C1A3514-C569-4768-8F1A-7FCC7140768C@omniti.com> <21873.44991.757078.936479@glaurung.bb-c.de> <21873.51090.373145.14761@glaurung.bb-c.de> Message-ID: Hi All, Figure I don't have time to do this so I went on to update my test environment from R150008 to R151012 to R151014 and it works but from R150008 to R151014, it keeps failling. Can the folks at Omniti check what is wrong? Hi Volker, Will try to do what you have teach here. Thanks & Regards. On Sat, Jun 6, 2015 at 12:00 AM, Volker A. Brandt wrote: > Hi Lawrence! > > > > I am trying to upgrade OmniOS but I keep getting network issue so I > > am think of setting up a local repo so that my internal servers can > > grab the upgrade file from the local repo, can someone teach me how > > to do? I followed the wiki but it is very confusing and I need to do > > this tomorrow. > > I think you were almost there. Quoting from your initial mail: > > > # zfs create tankNodeR/NFS/omniosR151014repo > > # pkgrepo create /tankNodeR/NFS/omniosR151014repo > > Yes. > > > # pkgrepo set -s /tankNodeR/NFS/omniosR151014repo > > publisher/prefix=OmniOSr151014.abc.internal > > No. You should use the same publisher as the repo you are copying from: > > /usr/bin/pkgrepo add-publisher -s /tankNodeR/NFS/omniosR151014repo omnios > > Then you copy the repo from the source: > > > # pkgrecv -s http://pkg.omniti.com/omnios/r151014 -d > > file:///tankNodeR/NFS/omniosR151014repo/ 'pkg:/*' > > Yes, except that '*' will be enough. Why did you want to exclude the > two kayak packages here? No need really, just get everything. Or is > it the kayak stuff that triggers the network issue? > > Anyway, then you set up the pkg server: > > > # svccfg -s pkg/server setprop pkg/inst_root = > > /tankNodeR/NFS/omniosR151014repo > > # svccfg -s pkg/server setprop pkg/port = 10000 > > # svcadm refresh pkg/server > > # svcadm enable pkg/server > > Yes. Note that you set the port to be 10000 here. > > Verify that your server works by connecting to > http://OmniOSr151014.abc.internal:10000/ with your browser (note the > port). > > Then, on the client do: > > > root at sgdevOmniOS:~# pkg unset-publisher omnios > > root at sgdevOmniOS:~# pkg set-publisher -P --set-property > signature-policy=require-signatures -g http://OmniOSr151014.abc.internal/ > omnios > > Almost. You forgot the port. Also, you can do it in one step: > > # pkg set-publisher -P --set-property > signature-policy=require-signatures -G > '*' -g http://OmniOSr151014.abc.internal:10000/ omnios > > Before you update the client, refresh the local catalog: > > # pkg refresh --full omnios > > Then update the client: > > # pkg update -v --be-name=omnios-r151014 entire > > I usually do a dryrun with -n before the actual update. > > > Hope this helps -- Volker > -- > ------------------------------------------------------------------------ > Volker A. Brandt Consulting and Support for Oracle Solaris > Brandt & Brandt Computer GmbH WWW: http://www.bb-c.de/ > Am Wiesenpfad 6, 53340 Meckenheim, GERMANY Email: vab at bb-c.de > Handelsregister: Amtsgericht Bonn, HRB 10513 Schuhgr??e: 46 > Gesch?ftsf?hrer: Rainer J.H. Brandt und Volker A. Brandt > > "When logic and proportion have fallen sloppy dead" > -------------- next part -------------- An HTML attachment was scrubbed... URL: From danmcd at omniti.com Fri Jun 5 19:46:28 2015 From: danmcd at omniti.com (Dan McDonald) Date: Fri, 5 Jun 2015 15:46:28 -0400 Subject: [OmniOS-discuss] Hello bloody folks! Check out preliminary gcc5.1 support Message-ID: Please point your code-review eyes and browser tabs here: http://kebe.com/~danmcd/webrevs/oob-gcc51/ It's a LARGE series of changes (8 distinct commits at the moment, over 56 files) that I'm confirming now all build gcc5.1 in lieu of gcc4.8 for non-illumos bits of OmniOS. Moving compilers is a tedious, multi-step process. I *hope* to document it all, but it's very difficult, with lots of stops-and-starts. It's made worse by the property that the OmniOS build system assumes you've done these steps, and *installed* the built system (build with hacked up build.sh files that aren't initially dependent on certain packages). Nonetheless, I'm now making decent forward progress. I hope to have a bloody update with gcc48 swapped out for gcc51, AND having it all work as well. Any code-minded people in the community should have a look at the above webrev, and tell me what you think. Unlike the last compiler change done for r151008, this is a jump from 4 to 5, so "g{cc,++}-N-runtime" has a new value of N, namely 5. The prior versions of gcc runtime are in place (.so.6.0.XX), but the package name has been changed. I've use the "renamed" attribute for gcc/++-4-runtime to aid in this regard. Finally -- I'll have highly variable response latency for the next 10 days or so. I won't be updating the bloody repo until I'm comfortable with the changes, so don't be alarmed I'm disappearing after having altered something. Thanks for your patience. Thanks! Dan From heinz at licenser.net Sat Jun 6 20:45:10 2015 From: heinz at licenser.net (Heinz Nikolaus Gies) Date: Sat, 6 Jun 2015 22:45:10 +0200 Subject: [OmniOS-discuss] ipadm failing after first zone boot Message-ID: <0159FBD5-7904-4892-9B06-A12564FA28A5@licenser.net> I noticed some rather odd behavior, to set up networking I use the following commands: [GZ] dladm create-vnic -l bge0 net4 -p allowed-ips=10.0.0.111/32 (do zone creation wait until svc has reached svc:/milestone/sysconfig:default) [NGZ] ipadm create-if net4 [NGZ] ipadm create-addr -T static -a 10.0.0.111/24 net4/v4 ipadm: Could not create address: Operation not permitted on from-gz interface Now thing become rather strange. Next, if I then reboot the zone and execute the failed command (create-addr) again it works and persists over reboots, however after the first boot it never does. This smells like some kind of bug to me. --- Cheers, Heinz Nikolaus Gies heinz at licenser.net -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 801 bytes Desc: Message signed with OpenPGP using GPGMail URL: From gate03 at landcroft.co.uk Mon Jun 8 02:23:34 2015 From: gate03 at landcroft.co.uk (Michael Mounteney) Date: Mon, 8 Jun 2015 12:23:34 +1000 Subject: [OmniOS-discuss] failure to mount all zfs volumes Message-ID: <20150608122334.7cbb25d9@emerior> This has happened to me on several occasions so I assume it must have happened to others. During boot, one or more zfs filesystems cannot be mounted, for whatever reason, including something as trivial as the mountpoint not being empty. When this happens, dependent services and in particular sshd are not started. Thus, on a headless box, the problem cannot even be diagnosed. The last time this happened to me, it was because two filesystems were both mounted on /home. As this is a polite mailing list I won't say what I really think of this, but frankly it is nonsense of the sort one expects to come out of Redmond, not from a serious server operating system. It seems to me that zfs mounting services should be split between 'system' and 'user' at the very least, and as many services as possible dependent on 'zfs-local-system', OR services should be dependent on the existence of needed configuration files only; e.g., /etc/ssh/sshd_config. Input ? Michael. From mailinglists at qutic.com Mon Jun 8 09:08:06 2015 From: mailinglists at qutic.com (qutic development) Date: Mon, 8 Jun 2015 11:08:06 +0200 Subject: [OmniOS-discuss] FiFo on OmniOS In-Reply-To: <3D2F3906-CA3E-493E-A05B-A0D8FB1DC51C@licenser.net> References: <3D2F3906-CA3E-493E-A05B-A0D8FB1DC51C@licenser.net> Message-ID: <4983AED4-FA43-412F-9C51-4474E39A738A@qutic.com> Hi Heinz, On 04.06.2015, at 02:35, Heinz Nikolaus Gies wrote: > some of you might already noticed by some tweets flying around but Dan and us have started working making OmniOS a possible hypervisor for Project-FiFo, or if you want to put it the other way around make Project-FiFo an possible management system for OmniOS. Sounds great and I would like to test it. How do I install it on OmniOS? Something like `pkg install project-fifo` would be fine :) Best regards Stefan From eric.sproul at circonus.com Mon Jun 8 14:46:21 2015 From: eric.sproul at circonus.com (Eric Sproul) Date: Mon, 8 Jun 2015 10:46:21 -0400 Subject: [OmniOS-discuss] Hello bloody folks! Check out preliminary gcc5.1 support In-Reply-To: References: Message-ID: On Fri, Jun 5, 2015 at 3:46 PM, Dan McDonald wrote: > Any code-minded people in the community should have a look at the above webrev, and tell me what you think. Unlike the last compiler change done for r151008, this is a jump from 4 to 5, so "g{cc,++}-N-runtime" has a new value of N, namely 5. The prior versions of gcc runtime are in place (.so.6.0.XX), but the package name has been changed. I've use the "renamed" attribute for gcc/++-4-runtime to aid in this regard. Dan, Great work so far. Mostly looks good to me, though I'm curious if pkglint is/will be complaining about redundant require dependencies. A lot of these build scripts predate the introduction of pkgdepend/pkglint, and IMO the changes where it's just updating DEPENDS_IPS could just be a straight removal of the entire DEPENDS_IPS line. pkgdepend should easily take care of any/all ELF dependencies now. Also, NSPR: there's version changes for the package along with the compiler change. Was that intentional? I don't much care either way, just curious if something slipped in by mistake. Eric From mir at miras.org Mon Jun 8 18:08:50 2015 From: mir at miras.org (Michael Rasmussen) Date: Mon, 8 Jun 2015 20:08:50 +0200 Subject: [OmniOS-discuss] CPU advice Message-ID: <20150608200850.355ba110@sleipner.datanom.net> Hi all, I am considering to build a new small ZFS storage server which will host 4 to 8 SATA disks connected on a HBA and SSD for log and cache connected through on-board SATA 3. Motherboard is Asrock E3C224 with 2x8 GB ECC RAM. I have these 3 CPU's in consideration: 1) Pentium G3450T 2) Core i3 I3-4130T 3) Xeon E3-1220V3 Taking the price tag and power consumption into consideration what would be the optimal choice? What performance can I expect from the above listed CPU's? If performance is also taking into consideration will that change the optimal choice? -- Hilsen/Regards Michael Rasmussen Get my public GnuPG keys: michael rasmussen cc http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E mir datanom net http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C mir miras org http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 -------------------------------------------------------------- /usr/games/fortune -es says: Rule the Empire through force. -- Shogun Tokugawa -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From matthew.lagoe at subrigo.net Mon Jun 8 18:14:32 2015 From: matthew.lagoe at subrigo.net (Matthew Lagoe) Date: Mon, 8 Jun 2015 11:14:32 -0700 Subject: [OmniOS-discuss] CPU advice In-Reply-To: <20150608200850.355ba110@sleipner.datanom.net> References: <20150608200850.355ba110@sleipner.datanom.net> Message-ID: <011501d0a216$fb3fb160$f1bf1420$@subrigo.net> I have never really ran into cpu issues using ZFS with LZ4 compression even at fairly high usage so I imagine any of them would work fine. -----Original Message----- From: OmniOS-discuss [mailto:omnios-discuss-bounces at lists.omniti.com] On Behalf Of Michael Rasmussen Sent: Monday, June 08, 2015 11:09 AM To: omnios-discuss Subject: [OmniOS-discuss] CPU advice Hi all, I am considering to build a new small ZFS storage server which will host 4 to 8 SATA disks connected on a HBA and SSD for log and cache connected through on-board SATA 3. Motherboard is Asrock E3C224 with 2x8 GB ECC RAM. I have these 3 CPU's in consideration: 1) Pentium G3450T 2) Core i3 I3-4130T 3) Xeon E3-1220V3 Taking the price tag and power consumption into consideration what would be the optimal choice? What performance can I expect from the above listed CPU's? If performance is also taking into consideration will that change the optimal choice? -- Hilsen/Regards Michael Rasmussen Get my public GnuPG keys: michael rasmussen cc http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E mir datanom net http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C mir miras org http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 -------------------------------------------------------------- /usr/games/fortune -es says: Rule the Empire through force. -- Shogun Tokugawa From bfriesen at simple.dallas.tx.us Mon Jun 8 19:07:38 2015 From: bfriesen at simple.dallas.tx.us (Bob Friesenhahn) Date: Mon, 8 Jun 2015 14:07:38 -0500 (CDT) Subject: [OmniOS-discuss] CPU advice In-Reply-To: <20150608200850.355ba110@sleipner.datanom.net> References: <20150608200850.355ba110@sleipner.datanom.net> Message-ID: On Mon, 8 Jun 2015, Michael Rasmussen wrote: > > I have these 3 CPU's in consideration: > 1) Pentium G3450T > 2) Core i3 I3-4130T > 3) Xeon E3-1220V3 > > Taking the price tag and power consumption into consideration what > would be the optimal choice? Choose one which lists support for ECC. You need to check Intel's product page on each to make sure that the CPU also supports ECC. Chips with 'Xeon' in the name are usually the server-centric models, with more memory bandwidth. > What performance can I expect from the above listed CPU's? For zfs, the CPU itself is rarely the bottleneck for performance. Spend your money on more RAM. Bob -- Bob Friesenhahn bfriesen at simple.dallas.tx.us, http://www.simplesystems.org/users/bfriesen/ GraphicsMagick Maintainer, http://www.GraphicsMagick.org/ From lists at marzocchi.net Mon Jun 8 19:13:42 2015 From: lists at marzocchi.net (Olaf Marzocchi) Date: Mon, 8 Jun 2015 21:13:42 +0200 Subject: [OmniOS-discuss] CPU advice In-Reply-To: <011501d0a216$fb3fb160$f1bf1420$@subrigo.net> References: <20150608200850.355ba110@sleipner.datanom.net> <011501d0a216$fb3fb160$f1bf1420$@subrigo.net> Message-ID: For that kind of work, I would go for the middle-tier i3, the Pentium may be too little (especially if you do more, later) and the Xeon is overkill. I have the Xeon E3-1220v2 and it is idling all the time. Idle power should be comparable, a little better on the i3. If you need VT-D then it?s only the Xeon. Olaf > Il giorno 08/giu/2015, alle ore 20:14, Matthew Lagoe ha scritto: > > I have never really ran into cpu issues using ZFS with LZ4 compression even > at fairly high usage so I imagine any of them would work fine. > > -----Original Message----- > From: OmniOS-discuss [mailto:omnios-discuss-bounces at lists.omniti.com] On > Behalf Of Michael Rasmussen > Sent: Monday, June 08, 2015 11:09 AM > To: omnios-discuss > Subject: [OmniOS-discuss] CPU advice > > Hi all, > > I am considering to build a new small ZFS storage server which will host 4 > to 8 SATA disks connected on a HBA and SSD for log and cache connected > through on-board SATA 3. Motherboard is Asrock E3C224 with > 2x8 GB ECC RAM. > > I have these 3 CPU's in consideration: > 1) Pentium G3450T > 2) Core i3 I3-4130T > 3) Xeon E3-1220V3 > > Taking the price tag and power consumption into consideration what would be > the optimal choice? > > What performance can I expect from the above listed CPU's? > > If performance is also taking into consideration will that change the > optimal choice? > > -- > Hilsen/Regards > Michael Rasmussen > > Get my public GnuPG keys: > michael rasmussen cc > http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E > mir datanom net > http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C > mir miras org > http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 > -------------------------------------------------------------- > /usr/games/fortune -es says: > Rule the Empire through force. > -- Shogun Tokugawa > > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss From skiselkov.ml at gmail.com Mon Jun 8 20:05:56 2015 From: skiselkov.ml at gmail.com (Saso Kiselkov) Date: Mon, 08 Jun 2015 22:05:56 +0200 Subject: [OmniOS-discuss] CPU advice In-Reply-To: <20150608200850.355ba110@sleipner.datanom.net> References: <20150608200850.355ba110@sleipner.datanom.net> Message-ID: <5575F5A4.7030109@gmail.com> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA256 On 6/8/15 8:08 PM, Michael Rasmussen wrote: > Hi all, > > I am considering to build a new small ZFS storage server which > will host 4 to 8 SATA disks connected on a HBA and SSD for log and > cache connected through on-board SATA 3. Motherboard is Asrock > E3C224 with 2x8 GB ECC RAM. > > I have these 3 CPU's in consideration: 1) Pentium G3450T 2) Core i3 > I3-4130T 3) Xeon E3-1220V3 > > Taking the price tag and power consumption into consideration what > would be the optimal choice? > > What performance can I expect from the above listed CPU's? > > If performance is also taking into consideration will that change > the optimal choice? Any one of these is a fine choice, even the G3450T is plenty fast for most "small user" needs and will happily saturate a couple of 1Gbit/s links with data. And all of the models you listed support ECC - wise choice. For me, I'd get the G3450T and spend the extra money on an extra 16GB of memory, or maybe a decent SAS HBA. Cheers, - -- Saso -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQIcBAEBCAAGBQJVdfWkAAoJEAREJfuiSrcaPJoP/3vTtEhFWzjkIg2ZWJNVqHg5 9wSNg3IuIUHtyTv/mtoEWSK5FvwwDGuO2LKKiOOc++xv5eFrRIB0OGWZHQKSTGuh Kna3sq0mAYIG0liw1l9B8HKzPhpJfUFyHIFJ8L4VqUz6K+sTv/4btXJT9rG1Ed1f EEJslqOvymj6O33kSlAUiuDA6pZFQURVR1508PoRHFfXKuLMww5kFutzPASHYhcm Z5kDF0vm1XaEjTUtCSEsyQgC0k8ZLashXnAHRuuC6W7lCuwcBTG8q+tVFUqxPm+k WxgOthhXTNxroK2k3tp1X3bfDdfc64NJHNgIwD8sjUokFDzgtaRCQrirJnTeM6R2 Cl1ORfdNE6FAKaRX7P2jgFZUuRMGZRfVmksqfp2pcbQsQzcddt4bD2Ryz8wqN50x rQUl+cQGiewpb1jfslp3wZBWkDYxoPALZQS4nweMH9yA8HL8zxTDlW3p8eK2vQ4E vgu+xrixSvLn6oDx7pxOVg7EnoRe0Lr0o0dsP7yE/y2IXl+ythcsjNMEPBlMl5f7 CpHs7HeekX0AlcE3fhhKyI4y2KIL8vOURu/ENhgfM2pujjsCAnh9se5WeeKuCre/ 1JG/s/oqH7KeGo6WU1nlzKYTCnqJzl82LeUcVQ2s36doOJZGPRVIHUYt1T3ZBzwW 0khd2iK/WrgofuhIPIAp =cnhA -----END PGP SIGNATURE----- From mir at miras.org Mon Jun 8 20:55:37 2015 From: mir at miras.org (Michael Rasmussen) Date: Mon, 8 Jun 2015 22:55:37 +0200 Subject: [OmniOS-discuss] CPU advice In-Reply-To: References: <20150608200850.355ba110@sleipner.datanom.net> Message-ID: <20150608225537.71f69bfe@sleipner.datanom.net> On Mon, 8 Jun 2015 14:07:38 -0500 (CDT) Bob Friesenhahn wrote: > On Mon, 8 Jun 2015, Michael Rasmussen wrote: > > > Choose one which lists support for ECC. You need to check Intel's product page on each to make sure that the CPU also supports ECC. Chips with 'Xeon' in the name are usually the server-centric models, with more memory bandwidth. > All Haswell Pentium and I3 should support ECC. > > For zfs, the CPU itself is rarely the bottleneck for performance. Spend your money on more RAM. > See my reply to Saso -- Hilsen/Regards Michael Rasmussen Get my public GnuPG keys: michael rasmussen cc http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E mir datanom net http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C mir miras org http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 -------------------------------------------------------------- /usr/games/fortune -es says: Q: How many IBM 370's does it take to execute a job? A: Four, three to hold it down, and one to rip its head off. -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From mir at miras.org Mon Jun 8 21:00:57 2015 From: mir at miras.org (Michael Rasmussen) Date: Mon, 8 Jun 2015 23:00:57 +0200 Subject: [OmniOS-discuss] CPU advice In-Reply-To: <5575F5A4.7030109@gmail.com> References: <20150608200850.355ba110@sleipner.datanom.net> <5575F5A4.7030109@gmail.com> Message-ID: <20150608230057.37dd2fa6@sleipner.datanom.net> On Mon, 08 Jun 2015 22:05:56 +0200 Saso Kiselkov wrote: > > Any one of these is a fine choice, even the G3450T is plenty fast for > most "small user" needs and will happily saturate a couple of 1Gbit/s > links with data. And all of the models you listed support ECC - wise > choice. For me, I'd get the G3450T and spend the extra money on an > extra 16GB of memory, or maybe a decent SAS HBA. > The usecase for the storage is LUN's for KVM's via Comstar and I should think Comstar can take advantage of more CPU threads? As another side note the network is Infiband 10Gbit. Taking this into the equation I am a bit worried about performance with the Pentium since this only have 2 threads. Should I be worried or is the Pentium still sufficient? -- Hilsen/Regards Michael Rasmussen Get my public GnuPG keys: michael rasmussen cc http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E mir datanom net http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C mir miras org http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 -------------------------------------------------------------- /usr/games/fortune -es says: Always the dullness of the fool is the whetstone of the wits. -- William Shakespeare, "As You Like It" -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From heinz at licenser.net Mon Jun 8 21:01:04 2015 From: heinz at licenser.net (Heinz Nikolaus Gies) Date: Mon, 8 Jun 2015 23:01:04 +0200 Subject: [OmniOS-discuss] FiFo on OmniOS In-Reply-To: <4983AED4-FA43-412F-9C51-4474E39A738A@qutic.com> References: <3D2F3906-CA3E-493E-A05B-A0D8FB1DC51C@licenser.net> <4983AED4-FA43-412F-9C51-4474E39A738A@qutic.com> Message-ID: <2D5A03F0-4244-4D91-96FD-A80398DE8583@licenser.net> Hi Stevan, yap I suppose that would be great but I haven?t yet figured out the ipgk build system (any help appreciated ;). For now it?s manual compilation on OmniOS (not as horrible as it sounds!) or using a SmartOs system with pkgin. if you want to compile it it?s rather simple actually this is for SmartOS https://gist.github.com/Licenser/c0573322309d973c73b1#file-setup-sh the main difference will be that you don?t have pkgin but pkg but it should be nearly the same (I?d try to look them up but the pkg search is giving 503?s here) Please be aware so that it is in the test branch and highly experimental ;) Cheers, Heinz --- Cheers, Heinz Nikolaus Gies heinz at licenser.net > On Jun 8, 2015, at 11:08, qutic development wrote: > > Hi Heinz, > > On 04.06.2015, at 02:35, Heinz Nikolaus Gies wrote: > >> some of you might already noticed by some tweets flying around but Dan and us have started working making OmniOS a possible hypervisor for Project-FiFo, or if you want to put it the other way around make Project-FiFo an possible management system for OmniOS. > > Sounds great and I would like to test it. How do I install it on OmniOS? Something like `pkg install project-fifo` would be fine :) > > Best regards > Stefan -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 801 bytes Desc: Message signed with OpenPGP using GPGMail URL: From Robert.Brock at 2hoffshore.com Tue Jun 9 15:05:20 2015 From: Robert.Brock at 2hoffshore.com (Robert A. Brock) Date: Tue, 9 Jun 2015 15:05:20 +0000 Subject: [OmniOS-discuss] Zpool export while resilvering? Message-ID: <2859482C466CCA42AD9B84B9F56212301023C30A@2H199.2hukwok2.local> List, This is probably a silly question, but I've honestly never tried this and don't have a test machine handy at the moment - can a pool be safely exported and re-imported later if it is currently resilvering? In the way of a bit of background, I have a pool made up with thirty or so 4TB Seagate disks with a firmware issue that results in their max temp being set at 40C as opposed to 60C. This particular pool is in an office building in Texas, in an air-conditioned server room. The condenser for this unit is in the building's plenum and when the building a/c goes off over weekends in the summer my server room a/c struggles and temps run up to about 85F or so. This is causing my pool to drop random disks lately (fmadm reports high temp and they get marked as removed from the pool), and I've only just narrowed it down to this firmware issue. Seagate firmware update utility is apparently Windows only, so the disks must come out for the firmware update, but the pool is resilvering several disks with days remaining, hence my original query. Regards, Rob -------------- next part -------------- An HTML attachment was scrubbed... URL: From mir at miras.org Tue Jun 9 15:45:13 2015 From: mir at miras.org (Michael Rasmussen) Date: Tue, 9 Jun 2015 17:45:13 +0200 Subject: [OmniOS-discuss] bug in powertop Message-ID: <20150609174513.66050f42@sleipner.datanom.net> Hi all, Anybody able to start powertop using some of the available options? root at nas:/root# powertop -d 1 Segmentation Fault (core dumped) root at nas:/root# powertop -t 10 Segmentation Fault (core dumped) root at nas:/root# powertop -d 1 -v Segmentation Fault (core dumped) root at nas:/root# powertop -c 0 Segmentation Fault (core dumped) The only way I am able to start powertop is using defaults. Omnios: OmniOS v11 r151014 -- Hilsen/Regards Michael Rasmussen Get my public GnuPG keys: michael rasmussen cc http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E mir datanom net http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C mir miras org http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 -------------------------------------------------------------- /usr/games/fortune -es says: He hated being thought of as one of those people that wore stupid ornamental armour. It was gilt by association. -- Terry Pratchett, "Night Watch" -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From dave-oo at pooserville.com Tue Jun 9 18:31:38 2015 From: dave-oo at pooserville.com (Dave Pooser) Date: Tue, 09 Jun 2015 13:31:38 -0500 Subject: [OmniOS-discuss] Zpool export while resilvering? Message-ID: >This is probably a silly question, but I?ve honestly never tried this and >don?t have a test machine handy at the moment ? can a pool be safely >exported and re-imported later if it is currently resilvering? > >In the way of a bit of background, I have a pool made up with thirty or >so 4TB Seagate disks with a firmware issue that results in their max temp >being set at 40C as opposed to 60C. This particular pool is > in an office building in Texas, in an air-conditioned server room. The >condenser for this unit is in the building?s plenum and when the building >a/c goes off over weekends in the summer my server room a/c struggles and >temps run up to about 85F or so. This > is causing my pool to drop random disks lately (fmadm reports high temp >and they get marked as removed from the pool), and I?ve only just >narrowed it down to this firmware issue. Seagate firmware update utility >is apparently Windows only, so the disks must > come out for the firmware update, but the pool is resilvering several >disks with days remaining, hence my original query. Not an answer to your question, but the approach I'd take is renting a portable 110V or 220V A/C unit from somebody like spot-coolers.com to get you through the resilver, then apply the firmware update. (And then I'd start trying to convince management that it's worth adding a unit permanently -- our Office Pro 24 24000BTU/hr cost us under $4k back in 2011.) -- Dave Pooser Cat-Herder-in-Chief, Pooserville.com From richard.elling at richardelling.com Tue Jun 9 18:56:16 2015 From: richard.elling at richardelling.com (Richard Elling) Date: Tue, 9 Jun 2015 11:56:16 -0700 Subject: [OmniOS-discuss] Zpool export while resilvering? In-Reply-To: <2859482C466CCA42AD9B84B9F56212301023C30A@2H199.2hukwok2.local> References: <2859482C466CCA42AD9B84B9F56212301023C30A@2H199.2hukwok2.local> Message-ID: <5BC06552-FEA8-4926-877C-D48166EA92A9@richardelling.com> > On Jun 9, 2015, at 8:05 AM, Robert A. Brock wrote: > > List, > > This is probably a silly question, but I?ve honestly never tried this and don?t have a test machine handy at the moment ? can a pool be safely exported and re-imported later if it is currently resilvering? yes. > > In the way of a bit of background, I have a pool made up with thirty or so 4TB Seagate disks with a firmware issue that results in their max temp being set at 40C as opposed to 60C. yep, this is the broken 003 firmware from Seagate, know it well :-P > This particular pool is in an office building in Texas, in an air-conditioned server room. The condenser for this unit is in the building?s plenum and when the building a/c goes off over weekends in the summer my server room a/c struggles and temps run up to about 85F or so. This is causing my pool to drop random disks lately (fmadm reports high temp and they get marked as removed from the pool), and I?ve only just narrowed it down to this firmware issue. Seagate firmware update utility is apparently Windows only, so the disks must come out for the firmware update, but the pool is resilvering several disks with days remaining, hence my original query. fwflash might work, but it is unlikely Seagate knows anything about it. In any case, firmware upgrades on production system is not a best practice. You can also disable the FMA agent, disk-transport, which is the agent responsible for watching to ensure the temperature does not exceed the "temperature at which the drive vendor says the drive should not be operated" The impact to you is that the same agent detects other failures, such as predicted failures, that probably do need to be noticed. For a short window, this option might work for you. Useful commands: fmstat - shows the current FMA modules, and should include disk-transport fmadm unload disk-transport fmadm load disk-transport The temp checks (and PFA) are done once per hour, by default. -- richard > > Regards, > Rob > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss -------------- next part -------------- An HTML attachment was scrubbed... URL: From narayan.desai at gmail.com Tue Jun 9 19:00:35 2015 From: narayan.desai at gmail.com (Narayan Desai) Date: Tue, 9 Jun 2015 14:00:35 -0500 Subject: [OmniOS-discuss] Zpool export while resilvering? In-Reply-To: References: Message-ID: You might also crank up the priority on your resilver, particularly if it is getting tripped all of the time: http://broken.net/uncategorized/zfs-performance-tuning-for-scrubs-and-resilvers/ -nld On Tue, Jun 9, 2015 at 1:31 PM, Dave Pooser wrote: > >This is probably a silly question, but I?ve honestly never tried this and > >don?t have a test machine handy at the moment ? can a pool be safely > >exported and re-imported later if it is currently resilvering? > > > >In the way of a bit of background, I have a pool made up with thirty or > >so 4TB Seagate disks with a firmware issue that results in their max temp > >being set at 40C as opposed to 60C. This particular pool is > > in an office building in Texas, in an air-conditioned server room. The > >condenser for this unit is in the building?s plenum and when the building > >a/c goes off over weekends in the summer my server room a/c struggles and > >temps run up to about 85F or so. This > > is causing my pool to drop random disks lately (fmadm reports high temp > >and they get marked as removed from the pool), and I?ve only just > >narrowed it down to this firmware issue. Seagate firmware update utility > >is apparently Windows only, so the disks must > > come out for the firmware update, but the pool is resilvering several > >disks with days remaining, hence my original query. > > Not an answer to your question, but the approach I'd take is renting a > portable 110V or 220V A/C unit from somebody like spot-coolers.com to get > you through the resilver, then apply the firmware update. (And then I'd > start trying to convince management that it's worth adding a unit > permanently -- our Office Pro 24 24000BTU/hr cost us under $4k back in > 2011.) > -- > Dave Pooser > Cat-Herder-in-Chief, Pooserville.com > > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss > -------------- next part -------------- An HTML attachment was scrubbed... URL: From richard.elling at richardelling.com Tue Jun 9 19:16:31 2015 From: richard.elling at richardelling.com (Richard Elling) Date: Tue, 9 Jun 2015 12:16:31 -0700 Subject: [OmniOS-discuss] Zpool export while resilvering? In-Reply-To: References: Message-ID: <400CBD87-4DE7-496B-9EF4-15C82C828037@richardelling.com> > On Jun 9, 2015, at 12:00 PM, Narayan Desai wrote: > > You might also crank up the priority on your resilver, particularly if it is getting tripped all of the time: > http://broken.net/uncategorized/zfs-performance-tuning-for-scrubs-and-resilvers/ > -nld In general, yes this is a very good post. However, for more recent ZFS and certainly the lastest OmniOS something-14 release, the write throttle has been completely rewritten, positively impacting resilvers. And, with that rewrite, there is a few more tunables at your disposal, while the old ones fade to the bucket of bad memories :-) In most cases, resilver is capped by the time to write to the resilvering device. You can see this in iostat "-x" as the device that is 100% busy with write workload. That said, for this specific case, the drives are not actually failed, just taken offline, so you could have a short resilver session, once they are brought back online. -- richard > > On Tue, Jun 9, 2015 at 1:31 PM, Dave Pooser > wrote: > >This is probably a silly question, but I?ve honestly never tried this and > >don?t have a test machine handy at the moment ? can a pool be safely > >exported and re-imported later if it is currently resilvering? > > > >In the way of a bit of background, I have a pool made up with thirty or > >so 4TB Seagate disks with a firmware issue that results in their max temp > >being set at 40C as opposed to 60C. This particular pool is > > in an office building in Texas, in an air-conditioned server room. The > >condenser for this unit is in the building?s plenum and when the building > >a/c goes off over weekends in the summer my server room a/c struggles and > >temps run up to about 85F or so. This > > is causing my pool to drop random disks lately (fmadm reports high temp > >and they get marked as removed from the pool), and I?ve only just > >narrowed it down to this firmware issue. Seagate firmware update utility > >is apparently Windows only, so the disks must > > come out for the firmware update, but the pool is resilvering several > >disks with days remaining, hence my original query. > > Not an answer to your question, but the approach I'd take is renting a > portable 110V or 220V A/C unit from somebody like spot-coolers.com to get > you through the resilver, then apply the firmware update. (And then I'd > start trying to convince management that it's worth adding a unit > permanently -- our Office Pro 24 24000BTU/hr cost us under $4k back in > 2011.) > -- > Dave Pooser > Cat-Herder-in-Chief, Pooserville.com > > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss -------------- next part -------------- An HTML attachment was scrubbed... URL: From Robert.Brock at 2hoffshore.com Tue Jun 9 19:30:53 2015 From: Robert.Brock at 2hoffshore.com (Robert A. Brock) Date: Tue, 9 Jun 2015 19:30:53 +0000 Subject: [OmniOS-discuss] Zpool export while resilvering? In-Reply-To: <400CBD87-4DE7-496B-9EF4-15C82C828037@richardelling.com> References: <400CBD87-4DE7-496B-9EF4-15C82C828037@richardelling.com> Message-ID: <2859482C466CCA42AD9B84B9F56212301023CC54@2H199.2hukwok2.local> They are failed as far as OmniOS is concerned, from what I can tell: Jun 08 01:08:54 710768e8-2f2b-4b3d-9d4b-a85ef5617219 DISK-8000-12 Major Host : 2hus291 Platform : S5500BC Chassis_id : ............ Product_sn : Fault class : fault.io.disk.over-temperature Affects : dev:///:devid=id1,sd at n5000c5007242271f//scsi_vhci/disk at g5000c5007242271f faulted and taken out of service FRU : "Slot 21" (hc://:product-id=LSI-SAS2X36:server-id=:chassis-id=500304800033213f:serial=S1Z02A8M0000K4361NF4:part=SEAGATE-ST4000NM0023:revision=0003/ses-enclosure=1/bay=20/disk=0) faulty Description : A disk's temperature exceeded the limits established by its manufacturer. Refer to http://illumos.org/msg/DISK-8000-12 for more information. root at 2hus291:/root# zpool status pool0 pool: pool0 state: DEGRADED status: One or more devices is currently being resilvered. The pool will continue to function, possibly in a degraded state. action: Wait for the resilver to complete. scan: resilver in progress since Tue Jun 9 11:11:16 2015 18.8T scanned out of 91.7T at 667M/s, 31h48m to go 591G resilvered, 20.55% done config: NAME STATE READ WRITE CKSUM pool0 DEGRADED 0 0 0 raidz2-0 DEGRADED 0 0 0 c0t5000C50055ECA49Bd0 ONLINE 0 0 0 c0t5000C50055ECA4B3d0 ONLINE 0 0 0 c0t5000C50055ECA587d0 ONLINE 0 0 0 c0t5000C50055ECA6CFd0 ONLINE 0 0 0 c0t5000C50055ECA7F3d0 ONLINE 0 0 0 spare-5 REMOVED 0 0 0 c0t5000C5007242271Fd0 REMOVED 0 0 0 c0t5000C50055EF8A6Fd0 ONLINE 0 0 0 (resilvering) c0t5000C50055ECAB23d0 ONLINE 0 0 0 c0t5000C50055ECABABd0 ONLINE 0 0 0 raidz2-1 ONLINE 0 0 0 c0t5000C50055EE9D87d0 ONLINE 0 0 0 c0t5000C50055EE9E43d0 ONLINE 0 0 0 c0t5000C50055EEA5ABd0 ONLINE 0 0 0 c0t5000C50055EEBA5Fd0 ONLINE 0 0 0 c0t5000C50055EEC1E3d0 ONLINE 0 0 0 c0t5000C500636670BFd0 ONLINE 0 0 0 c0t5000C50055EF8CBBd0 ONLINE 0 0 0 c0t5000C50055EF8D33d0 ONLINE 0 0 0 raidz2-2 ONLINE 0 0 0 c0t5000C50055F7942Fd0 ONLINE 0 0 0 c0t5000C50055F79E03d0 ONLINE 0 0 0 c0t5000C50055F7A8DFd0 ONLINE 0 0 0 c0t5000C50055F81C1Bd0 ONLINE 0 0 0 c0t5000C5005604A42Bd0 ONLINE 0 0 0 c0t5000C5005604A487d0 ONLINE 0 0 0 c0t5000C5005604A74Bd0 ONLINE 0 0 0 c0t5000C5005604A91Bd0 ONLINE 0 0 0 raidz2-4 DEGRADED 0 0 0 c0t5000C500562ED6A3d0 ONLINE 0 0 0 c0t5000C500562F8DEFd0 ONLINE 0 0 0 c0t5000C500562F92D7d0 ONLINE 0 0 0 c0t5000C500562FA0DFd0 ONLINE 0 0 0 c0t5000C500636679EBd0 ONLINE 0 0 0 spare-5 DEGRADED 0 0 14 c0t5000C50057FBB127d0 REMOVED 0 0 0 c0t5000C5006366906Bd0 ONLINE 0 0 0 c0t5000C5006366808Fd0 ONLINE 0 0 0 spare-7 REMOVED 0 0 0 c0t5000C50057FC84F3d0 REMOVED 0 0 0 c0t5000C50063669937d0 ONLINE 0 0 0 logs mirror-3 ONLINE 0 0 0 c13t5003048000308398d0 ONLINE 0 0 0 c13t5003048000308399d0 ONLINE 0 0 0 cache c13t5E83A97000005BC3d0 ONLINE 0 0 0 spares c0t5000C5006366906Bd0 INUSE currently in use c0t5000C50063669937d0 INUSE currently in use c0t5000C50055EF8A6Fd0 INUSE currently in use c0t5000C5006366994Bd0 AVAIL Seems to be this that?s got me: http://www.bigdatajunkie.com/index.php/10-hardware/19-seagate-constellation-es-3-firmware-0003 Can?t interact with them unless I bring them ?out of retirement?: root at 2hus291:/root# cat /etc/devices/retire_store ??Y^P"/scsi_vhci/disk at g5000c500724eacb70rio-store-version(rio-store-magic??`(rio-store-flagsP"/scsi_vhci/disk at g5000c50057fbc1c30rio-store-version(rio-store-magic??`(rio-store-flagsP"/scsi_vhci/disk at g5000c50057fbaf030rio-store-version(rio-store-magic??`(rio-store-flagsP"/scsi_vhci/disk at g5000c50057fbb1270rio-store-version(rio-store-magic??`(rio-store-flagsP"/scsi_vhci/disk at g5000c50057fc84f30rio-store-version(rio-store-magic??`(rio-store-flagsP"/scsi_vhci/disk at g5000c5007242271f0rio-store-version(rio-store-magic??`(rio-store-flags Which, as I understand it, will necessitate a reboot. I?m going to try fwflash to see if I can get the firmware upgraded without having to pull all the disks and hook them up to a windows machine ? has anyone else tried fwflash and Seagate disk firmware with any success? Fwflash seems to be able to see the disks, so I?m hopeful: root at 2hus291:/root# fwflash -l | grep ST4000NM0023 Product : ST4000NM0023 Product : ST4000NM0023 Product : ST4000NM0023 Product : ST4000NM0023 Product : ST4000NM0023 Product : ST4000NM0023 Product : ST4000NM0023 Product : ST4000NM0023 Product : ST4000NM0023 Product : ST4000NM0023 Product : ST4000NM0023 Product : ST4000NM0023 Product : ST4000NM0023 Product : ST4000NM0023 Product : ST4000NM0023 Product : ST4000NM0023 Product : ST4000NM0023 Product : ST4000NM0023 Product : ST4000NM0023 Product : ST4000NM0023 Product : ST4000NM0023 Product : ST4000NM0023 Product : ST4000NM0023 Product : ST4000NM0023 Product : ST4000NM0023 Product : ST4000NM0023 Product : ST4000NM0023 Product : ST4000NM0023 Product : ST4000NM0023 Product : ST4000NM0023 Product : ST4000NM0023 Product : ST4000NM0023 Product : ST4000NM0023 From: OmniOS-discuss [mailto:omnios-discuss-bounces at lists.omniti.com] On Behalf Of Richard Elling Sent: 09 June 2015 20:17 To: Narayan Desai Cc: omnios-discuss at lists.omniti.com Subject: Re: [OmniOS-discuss] Zpool export while resilvering? On Jun 9, 2015, at 12:00 PM, Narayan Desai > wrote: You might also crank up the priority on your resilver, particularly if it is getting tripped all of the time: http://broken.net/uncategorized/zfs-performance-tuning-for-scrubs-and-resilvers/ -nld In general, yes this is a very good post. However, for more recent ZFS and certainly the lastest OmniOS something-14 release, the write throttle has been completely rewritten, positively impacting resilvers. And, with that rewrite, there is a few more tunables at your disposal, while the old ones fade to the bucket of bad memories :-) In most cases, resilver is capped by the time to write to the resilvering device. You can see this in iostat "-x" as the device that is 100% busy with write workload. That said, for this specific case, the drives are not actually failed, just taken offline, so you could have a short resilver session, once they are brought back online. -- richard >This is probably a silly question, but I?ve honestly never tried this and >don?t have a test machine handy at the moment ? can a pool be safely >exported and re-imported later if it is currently resilvering? > -------------- next part -------------- An HTML attachment was scrubbed... URL: From chip at innovates.com Tue Jun 9 20:25:13 2015 From: chip at innovates.com (Schweiss, Chip) Date: Tue, 9 Jun 2015 15:25:13 -0500 Subject: [OmniOS-discuss] Zpool export while resilvering? In-Reply-To: <2859482C466CCA42AD9B84B9F56212301023CC54@2H199.2hukwok2.local> References: <400CBD87-4DE7-496B-9EF4-15C82C828037@richardelling.com> <2859482C466CCA42AD9B84B9F56212301023CC54@2H199.2hukwok2.local> Message-ID: I went through this problem a while back. There are some gotchas in getting them back online and firmware upgraded. The is will not talk to the drive until it has its firmware upgraded or cleared from the fault database. This drives will not flash with multipath enabled either. I ended up clearing the fault manager's database, disabling it and disconnecting half the SAS cables to get them flashed. -Chip On Jun 9, 2015 2:32 PM, "Robert A. Brock" wrote: > They are failed as far as OmniOS is concerned, from what I can tell: > > > > Jun 08 01:08:54 710768e8-2f2b-4b3d-9d4b-a85ef5617219 DISK-8000-12 Major > > > > Host : 2hus291 > > Platform : S5500BC Chassis_id : ............ > > Product_sn : > > > > Fault class : fault.io.disk.over-temperature > > Affects : dev:///:devid=id1,sd at n5000c5007242271f > //scsi_vhci/disk at g5000c5007242271f > > faulted and taken out of service > > FRU : "Slot 21" > (hc://:product-id=LSI-SAS2X36:server-id=:chassis-id=500304800033213f:serial=S1Z02A8M0000K4361NF4:part=SEAGATE-ST4000NM0023:revision=0003/ses-enclosure=1/bay=20/disk=0) > > faulty > > > > Description : A disk's temperature exceeded the limits established by > > its manufacturer. > > Refer to http://illumos.org/msg/DISK-8000-12 for more > > information. > > > > root at 2hus291:/root# zpool status pool0 > > pool: pool0 > > state: DEGRADED > > status: One or more devices is currently being resilvered. The pool will > > continue to function, possibly in a degraded state. > > action: Wait for the resilver to complete. > > scan: resilver in progress since Tue Jun 9 11:11:16 2015 > > 18.8T scanned out of 91.7T at 667M/s, 31h48m to go > > 591G resilvered, 20.55% done > > config: > > > > NAME STATE READ WRITE CKSUM > > pool0 DEGRADED 0 0 0 > > raidz2-0 DEGRADED 0 0 0 > > c0t5000C50055ECA49Bd0 ONLINE 0 0 0 > > c0t5000C50055ECA4B3d0 ONLINE 0 0 0 > > c0t5000C50055ECA587d0 ONLINE 0 0 0 > > c0t5000C50055ECA6CFd0 ONLINE 0 0 0 > > c0t5000C50055ECA7F3d0 ONLINE 0 0 0 > > spare-5 REMOVED 0 0 0 > > c0t5000C5007242271Fd0 REMOVED 0 0 0 > > c0t5000C50055EF8A6Fd0 ONLINE 0 0 0 > (resilvering) > > c0t5000C50055ECAB23d0 ONLINE 0 0 0 > > c0t5000C50055ECABABd0 ONLINE 0 0 0 > > raidz2-1 ONLINE 0 0 0 > > c0t5000C50055EE9D87d0 ONLINE 0 0 0 > > c0t5000C50055EE9E43d0 ONLINE 0 0 0 > > c0t5000C50055EEA5ABd0 ONLINE 0 0 0 > > c0t5000C50055EEBA5Fd0 ONLINE 0 0 0 > > c0t5000C50055EEC1E3d0 ONLINE 0 0 0 > > c0t5000C500636670BFd0 ONLINE 0 0 0 > > c0t5000C50055EF8CBBd0 ONLINE 0 0 0 > > c0t5000C50055EF8D33d0 ONLINE 0 0 0 > > raidz2-2 ONLINE 0 0 0 > > c0t5000C50055F7942Fd0 ONLINE 0 0 0 > > c0t5000C50055F79E03d0 ONLINE 0 0 0 > > c0t5000C50055F7A8DFd0 ONLINE 0 0 0 > > c0t5000C50055F81C1Bd0 ONLINE 0 0 0 > > c0t5000C5005604A42Bd0 ONLINE 0 0 0 > > c0t5000C5005604A487d0 ONLINE 0 0 0 > > c0t5000C5005604A74Bd0 ONLINE 0 0 0 > > c0t5000C5005604A91Bd0 ONLINE 0 0 0 > > raidz2-4 DEGRADED 0 0 0 > > c0t5000C500562ED6A3d0 ONLINE 0 0 0 > > c0t5000C500562F8DEFd0 ONLINE 0 0 0 > > c0t5000C500562F92D7d0 ONLINE 0 0 0 > > c0t5000C500562FA0DFd0 ONLINE 0 0 0 > > c0t5000C500636679EBd0 ONLINE 0 0 0 > > spare-5 DEGRADED 0 0 14 > > c0t5000C50057FBB127d0 REMOVED 0 0 0 > > c0t5000C5006366906Bd0 ONLINE 0 0 0 > > c0t5000C5006366808Fd0 ONLINE 0 0 0 > > spare-7 REMOVED 0 0 0 > > c0t5000C50057FC84F3d0 REMOVED 0 0 0 > > c0t5000C50063669937d0 ONLINE 0 0 0 > > logs > > mirror-3 ONLINE 0 0 0 > > c13t5003048000308398d0 ONLINE 0 0 0 > > c13t5003048000308399d0 ONLINE 0 0 0 > > cache > > c13t5E83A97000005BC3d0 ONLINE 0 0 0 > > spares > > c0t5000C5006366906Bd0 INUSE currently in use > > c0t5000C50063669937d0 INUSE currently in use > > c0t5000C50055EF8A6Fd0 INUSE currently in use > > c0t5000C5006366994Bd0 AVAIL > > > > Seems to be this that?s got me: > > > > > http://www.bigdatajunkie.com/index.php/10-hardware/19-seagate-constellation-es-3-firmware-0003 > > > > Can?t interact with them unless I bring them ?out of retirement?: > > > > root at 2hus291:/root# cat /etc/devices/retire_store > > ??Y^P"/scsi_vhci/disk at g5000c500724eacb70rio-store-version(rio-store-magic > ??`(rio-store-flagsP"/scsi_vhci/disk at g5000c50057fbc1c30rio-store-version > (rio-store-magic?? > `(rio-store-flagsP"/scsi_vhci/disk at g5000c50057fbaf030rio-store-version > (rio-store-magic?? > `(rio-store-flagsP"/scsi_vhci/disk at g5000c50057fbb1270rio-store-version > (rio-store-magic?? > `(rio-store-flagsP"/scsi_vhci/disk at g5000c50057fc84f30rio-store-version > (rio-store-magic?? > `(rio-store-flagsP"/scsi_vhci/disk at g5000c5007242271f0rio-store-version > (rio-store-magic??`(rio-store-flags > > > > Which, as I understand it, will necessitate a reboot. I?m going to try > fwflash to see if I can get the firmware upgraded without having to pull > all the disks and hook them up to a windows machine ? has anyone else tried > fwflash and Seagate disk firmware with any success? Fwflash seems to be > able to see the disks, so I?m hopeful: > > > > root at 2hus291:/root# fwflash -l | grep ST4000NM0023 > > Product : ST4000NM0023 > > Product : ST4000NM0023 > > Product : ST4000NM0023 > > Product : ST4000NM0023 > > Product : ST4000NM0023 > > Product : ST4000NM0023 > > Product : ST4000NM0023 > > Product : ST4000NM0023 > > Product : ST4000NM0023 > > Product : ST4000NM0023 > > Product : ST4000NM0023 > > Product : ST4000NM0023 > > Product : ST4000NM0023 > > Product : ST4000NM0023 > > Product : ST4000NM0023 > > Product : ST4000NM0023 > > Product : ST4000NM0023 > > Product : ST4000NM0023 > > Product : ST4000NM0023 > > Product : ST4000NM0023 > > Product : ST4000NM0023 > > Product : ST4000NM0023 > > Product : ST4000NM0023 > > Product : ST4000NM0023 > > Product : ST4000NM0023 > > Product : ST4000NM0023 > > Product : ST4000NM0023 > > Product : ST4000NM0023 > > Product : ST4000NM0023 > > Product : ST4000NM0023 > > Product : ST4000NM0023 > > Product : ST4000NM0023 > > Product : ST4000NM0023 > > > > > > *From:* OmniOS-discuss [mailto:omnios-discuss-bounces at lists.omniti.com] *On > Behalf Of *Richard Elling > *Sent:* 09 June 2015 20:17 > *To:* Narayan Desai > *Cc:* omnios-discuss at lists.omniti.com > *Subject:* Re: [OmniOS-discuss] Zpool export while resilvering? > > > > > > On Jun 9, 2015, at 12:00 PM, Narayan Desai > wrote: > > > > You might also crank up the priority on your resilver, particularly if it > is getting tripped all of the time: > > > http://broken.net/uncategorized/zfs-performance-tuning-for-scrubs-and-resilvers/ > > -nld > > > > In general, yes this is a very good post. However, for more recent ZFS and > certainly > > the lastest OmniOS something-14 release, the write throttle has been > completely > > rewritten, positively impacting resilvers. And, with that rewrite, there > is a few more > > tunables at your disposal, while the old ones fade to the bucket of bad > memories :-) > > > > In most cases, resilver is capped by the time to write to the resilvering > device. > > You can see this in iostat "-x" as the device that is 100% busy with write > workload. > > That said, for this specific case, the drives are not actually failed, > just taken offline, > > so you could have a short resilver session, once they are brought back > online. > > -- richard > > > > >This is probably a silly question, but I?ve honestly never tried this > and > >don?t have a test machine handy at the moment ? can a pool be safely > >exported and re-imported later if it is currently resilvering? > > > > > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From graham at thestephensdomain.com Tue Jun 9 20:49:41 2015 From: graham at thestephensdomain.com (Graham Stephens) Date: Tue, 09 Jun 2015 21:49:41 +0100 Subject: [OmniOS-discuss] kvm networking Message-ID: <55775165.50703@thestephensdomain.com> This must have been done millions of times already, but as a newbie I'm having great difficulty getting a kvm guest inside a zone to network properly. It seems to me that I've tried every combination of configs I can think of (except one, no doubt ;) ) but can't get anything outside the zone to talk to the guest, with the exception of vnc and ssh (which I presume must be built into the kvm mechanism?). To the best of my knowledge (which isn't that much in this case), I've tried "-net user", "-net tap", "-redir", "hostfwd" in addition to the "-net nic" as shown on the wiki. Linux help pages suggest it should be done by bridging, but "create-bridge" won't let me use the visible vnic as a link. After two days of head-scratching I've decided I need assistance; if anyone can please point me at a working methodology I'd appreciate it. Be warned though I might need it spelling out :) From Robert.Brock at 2hoffshore.com Tue Jun 9 20:59:05 2015 From: Robert.Brock at 2hoffshore.com (Robert A. Brock) Date: Tue, 9 Jun 2015 20:59:05 +0000 Subject: [OmniOS-discuss] Zpool export while resilvering? References: <400CBD87-4DE7-496B-9EF4-15C82C828037@richardelling.com> Message-ID: <2859482C466CCA42AD9B84B9F56212301023D12F@2H199.2hukwok2.local> Fwflash has an issue: root at 2hus291:/root# fwflash -f MegalodonES3-SAS-STD-0004.LOD -d /devices/pci at 0,0/pci8086,340e at 7/pci1000,3040 at 0/iport at f/disk at w5000c50057fbaf01,0:a,raw fwflash: sd-GENERIC firmware image verifier: supplied filename MegalodonES3-SAS-STD-0004.LOD exceeds maximum allowable size of 1468006 bytes Seems I?m not the first: https://www.illumos.org/issues/5012 Argh. I guess I have to pull the disks? From: Robert A. Brock Sent: 09 June 2015 20:31 To: omnios-discuss at lists.omniti.com Subject: RE: [OmniOS-discuss] Zpool export while resilvering? -------------- next part -------------- An HTML attachment was scrubbed... URL: From jdg117 at elvis.arl.psu.edu Tue Jun 9 21:22:40 2015 From: jdg117 at elvis.arl.psu.edu (John D Groenveld) Date: Tue, 09 Jun 2015 17:22:40 -0400 Subject: [OmniOS-discuss] kvm networking In-Reply-To: Your message of "Tue, 09 Jun 2015 21:49:41 BST." <55775165.50703@thestephensdomain.com> References: <55775165.50703@thestephensdomain.com> Message-ID: <201506092122.t59LMejr017004@elvis.arl.psu.edu> In message <55775165.50703 at thestephensdomain.com>, Graham Stephens writes: >This must have been done millions of times already, but as a newbie I'm >having great difficulty getting a kvm guest inside a zone to network With 151014, you don't need to explicitly add /dev/kvm and /dev/dld John groenveld at acm.org From graham at thestephensdomain.com Tue Jun 9 22:27:09 2015 From: graham at thestephensdomain.com (graham at thestephensdomain.com) Date: Tue, 09 Jun 2015 22:27:09 +0000 Subject: [OmniOS-discuss] kvm networking In-Reply-To: <55775BF9.2060304@gmx.li> References: <55775165.50703@thestephensdomain.com> <55775BF9.2060304@gmx.li> Message-ID: <20150609222709.14103w9no6entwe5@202.146.209.5> Quoting "Dominik Hassler" : > you need a dedicated vnic for the kvm. I assume you have one vnic > and use that one for the zone so it is not exclusive for the kvm > guest anymore and won't work to my knowledge. You can always bind > VNC to a unix socket so you don't acutally need a vnic for the zone > (except that one which is used by the zone but most not be > configured within the zone at all). I found that if the vnic has an address for the zone, vnc works but the network access from the guest is sporadic at best. I tried with no address on the vnic, which means vnc doesn't work - but in my case that's not a problem as it's for an email server - and yes, outgoing connections work great, so it is probably how it should be. I can't talk to the email server from outside the zone, though. I don't know how I can pass the vnic to the zone without it being 'aware', unless you can add it as a device instead of a network interface? > Did you have a look at https://github.com/hadfl/kvmadm ? It'll do > all the necessary setup for you for running kvms within zones. I'll have to take a look at that, thanks. From wzmalone at gmail.com Wed Jun 10 05:21:52 2015 From: wzmalone at gmail.com (Zach Malone) Date: Tue, 9 Jun 2015 23:21:52 -0600 Subject: [OmniOS-discuss] bug in powertop In-Reply-To: <20150609174513.66050f42@sleipner.datanom.net> References: <20150609174513.66050f42@sleipner.datanom.net> Message-ID: I gave this a go on 151012 (014 isn't available as an AMI yet, and I can't get the 006 AMI to accept a ssh key pair on instance creation), and saw the same thing. powertop appears to segfault in string_to_decimal / libc.so.1. I then tried rebuilding powertop from the OmniTI illumos-gate build scripts, but found that omnios-build:build/illumos/build.sh is a little fussy for the uninitiated. This sounds like a libc or kernel change, but I can't imagine why it would show up on both 012 and 014. Maybe someone with a working build environment can dig further? --Zach Malone On Tue, Jun 9, 2015 at 9:45 AM, Michael Rasmussen wrote: > Hi all, > > Anybody able to start powertop using some of the available options? > root at nas:/root# powertop -d 1 > Segmentation Fault (core dumped) > root at nas:/root# powertop -t 10 > Segmentation Fault (core dumped) > root at nas:/root# powertop -d 1 -v > Segmentation Fault (core dumped) > root at nas:/root# powertop -c 0 > Segmentation Fault (core dumped) > > The only way I am able to start powertop is using defaults. > > Omnios: OmniOS v11 r151014 > > -- > Hilsen/Regards > Michael Rasmussen > > Get my public GnuPG keys: > michael rasmussen cc > http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E > mir datanom net > http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C > mir miras org > http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 > -------------------------------------------------------------- > /usr/games/fortune -es says: > He hated being thought of as one of those people that wore stupid > ornamental armour. It was gilt by association. > -- Terry Pratchett, "Night Watch" > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss > From minikola at gmail.com Wed Jun 10 07:23:53 2015 From: minikola at gmail.com (Nikola M) Date: Wed, 10 Jun 2015 09:23:53 +0200 Subject: [OmniOS-discuss] Wiki change In-Reply-To: <556EE25E.8010404@thestephensdomain.com> References: <556ECF13.5040608@thestephensdomain.com> <556EE25E.8010404@thestephensdomain.com> Message-ID: <5577E609.7030208@gmail.com> On 06/ 3/15 01:17 PM, Graham Stephens wrote: > I am going off the Oracle docs here: > > http://docs.oracle.com/cd/E19253-01/819-5461/gbebi/index.html > > which state: > > "Note that adding a raw volume to a zone has implicit security risks, > even if the volume doesn't correspond to a physical device. In > particular, the zone administrator could create malformed file systems > that would panic the system when a mount is attempted." > > It may be overkill, but considering my initial problem was one of a > system crash, it stuck in my head. It is for Zones, right? Not to let them use rdsk. On the other hand I need to use raw disk for Virtualbox to make vmdk out of zfs volume. (and for kvm?) Wonder about recommendations for exporting volume to ISCSI. From Robert.Brock at 2hoffshore.com Wed Jun 10 09:09:04 2015 From: Robert.Brock at 2hoffshore.com (Robert A. Brock) Date: Wed, 10 Jun 2015 09:09:04 +0000 Subject: [OmniOS-discuss] Zpool export while resilvering? In-Reply-To: References: <400CBD87-4DE7-496B-9EF4-15C82C828037@richardelling.com> <2859482C466CCA42AD9B84B9F56212301023CC54@2H199.2hukwok2.local> Message-ID: <2859482C466CCA42AD9B84B9F562123010240CD5@2H199.2hukwok2.local> What did you use to flash them? Fwflash just gives an error about firmware file being too large. From: Schweiss, Chip [mailto:chip at innovates.com] Sent: 09 June 2015 21:25 To: Robert A. Brock Cc: omnios-discuss Subject: Re: [OmniOS-discuss] Zpool export while resilvering? I went through this problem a while back. There are some gotchas in getting them back online and firmware upgraded. The is will not talk to the drive until it has its firmware upgraded or cleared from the fault database. This drives will not flash with multipath enabled either. I ended up clearing the fault manager's database, disabling it and disconnecting half the SAS cables to get them flashed. -Chip -------------- next part -------------- An HTML attachment was scrubbed... URL: From graham at thestephensdomain.com Wed Jun 10 10:10:43 2015 From: graham at thestephensdomain.com (Graham Stephens) Date: Wed, 10 Jun 2015 11:10:43 +0100 Subject: [OmniOS-discuss] kvm networking In-Reply-To: <5577698F.3040904@gmx.li> References: <55775165.50703@thestephensdomain.com> <55775BF9.2060304@gmx.li> <20150609222709.14103w9no6entwe5@202.146.209.5> <5577698F.3040904@gmx.li> Message-ID: <55780D23.9070202@thestephensdomain.com> Thanks Dominik, that fits with the way it worked best during my experimentation. The unix socket is a good idea, I had wondered how to get round that. It turns out that my issue was a complete newb one - I used tcpdump from within the guest and found the packets were reaching the OS after all. I had used the CLI to add rules to iptables on the guest for the mail server, not realising that it was dumb and was appending the rules AFTER the default-supplied REJECT rule. After re-writing and reloading the ruleset, all is now well. Thanks for the assistance and the tips. ------------------------------------------- On 09/06/2015 23:32, Dominik Hassler wrote: > Graham, > > With "dedicated vnic" I meant that you don't apply an address for it > within the zone. The zone is aware of it but you should not set it up. > Sorry if my description was a bit misleading... > > If you want to use VNC you'll have to setup a second vnic in the zone > config and apply an address for that one or use a unix socket for VNC > which is my preferred option. > > Adding the vnics as network resources is perfectly fine. > From graham at thestephensdomain.com Wed Jun 10 10:44:53 2015 From: graham at thestephensdomain.com (Graham Stephens) Date: Wed, 10 Jun 2015 11:44:53 +0100 Subject: [OmniOS-discuss] Wiki change In-Reply-To: <5577E609.7030208@gmail.com> References: <556ECF13.5040608@thestephensdomain.com> <556EE25E.8010404@thestephensdomain.com> <5577E609.7030208@gmail.com> Message-ID: <55781525.3030203@thestephensdomain.com> Using dsk for kvm seems to be working for me so far, but I've nothing to compare it with - some of the other guys thought there may be a performance penalty? ------------------------------------ On 10/06/2015 08:23, Nikola M wrote: > On 06/ 3/15 01:17 PM, Graham Stephens wrote: >> >> "Note that adding a raw volume to a zone has implicit security risks, >> even if the volume doesn't correspond to a physical device. In >> particular, the zone administrator could create malformed file systems >> that would panic the system when a mount is attempted." > > It is for Zones, right? Not to let them use rdsk. > On the other hand I need to use raw disk for Virtualbox to make vmdk out > of zfs volume. (and for kvm?) > Wonder about recommendations for exporting volume to ISCSI. From chip at innovates.com Wed Jun 10 10:49:25 2015 From: chip at innovates.com (Schweiss, Chip) Date: Wed, 10 Jun 2015 05:49:25 -0500 Subject: [OmniOS-discuss] Zpool export while resilvering? In-Reply-To: <2859482C466CCA42AD9B84B9F562123010240CD5@2H199.2hukwok2.local> References: <400CBD87-4DE7-496B-9EF4-15C82C828037@richardelling.com> <2859482C466CCA42AD9B84B9F56212301023CC54@2H199.2hukwok2.local> <2859482C466CCA42AD9B84B9F562123010240CD5@2H199.2hukwok2.local> Message-ID: On Wed, Jun 10, 2015 at 4:09 AM, Robert A. Brock < Robert.Brock at 2hoffshore.com> wrote: > What did you use to flash them? Fwflash just gives an error about > firmware file being too large. > Santools. -Chip > > > *From:* Schweiss, Chip [mailto:chip at innovates.com] > *Sent:* 09 June 2015 21:25 > *To:* Robert A. Brock > *Cc:* omnios-discuss > *Subject:* Re: [OmniOS-discuss] Zpool export while resilvering? > > > > I went through this problem a while back. There are some gotchas in > getting them back online and firmware upgraded. The is will not talk to > the drive until it has its firmware upgraded or cleared from the fault > database. > > This drives will not flash with multipath enabled either. > > I ended up clearing the fault manager's database, disabling it and > disconnecting half the SAS cables to get them flashed. > > -Chip > > > > > 2H Offshore Engineering Ltd | Registered in England No. 02790139 | > Registered office: Ferryside, Ferry Road, Norwich NR1 1SW. > > > 2H Offshore is an Acteon company specializing in the design, monitoring > and integrity management of offshore riser and conductor systems. Acteon is > a group of specialist international engineering companies serving the > offshore oil and gas industry. Its focus is on subsea services spanning the > entire life of field. For more information, visit www.acteon.com > > > The information in and/or accompanying this email is intended for the > use of the stated recipient only and may be confidential and/or privileged. > It should not be forwarded or copied nor should its contents be disclosed > in any manner without the express consent of the sender/author. Any views > or opinions presented are solely those of the author and do not necessarily > represent those of 2H. > -------------- next part -------------- An HTML attachment was scrubbed... URL: From danmcd at omniti.com Wed Jun 10 12:36:34 2015 From: danmcd at omniti.com (Dan McDonald) Date: Wed, 10 Jun 2015 08:36:34 -0400 Subject: [OmniOS-discuss] ipadm failing after first zone boot In-Reply-To: <0159FBD5-7904-4892-9B06-A12564FA28A5@licenser.net> References: <0159FBD5-7904-4892-9B06-A12564FA28A5@licenser.net> Message-ID: > On Jun 6, 2015, at 4:45 PM, Heinz Nikolaus Gies wrote: > > I noticed some rather odd behavior, to set up networking I use the following commands: > > [GZ] dladm create-vnic -l bge0 net4 -p allowed-ips=10.0.0.111/32 > (do zone creation wait until svc has reached svc:/milestone/sysconfig:default) I'm assuming you've done a zone boot at this point as well? (You don't say it, but "reached milestone/sysconfig" suggests you have.) > [NGZ] ipadm create-if net4 > [NGZ] ipadm create-addr -T static -a 10.0.0.111/24 net4/v4 > ipadm: Could not create address: Operation not permitted on from-gz interface That *is* really weird. I'm assuming you can reproduce this bug? Also, have you tried just doing the create-addr w/o the create-if? I'm pretty sure you can shortcut, and perhaps it works around the not-permitted problem? Thanks, Dan From danmcd at omniti.com Wed Jun 10 12:46:44 2015 From: danmcd at omniti.com (Dan McDonald) Date: Wed, 10 Jun 2015 08:46:44 -0400 Subject: [OmniOS-discuss] Hello bloody folks! Check out preliminary gcc5.1 support In-Reply-To: References: Message-ID: <5FF7750C-3CFC-434C-A563-2AD1A882096A@omniti.com> > On Jun 8, 2015, at 10:46 AM, Eric Sproul wrote: > > On Fri, Jun 5, 2015 at 3:46 PM, Dan McDonald wrote: >> Any code-minded people in the community should have a look at the above webrev, and tell me what you think. Unlike the last compiler change done for r151008, this is a jump from 4 to 5, so "g{cc,++}-N-runtime" has a new value of N, namely 5. The prior versions of gcc runtime are in place (.so.6.0.XX), but the package name has been changed. I've use the "renamed" attribute for gcc/++-4-runtime to aid in this regard. > > Dan, > Great work so far. Mostly looks good to me, though I'm curious if > pkglint is/will be complaining about redundant require dependencies. Remember -- we don't run pkglint by default. (Some existing package naming schemes make that difficult/impossible.) > A lot of these build scripts predate the introduction of > pkgdepend/pkglint, and IMO the changes where it's just updating > DEPENDS_IPS could just be a straight removal of the entire DEPENDS_IPS > line. pkgdepend should easily take care of any/all ELF dependencies > now. pkgdepend, OTOH, runs. The whole DEPENDS_IPS/BUILD_DEPENDS_IPS/PKG_DEPENDS_IPS is screwy to me. A bit of history, and a suggested what's-right course of action may be in order. > Also, NSPR: there's version changes for the package along with the > compiler change. Was that intentional? I don't much care either way, > just curious if something slipped in by mistake. Threw in the update to see if it was gcc51 ready, it wasn't, but I kept the updates in anyway. Dan From danmcd at omniti.com Wed Jun 10 12:50:23 2015 From: danmcd at omniti.com (Dan McDonald) Date: Wed, 10 Jun 2015 08:50:23 -0400 Subject: [OmniOS-discuss] FiFo on OmniOS In-Reply-To: <2D5A03F0-4244-4D91-96FD-A80398DE8583@licenser.net> References: <3D2F3906-CA3E-493E-A05B-A0D8FB1DC51C@licenser.net> <4983AED4-FA43-412F-9C51-4474E39A738A@qutic.com> <2D5A03F0-4244-4D91-96FD-A80398DE8583@licenser.net> Message-ID: <8604D9FF-0D82-4C45-9B30-F0C1A9A88017@omniti.com> > On Jun 8, 2015, at 5:01 PM, Heinz Nikolaus Gies wrote: > > Hi Stevan, > yap I suppose that would be great but I haven?t yet figured out the ipgk build system (any help appreciated ;). For now it?s manual compilation on OmniOS (not as horrible as it sounds!) or using a SmartOs system with pkgin. When you're closer to ready, I can package it up -- this wouldn't be part of OmniOS's repo, but I suspect it'd have a nice home in the "ms.omniti.com" repo. I'll see. Dan From eric.sproul at circonus.com Wed Jun 10 14:00:08 2015 From: eric.sproul at circonus.com (Eric Sproul) Date: Wed, 10 Jun 2015 10:00:08 -0400 Subject: [OmniOS-discuss] Hello bloody folks! Check out preliminary gcc5.1 support In-Reply-To: <5FF7750C-3CFC-434C-A563-2AD1A882096A@omniti.com> References: <5FF7750C-3CFC-434C-A563-2AD1A882096A@omniti.com> Message-ID: On Wed, Jun 10, 2015 at 8:46 AM, Dan McDonald wrote: > >> On Jun 8, 2015, at 10:46 AM, Eric Sproul wrote: >> >> On Fri, Jun 5, 2015 at 3:46 PM, Dan McDonald wrote: >>> Any code-minded people in the community should have a look at the above webrev, and tell me what you think. Unlike the last compiler change done for r151008, this is a jump from 4 to 5, so "g{cc,++}-N-runtime" has a new value of N, namely 5. The prior versions of gcc runtime are in place (.so.6.0.XX), but the package name has been changed. I've use the "renamed" attribute for gcc/++-4-runtime to aid in this regard. >> >> Dan, >> Great work so far. Mostly looks good to me, though I'm curious if >> pkglint is/will be complaining about redundant require dependencies. > > Remember -- we don't run pkglint by default. (Some existing package naming schemes make that difficult/impossible.) Ah, I just looked at the make_package() function. Out of (perhaps morbid) curiosity, what are the obstacles? > The whole DEPENDS_IPS/BUILD_DEPENDS_IPS/PKG_DEPENDS_IPS is screwy to me. A bit of history, and a suggested what's-right course of action may be in order. OK, so to go _really_ far back into the Dark Ages, our build system was originally created to make SVR4 packages for Solaris 10 in OmniTI's hosting environments. No fancy auto-dependency tools there, so the build had to declare all runtime dependencies manually. That persisted through the transition to SXCE, which was still SVR4. Fast-forward to IPS and OmniOS, which grew organically from the SVR4 builds, and the habit of specifying all deps manually was already deeply ingrained. In The Beginning, it was just "BUILD_DEPENDS" and "DEPENDS". The former obviously being build-time, the latter run-time. But for a while, as we started developing OmniOS, we had a hybrid build system (what would eventually become ms.omniti.com) that could build either SVR4 or IPS packages, depending on the desired target. That's where the "_IPS" variants came in, to distinguish from the (now defunct) "_SVR4" ones. We didn't really know about or grok pkgdepend in the beginning. So now, in the all-IPS world, with pkgdepend, we generally don't need to specify run-time dependencies. pkgdepend can handle ELF, Java, and some scripts (Python, Perl, shell), and maybe a few other minor things. It can't do Perl modules, that's one glaring issue that comes to mind, so there are still cases where you need to specify *some* runtime deps. BUT, since we don't run pkglint, we haven't been good about cleaning up the unnecessary DEPENDS_IPS items, and our packages will end up with either duplicate deps or unnecessary ones, e.g. $ pkg contents -t depend -o type,fmri bash TYPE FMRI require pkg:/SUNWcs at 0.5.11-0.151014 require pkg:/system/library at 0.5.11-0.151014 require system/library/gcc-4-runtime The first two came from pkgdepend, but bash doesn't actually need gcc-4-runtime (verified with ldd). It may have at one time, or someone was just over-zealous in applying an assumed dependency without validation. This is mostly harmless, but as in your case here, it's creating unnecessary work for you to change to gcc5. > >> Also, NSPR: there's version changes for the package along with the >> compiler change. Was that intentional? I don't much care either way, >> just curious if something slipped in by mistake. > > Threw in the update to see if it was gcc51 ready, it wasn't, but I kept the updates in anyway. Understood. Eric From bfriesen at simple.dallas.tx.us Wed Jun 10 14:34:03 2015 From: bfriesen at simple.dallas.tx.us (Bob Friesenhahn) Date: Wed, 10 Jun 2015 09:34:03 -0500 (CDT) Subject: [OmniOS-discuss] bug in powertop In-Reply-To: References: <20150609174513.66050f42@sleipner.datanom.net> Message-ID: On Tue, 9 Jun 2015, Zach Malone wrote: > > This sounds like a libc or kernel change, but I can't imagine why it > would show up on both 012 and 014. Maybe someone with a working build > environment can dig further? I have read that powertop uses DTrace probes under Illumos. Does it link against a dtrace library? Has the interface changed? Another possibility is the internationalization changes (from scratch re-write) in Illumos. Check the locale settings. Locale settings could influence string to decimal conversion. Bob -- Bob Friesenhahn bfriesen at simple.dallas.tx.us, http://www.simplesystems.org/users/bfriesen/ GraphicsMagick Maintainer, http://www.GraphicsMagick.org/ From mir at miras.org Wed Jun 10 14:46:54 2015 From: mir at miras.org (Michael Rasmussen) Date: Wed, 10 Jun 2015 16:46:54 +0200 Subject: [OmniOS-discuss] bug in powertop In-Reply-To: References: <20150609174513.66050f42@sleipner.datanom.net> Message-ID: <20150610164654.3d508fbd@sleipner.datanom.net> On Wed, 10 Jun 2015 09:34:03 -0500 (CDT) Bob Friesenhahn wrote: > > I have read that powertop uses DTrace probes under Illumos. Does it link against a dtrace library? Has the interface changed? > If it is linked dynamically to DTrace then you shouldn't see a segmentation fault unless the source does not check for null-pointer references. Statically linked to DTrace and you should not be able to pass the linker stage. > Another possibility is the internationalization changes (from scratch re-write) in Illumos. Check the locale settings. Locale settings could influence string to decimal conversion. > This could be an explanation but I doubt it. My default locale is C and I have run some test with LANG configured to en, en_US, en_UK but without any change. My best guess is a bug in the code so I will fetch the source and run it through gdb. -- Hilsen/Regards Michael Rasmussen Get my public GnuPG keys: michael rasmussen cc http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E mir datanom net http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C mir miras org http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 -------------------------------------------------------------- /usr/games/fortune -es says: Never keep up with the Joneses. Drag them down to your level. -- Quentin Crisp -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From peter.tribble at gmail.com Wed Jun 10 16:04:20 2015 From: peter.tribble at gmail.com (Peter Tribble) Date: Wed, 10 Jun 2015 17:04:20 +0100 Subject: [OmniOS-discuss] bug in powertop In-Reply-To: <20150609174513.66050f42@sleipner.datanom.net> References: <20150609174513.66050f42@sleipner.datanom.net> Message-ID: On Tue, Jun 9, 2015 at 4:45 PM, Michael Rasmussen wrote: > Hi all, > > Anybody able to start powertop using some of the available options? > root at nas:/root# powertop -d 1 > Segmentation Fault (core dumped) > Oh, that's bad. Look at: http://src.illumos.org/source/xref/illumos-gate/usr/src/cmd/powertop/common/powertop.c#81 which is: char *optarg; So, when getopt_long parses the arguments and pokes things back into optarg, where's it going to end up? Not the right optarg, hence the SEGV when strtod() is called on the local optarg. Delete that line in powertop.c and all should be well. (I get the following as well: powertop: failed to compile P-states (frequencies) program powertop: failed to compile C-states (idle power) program so I'm not sure it's actually functional.) -- -Peter Tribble http://www.petertribble.co.uk/ - http://ptribble.blogspot.com/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From heinz at licenser.net Wed Jun 10 17:46:10 2015 From: heinz at licenser.net (Heinz Nikolaus Gies) Date: Wed, 10 Jun 2015 19:46:10 +0200 Subject: [OmniOS-discuss] ipadm failing after first zone boot In-Reply-To: References: <0159FBD5-7904-4892-9B06-A12564FA28A5@licenser.net> Message-ID: > On Jun 10, 2015, at 14:36, Dan McDonald wrote: > > >> On Jun 6, 2015, at 4:45 PM, Heinz Nikolaus Gies wrote: >> >> I noticed some rather odd behavior, to set up networking I use the following commands: >> >> [GZ] dladm create-vnic -l bge0 net4 -p allowed-ips=10.0.0.111/32 >> (do zone creation wait until svc has reached svc:/milestone/sysconfig:default) > > I'm assuming you've done a zone boot at this point as well? (You don't say it, but "reached milestone/sysconfig" suggests you have.) Yup zone boot was done, and I waited till the last milestone was reached (I forgot that before ;) > >> [NGZ] ipadm create-if net4 >> [NGZ] ipadm create-addr -T static -a 10.0.0.111/24 net4/v4 >> ipadm: Could not create address: Operation not permitted on from-gz interface > > That *is* really weird. I'm assuming you can reproduce this bug? In 100% of the tests I?ve had done yes, if it were a one time thing I?d have just doubted myself and wrote it off as me probably having forgotten something. > > Also, have you tried just doing the create-addr w/o the create-if? I'm pretty sure you can shortcut, and perhaps it works around the not-permitted problem? That I have not tried but I?ll give it a try, my work around right now is actually doing a full re-boot of the zone so it?s not too huge of a problem, but it looked like a bug > > Thanks, > Dan > -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 801 bytes Desc: Message signed with OpenPGP using GPGMail URL: From mir at miras.org Wed Jun 10 18:12:21 2015 From: mir at miras.org (Michael Rasmussen) Date: Wed, 10 Jun 2015 20:12:21 +0200 Subject: [OmniOS-discuss] bug in powertop In-Reply-To: References: <20150609174513.66050f42@sleipner.datanom.net> Message-ID: <20150610201221.2337ccf1@sleipner.datanom.net> On Wed, 10 Jun 2015 17:04:20 +0100 Peter Tribble wrote: > > Delete that line in powertop.c and all should be well. > Do I need to build the entire illumos-gate to recompile powertop? > (I get the following as well: > powertop: failed to compile P-states (frequencies) program > powertop: failed to compile C-states (idle power) program > so I'm not sure it's actually functional.) > You need to run powertop either as root or using sudo. -- Hilsen/Regards Michael Rasmussen Get my public GnuPG keys: michael rasmussen cc http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E mir datanom net http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C mir miras org http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 -------------------------------------------------------------- /usr/games/fortune -es says: Q: What do little WASPs want to be when they grow up? A: The very best person they can possibly be. -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From rafibeyli at gmail.com Thu Jun 11 06:59:07 2015 From: rafibeyli at gmail.com (Hafiz Rafiyev) Date: Thu, 11 Jun 2015 09:59:07 +0300 (EEST) Subject: [OmniOS-discuss] upgrade issue from 151006 to 151014 Message-ID: <1983054235.1350712.1434005947130.JavaMail.zimbra@cantekstil.com.tr> Having trouble upgrading from 151006 to 151014,according http://omnios.omniti.com/wiki.php/Upgrade_to_r151014 getting these errors: /usr/bin/pkg update --be-name=omnios-r151014 entire at 11,5.11-0.151014 Creating Plan | pkg update: No matching version of entire can be installed: Reject: pkg://omnios/entire at 11,5.11-0.151014:20150402T192159Z Reason: All versions matching 'require' dependency pkg:/editor/vim at 7.3,5.11-0.151014 are rejected Reject: pkg://omnios/editor/vim at 7.4.45,5.11-0.151014:20150402T174104Z Reason: All versions matching 'require' dependency pkg:/runtime/perl at 5.16.1,5.11-0.151014 are rejected Reject: pkg://omnios/runtime/perl at 5.16.1,5.11-0.151014:20150402T192953Z Reason: This version is excluded by installed incorporation pkg://omnios/runtime/perl/manual at 5.16.1,5.11-0.151006:20130507T191120Z any helps? From danmcd at omniti.com Thu Jun 11 13:16:03 2015 From: danmcd at omniti.com (Dan McDonald) Date: Thu, 11 Jun 2015 09:16:03 -0400 Subject: [OmniOS-discuss] upgrade issue from 151006 to 151014 In-Reply-To: <1983054235.1350712.1434005947130.JavaMail.zimbra@cantekstil.com.tr> References: <1983054235.1350712.1434005947130.JavaMail.zimbra@cantekstil.com.tr> Message-ID: <940D2F2F-23BA-4BDA-94E5-59FF88363516@omniti.com> > On Jun 11, 2015, at 2:59 AM, Hafiz Rafiyev wrote: > > Having trouble upgrading from 151006 to 151014,according http://omnios.omniti.com/wiki.php/Upgrade_to_r151014 > > getting these errors: > > /usr/bin/pkg update --be-name=omnios-r151014 entire at 11,5.11-0.151014 > Creating Plan | > pkg update: No matching version of entire can be installed: > Reject: pkg://omnios/entire at 11,5.11-0.151014:20150402T192159Z > Reason: All versions matching 'require' dependency pkg:/editor/vim at 7.3,5.11-0.151014 are rejected > Reject: pkg://omnios/editor/vim at 7.4.45,5.11-0.151014:20150402T174104Z > Reason: All versions matching 'require' dependency pkg:/runtime/perl at 5.16.1,5.11-0.151014 are rejected > Reject: pkg://omnios/runtime/perl at 5.16.1,5.11-0.151014:20150402T192953Z > Reason: This version is excluded by installed incorporation pkg://omnios/runtime/perl/manual at 5.16.1,5.11-0.151006:20130507T191120Z > > any helps? Uninstall those ms.omniti.com packages and try again. THen reinstall. Sorry, Dan From wzmalone at gmail.com Thu Jun 11 14:28:01 2015 From: wzmalone at gmail.com (Zach Malone) Date: Thu, 11 Jun 2015 08:28:01 -0600 Subject: [OmniOS-discuss] upgrade issue from 151006 to 151014 In-Reply-To: <940D2F2F-23BA-4BDA-94E5-59FF88363516@omniti.com> References: <1983054235.1350712.1434005947130.JavaMail.zimbra@cantekstil.com.tr> <940D2F2F-23BA-4BDA-94E5-59FF88363516@omniti.com> Message-ID: On Thu, Jun 11, 2015 at 7:16 AM, Dan McDonald wrote: > >> On Jun 11, 2015, at 2:59 AM, Hafiz Rafiyev wrote: >> >> Having trouble upgrading from 151006 to 151014,according http://omnios.omniti.com/wiki.php/Upgrade_to_r151014 >> >> getting these errors: >> >> /usr/bin/pkg update --be-name=omnios-r151014 entire at 11,5.11-0.151014 >> Creating Plan | >> pkg update: No matching version of entire can be installed: >> Reject: pkg://omnios/entire at 11,5.11-0.151014:20150402T192159Z >> Reason: All versions matching 'require' dependency pkg:/editor/vim at 7.3,5.11-0.151014 are rejected >> Reject: pkg://omnios/editor/vim at 7.4.45,5.11-0.151014:20150402T174104Z >> Reason: All versions matching 'require' dependency pkg:/runtime/perl at 5.16.1,5.11-0.151014 are rejected >> Reject: pkg://omnios/runtime/perl at 5.16.1,5.11-0.151014:20150402T192953Z >> Reason: This version is excluded by installed incorporation pkg://omnios/runtime/perl/manual at 5.16.1,5.11-0.151006:20130507T191120Z >> >> any helps? > > Uninstall those ms.omniti.com packages and try again. THen reinstall. > > Sorry, > Dan I don't think those are omniti-ms packages, runtime/perl and editor/vim are both core OS packages. Do you have any package freezes in place as a result of the 151006-151008 upgrade issue? From graham at thestephensdomain.com Thu Jun 11 16:00:29 2015 From: graham at thestephensdomain.com (Graham Stephens) Date: Thu, 11 Jun 2015 17:00:29 +0100 Subject: [OmniOS-discuss] port 10926 Message-ID: <5579B09D.9090806@thestephensdomain.com> Guys, Is there something in OmniOS that 'phones home' occasionally? I happened to be looking at my ISP hub and noticed I had 59 attempts incoming to contact port 10926 from a Maryland U address. As I haven't had my OmniOS box on yet today, I was a bit surprised. Can anyone shed any light on this? From eric.sproul at circonus.com Thu Jun 11 16:20:43 2015 From: eric.sproul at circonus.com (Eric Sproul) Date: Thu, 11 Jun 2015 12:20:43 -0400 Subject: [OmniOS-discuss] port 10926 In-Reply-To: <5579B09D.9090806@thestephensdomain.com> References: <5579B09D.9090806@thestephensdomain.com> Message-ID: On Thu, Jun 11, 2015 at 12:00 PM, Graham Stephens wrote: > Guys, > > Is there something in OmniOS that 'phones home' occasionally? > > I happened to be looking at my ISP hub and noticed I had 59 attempts > incoming to contact port 10926 from a Maryland U address. As I haven't had > my OmniOS box on yet today, I was a bit surprised. Nothing in OmniOS "phones home". Do you have any community package repos configured? There is one run by a community member at http://pkg.cs.umd.edu/ ... that's the only UMD connection I can think of. Eric From danmcd at omniti.com Thu Jun 11 17:08:49 2015 From: danmcd at omniti.com (Dan McDonald) Date: Thu, 11 Jun 2015 13:08:49 -0400 Subject: [OmniOS-discuss] port 10926 In-Reply-To: <5579B09D.9090806@thestephensdomain.com> References: <5579B09D.9090806@thestephensdomain.com> Message-ID: <29F47EA8-D992-4550-B4EE-6BD2FA479279@omniti.com> > On Jun 11, 2015, at 12:00 PM, Graham Stephens wrote: > > Guys, > > Is there something in OmniOS that 'phones home' occasionally? > > I happened to be looking at my ISP hub and noticed I had 59 attempts incoming to contact port 10926 from a Maryland U address. As I haven't had my OmniOS box on yet today, I was a bit surprised. > > Can anyone shed any light on this? > Eric already covered the nothing-phones-home case. When you turn on your OmniOS box, utter: netstat -an | grep 10926 if it shows a result, then there is a process on your OmniOS box listening on that port. IF the netstat shows output, then find out which process has it open: sudo pfiles `ptree | awk '{print $1}'` | egrep '^[0-9]*:|10926' Most of the output will be: 478: /usr/sbin/syslogd 481: /usr/lib/ssh/sshd But you may see an extra line under a process with the matching port. It'll look like this (I used port 22 for an example): 481: /usr/lib/ssh/sshd sockname: AF_INET6 :: port: 22 Hope this helps, Dan From danmcd at omniti.com Thu Jun 11 17:12:10 2015 From: danmcd at omniti.com (Dan McDonald) Date: Thu, 11 Jun 2015 13:12:10 -0400 Subject: [OmniOS-discuss] port 10926 In-Reply-To: <29F47EA8-D992-4550-B4EE-6BD2FA479279@omniti.com> References: <5579B09D.9090806@thestephensdomain.com> <29F47EA8-D992-4550-B4EE-6BD2FA479279@omniti.com> Message-ID: > On Jun 11, 2015, at 1:08 PM, Dan McDonald wrote: > > > IF the netstat shows output, then find out which process has it open: > > sudo pfiles `ptree | awk '{print $1}'` | egrep '^[0-9]*:|10926' BTW, there is work going on upstream to add pid (or at least pid-that-opened) output to netstat(1M). That should make it easier. Until then pfiles(1) is your only friend about this. Dan From graham at thestephensdomain.com Thu Jun 11 18:12:57 2015 From: graham at thestephensdomain.com (Graham Stephens) Date: Thu, 11 Jun 2015 19:12:57 +0100 Subject: [OmniOS-discuss] port 10926 In-Reply-To: References: <5579B09D.9090806@thestephensdomain.com> Message-ID: <5579CFA9.8030507@thestephensdomain.com> On 11/06/2015 17:20, Eric Sproul wrote: > On Thu, Jun 11, 2015 at 12:00 PM, Graham Stephens > wrote: >> Guys, >> >> Is there something in OmniOS that 'phones home' occasionally? >> >> I happened to be looking at my ISP hub and noticed I had 59 attempts >> incoming to contact port 10926 from a Maryland U address. As I haven't had >> my OmniOS box on yet today, I was a bit surprised. > > Nothing in OmniOS "phones home". Do you have any community package > repos configured? There is one run by a community member at > http://pkg.cs.umd.edu/ ... that's the only UMD connection I can think > of. > > Eric > I do have the pkgin bootstrap, but only for bind/isc-dhcp. None of the packages on http://pkg.cs.umd.edu/ ring any bells. From eric.sproul at circonus.com Thu Jun 11 18:16:37 2015 From: eric.sproul at circonus.com (Eric Sproul) Date: Thu, 11 Jun 2015 14:16:37 -0400 Subject: [OmniOS-discuss] port 10926 In-Reply-To: References: <5579B09D.9090806@thestephensdomain.com> <29F47EA8-D992-4550-B4EE-6BD2FA479279@omniti.com> Message-ID: Graham, To make sure I understand, when you say "I had 59 attempts incoming", you meant "incoming from the internet to my machine", correct? As I thought about it, "phone home" could indicate a connection initiated *from* your machine outward to some remote destination, so I wanted to be sure I understood correctly. Eric From daleg at omniti.com Thu Jun 11 18:29:03 2015 From: daleg at omniti.com (Dale Ghent) Date: Thu, 11 Jun 2015 14:29:03 -0400 Subject: [OmniOS-discuss] port 10926 In-Reply-To: <5579B09D.9090806@thestephensdomain.com> References: <5579B09D.9090806@thestephensdomain.com> Message-ID: <25EB697C-7539-479D-9CD4-CD02FA9EA998@omniti.com> > On Jun 11, 2015, at 12:00 PM, Graham Stephens wrote: > > Guys, > > Is there something in OmniOS that 'phones home' occasionally? > > I happened to be looking at my ISP hub and noticed I had 59 attempts incoming to contact port 10926 from a Maryland U address. As I haven't had my OmniOS box on yet today, I was a bit surprised. *incoming* packets? That wouldn?t be phoning home? it sounds more like a probing attempt. OmniOS doesn?t actively (or passively) phone home in any capacity, unless you do something such as run `pkg update -nv` from cron on a daily basis or something of the sort. That would be the closest thing to it, and quite deliberate at that. It?s likely some umd host is misbehaving or is grossly misconfigured in some way. Having ran infrastructure once at one of the UMD campuses, depending on what the IP resolves to might indicate some student project gone awry, a compromised host that?s scanning for specific stuff, or both. Not knowing which campus your mystery packets are coming from, mailing security@ is a generally reliable way of getting in touch with someone regardless of the campus (of which are many, each with their own domain? umd.edu, umbc.edu, umuc.edu, and so on) /dale -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 494 bytes Desc: Message signed with OpenPGP using GPGMail URL: From graham at thestephensdomain.com Thu Jun 11 18:32:59 2015 From: graham at thestephensdomain.com (Graham Stephens) Date: Thu, 11 Jun 2015 19:32:59 +0100 Subject: [OmniOS-discuss] port 10926 In-Reply-To: <29F47EA8-D992-4550-B4EE-6BD2FA479279@omniti.com> References: <5579B09D.9090806@thestephensdomain.com> <29F47EA8-D992-4550-B4EE-6BD2FA479279@omniti.com> Message-ID: <5579D45B.8080907@thestephensdomain.com> On 11/06/2015 18:08, Dan McDonald wrote: > >> On Jun 11, 2015, at 12:00 PM, Graham Stephens wrote: >> >> Guys, >> >> Is there something in OmniOS that 'phones home' occasionally? >> >> I happened to be looking at my ISP hub and noticed I had 59 attempts incoming to contact port 10926 from a Maryland U address. As I haven't had my OmniOS box on yet today, I was a bit surprised. >> >> Can anyone shed any light on this? >> > > Eric already covered the nothing-phones-home case. > > When you turn on your OmniOS box, utter: > > netstat -an | grep 10926 > > if it shows a result, then there is a process on your OmniOS box listening on that port. > > IF the netstat shows output, then find out which process has it open: > > sudo pfiles `ptree | awk '{print $1}'` | egrep '^[0-9]*:|10926' > > Most of the output will be: > > 478: /usr/sbin/syslogd > 481: /usr/lib/ssh/sshd > > But you may see an extra line under a process with the matching port. It'll look like this (I used port 22 for an example): > > 481: /usr/lib/ssh/sshd > sockname: AF_INET6 :: port: 22 > > Hope this helps, > Dan > Thanks Dan. I was just about to report that there's nothing on this port and decided to re-check the hub; it looks as though the dumb log only reports the last port & address attempted, along with the total attempts count for the day, so it's misleading. It isn't telling me much other than it's pick on my IP day! (I wonder where they got my address from? Mmh.) If it continues I'll have to let the traffic through to my firewall instead and log it all, at least I'd get a better picture that way. Thanks again. From jimklimov at cos.ru Thu Jun 11 18:43:33 2015 From: jimklimov at cos.ru (Jim Klimov) Date: Thu, 11 Jun 2015 20:43:33 +0200 Subject: [OmniOS-discuss] port 10926 In-Reply-To: References: <5579B09D.9090806@thestephensdomain.com> Message-ID: <2444C5D0-5A5F-4EA1-83FB-8888D516FA8B@cos.ru> 11 ???? 2015??. 18:20:43 CEST, Eric Sproul ?????: >On Thu, Jun 11, 2015 at 12:00 PM, Graham Stephens > wrote: >> Guys, >> >> Is there something in OmniOS that 'phones home' occasionally? >> >> I happened to be looking at my ISP hub and noticed I had 59 attempts >> incoming to contact port 10926 from a Maryland U address. As I >haven't had >> my OmniOS box on yet today, I was a bit surprised. > >Nothing in OmniOS "phones home". Do you have any community package >repos configured? There is one run by a community member at >http://pkg.cs.umd.edu/ ... that's the only UMD connection I can think >of. > >Eric >_______________________________________________ >OmniOS-discuss mailing list >OmniOS-discuss at lists.omniti.com >http://lists.omniti.com/mailman/listinfo/omnios-discuss Also, if something random tries to contact your addresses, it might be port-probing (e.g. google if some viral rootkits or relays for any OS are known to run on such ports - maybe some "shepherds" are trying to find their "herd") or something else that was never demanded or implicitly initiated by your boxes. Just background radiation of the Internet. Maybe... ;) Jim -- Typos courtesy of K-9 Mail on my Samsung Android From graham at thestephensdomain.com Thu Jun 11 19:31:31 2015 From: graham at thestephensdomain.com (Graham Stephens) Date: Thu, 11 Jun 2015 20:31:31 +0100 Subject: [OmniOS-discuss] port 10926 In-Reply-To: References: <5579B09D.9090806@thestephensdomain.com> <29F47EA8-D992-4550-B4EE-6BD2FA479279@omniti.com> Message-ID: <5579E213.4070606@thestephensdomain.com> Eric, Yes, I meant incoming from the internet - I just wondered if OmniOS had kicked off something it was expecting to talk back. Of course if it was having a two-way conversation it wouldn't get caught on the way back in. --------------------------------------- On 11/06/2015 19:16, Eric Sproul wrote: > Graham, > To make sure I understand, when you say "I had 59 attempts incoming", > you meant "incoming from the internet to my machine", correct? As I > thought about it, "phone home" could indicate a connection initiated > *from* your machine outward to some remote destination, so I wanted to > be sure I understood correctly. > > Eric > From graham at thestephensdomain.com Thu Jun 11 19:34:35 2015 From: graham at thestephensdomain.com (Graham Stephens) Date: Thu, 11 Jun 2015 20:34:35 +0100 Subject: [OmniOS-discuss] port 10926 In-Reply-To: <2444C5D0-5A5F-4EA1-83FB-8888D516FA8B@cos.ru> References: <5579B09D.9090806@thestephensdomain.com> <2444C5D0-5A5F-4EA1-83FB-8888D516FA8B@cos.ru> Message-ID: <5579E2CB.2050707@thestephensdomain.com> On 11/06/2015 19:43, Jim Klimov wrote: > 11 ???? 2015 ?. 18:20:43 CEST, Eric Sproul ?????: >> On Thu, Jun 11, 2015 at 12:00 PM, Graham Stephens >> wrote: >>> Guys, >>> >>> Is there something in OmniOS that 'phones home' occasionally? >>> >>> I happened to be looking at my ISP hub and noticed I had 59 attempts >>> incoming to contact port 10926 from a Maryland U address. As I >> haven't had >>> my OmniOS box on yet today, I was a bit surprised. >> >> Nothing in OmniOS "phones home". Do you have any community package >> repos configured? There is one run by a community member at >> http://pkg.cs.umd.edu/ ... that's the only UMD connection I can think >> of. >> >> Eric >> _______________________________________________ >> OmniOS-discuss mailing list >> OmniOS-discuss at lists.omniti.com >> http://lists.omniti.com/mailman/listinfo/omnios-discuss > > Also, if something random tries to contact your addresses, it might be port-probing (e.g. google if some viral rootkits or relays for any OS are known to run on such ports - maybe some "shepherds" are trying to find their "herd") or something else that was never demanded or implicitly initiated by your boxes. Just background radiation of the Internet. Maybe... ;) > > Jim > -- > Typos courtesy of K-9 Mail on my Samsung Android > Jim, It was just a bit odd that all of a sudden I had 50+ hits when I'd had very little previously - the only thing I'd been playing with was OmniOS, so I thought I'd better check. Cheers. From graham at thestephensdomain.com Fri Jun 12 09:10:36 2015 From: graham at thestephensdomain.com (Graham Stephens) Date: Fri, 12 Jun 2015 10:10:36 +0100 Subject: [OmniOS-discuss] port 10926 In-Reply-To: <5579E2CB.2050707@thestephensdomain.com> References: <5579E2CB.2050707@thestephensdomain.com> Message-ID: <557AA20C.90208@thestephensdomain.com> >On 11/06/2015 19:43, Jim Klimov wrote: >> 11 ???? 2015 ?. 18:20:43 CEST, Eric Sproul ?????: >>> On Thu, Jun 11, 2015 at 12:00 PM, Graham Stephens >>> wrote: >>>> Guys, >>>> >>>> Is there something in OmniOS that 'phones home' occasionally? >>>> >>>> I happened to be looking at my ISP hub and noticed I had 59 attempts >>>> incoming to contact port 10926 from a Maryland U address. As I >>> haven't had >>>> my OmniOS box on yet today, I was a bit surprised. >>> >>> Nothing in OmniOS "phones home". Do you have any community package >>> repos configured? There is one run by a community member at >>> http://pkg.cs.umd.edu/ ... that's the only UMD connection I can think >>> of. >>> >>> Eric >>> _______________________________________________ >>> OmniOS-discuss mailing list >>> OmniOS-discuss at lists.omniti.com >>> http://lists.omniti.com/mailman/listinfo/omnios-discuss >> >> Also, if something random tries to contact your addresses, it might be port-probing (e.g. google if some viral rootkits or relays for any OS are known to run on such ports - maybe some "shepherds" are trying to find their "herd") or something else that was never demanded or implicitly initiated by your boxes. Just background radiation of the Internet. Maybe... ;) >> >> Jim >> -- >> Typos courtesy of K-9 Mail on my Samsung Android >> > >Jim, > >It was just a bit odd that all of a sudden I had 50+ hits when I'd had >very little previously - the only thing I'd been playing with was >OmniOS, so I thought I'd better check. > >Cheers. It looks like it might have been background noise - I checked my hub again today, it seems to have restarted the counter from this morning, and everything from yesterday has disappeared(!). My crappy hub's logging appears to be faulty, and is telling me porkies! I just didn't notice before. From tobi at oetiker.ch Fri Jun 12 15:07:55 2015 From: tobi at oetiker.ch (Tobias Oetiker) Date: Fri, 12 Jun 2015 17:07:55 +0200 (CEST) Subject: [OmniOS-discuss] 014 and X540 on intel s2600 Message-ID: We are trying to get an intel 10G network interface x540 to work on omnios r014. We do see the interface with $ dladm show-phys but it does not show any reaction when we plug an ethernet cable leading to a 1Gb switch port. It stays 'down' any ideas ? cheers tobi -- Tobi Oetiker, OETIKER+PARTNER AG, Aarweg 15 CH-4600 Olten, Switzerland www.oetiker.ch tobi at oetiker.ch +41 62 775 9902 From danmcd at omniti.com Fri Jun 12 16:33:56 2015 From: danmcd at omniti.com (Dan McDonald) Date: Fri, 12 Jun 2015 12:33:56 -0400 Subject: [OmniOS-discuss] 014 and X540 on intel s2600 In-Reply-To: References: Message-ID: > On Jun 12, 2015, at 11:07 AM, Tobias Oetiker wrote: > > We are trying to get an intel 10G network interface x540 to work > on omnios r014. We do see the interface with > > $ dladm show-phys > > but it does not show any reaction when we plug an ethernet cable > leading to a 1Gb switch port. It stays 'down' This sounds familiar, but I'd have to go digging to see why. Also, officially I'm on vacation, but because of the openssl ABI fiasco, I'm actually reading mail and fixing openssl today. Sorry I can't be of more immediate assistance, Dan From johan.kragsterman at capvert.se Fri Jun 12 16:50:35 2015 From: johan.kragsterman at capvert.se (Johan Kragsterman) Date: Fri, 12 Jun 2015 18:50:35 +0200 Subject: [OmniOS-discuss] Ang: Re: 014 and X540 on intel s2600 In-Reply-To: References: , Message-ID: Hi! -----"OmniOS-discuss" skrev: ----- Till: Tobias Oetiker Fr?n: Dan McDonald S?nt av: "OmniOS-discuss" Datum: 2015-06-12 18:35 Kopia: omnios-discuss at lists.omniti.com ?rende: Re: [OmniOS-discuss] 014 and X540 on intel s2600 > On Jun 12, 2015, at 11:07 AM, Tobias Oetiker wrote: > > We are trying to get an intel 10G network interface x540 to work > on omnios r014. We do see the interface with > > $ dladm show-phys > > but it does not show any reaction when we plug an ethernet cable > leading to a 1Gb switch port. It stays 'down' This sounds familiar, but I'd have to go digging to see why. It sounds familiar to me too. And I think I recall a discussion about the same issue, where the SFP's where involved. I believe they needed to be specific...? Rgrds Johan Also, officially I'm on vacation, but because of the openssl ABI fiasco, I'm actually reading mail and fixing openssl today. Sorry I can't be of more immediate assistance, Dan _______________________________________________ OmniOS-discuss mailing list OmniOS-discuss at lists.omniti.com http://lists.omniti.com/mailman/listinfo/omnios-discuss From danmcd at omniti.com Fri Jun 12 18:20:23 2015 From: danmcd at omniti.com (Dan McDonald) Date: Fri, 12 Jun 2015 14:20:23 -0400 Subject: [OmniOS-discuss] Update now --> OpenSSL to 1.0.2c or 1.0.1o Message-ID: For 014 and later, OpenSSL is now at 1.0.2c. For 006 and 012, OpenSSL is now at 1.0.1o. If you're running another release, STOP AND UPGRADE ALREADY! :) Thanks to the OpenSSL community for quickly fixing the ABI brokenness in 1.0.2b and 1.0.1n respectively. Thanks also to our SA team, who started to roll it out in 006 deployments and discovered the problem. You may need to "svcadm restart" services that depend upon OpenSSL. Happy updating! Dan From graham at thestephensdomain.com Sat Jun 13 12:54:03 2015 From: graham at thestephensdomain.com (Graham Stephens) Date: Sat, 13 Jun 2015 13:54:03 +0100 Subject: [OmniOS-discuss] Adding a disk array after boot Message-ID: <557C27EB.2060804@thestephensdomain.com> Perhaps another dumb question, but here goes... I currently have a FC disk array attached to a server acting, among other things, as a Samba file server. I don't need the files serving all the time, so mainly start the machine (it is normally off overnight due to the noise) without the disk array turned on unless I know I will need the files in advance. ZFS doesn't seem to mind as long as the disks are attached/not attached at boot, and I don't try turning the array on while the server is running. I am moving several boxes into one quieter zoned box that I would like to have on 24/7, and the Samba server is intended to go into one of those zones. I will eventually swap the external array over to internal disks, but it isn't going to happen straight away; so what I would like to ask is: Is there a way for me to be able to turn the disk array on and off as necessary (it will become the noisy part of the setup), without me having to reboot the main box (with all the zones, etc) every time? From stephan.budach at JVM.DE Sat Jun 13 14:01:25 2015 From: stephan.budach at JVM.DE (Stephan Budach) Date: Sat, 13 Jun 2015 16:01:25 +0200 Subject: [OmniOS-discuss] Adding a disk array after boot In-Reply-To: <557C27EB.2060804@thestephensdomain.com> References: <557C27EB.2060804@thestephensdomain.com> Message-ID: <557C37B5.8050001@jvm.de> Am 13.06.15 um 14:54 schrieb Graham Stephens: > Perhaps another dumb question, but here goes... > > I currently have a FC disk array attached to a server acting, among > other things, as a Samba file server. I don't need the files serving > all the time, so mainly start the machine (it is normally off > overnight due to the noise) without the disk array turned on unless I > know I will need the files in advance. ZFS doesn't seem to mind as > long as the disks are attached/not attached at boot, and I don't try > turning the array on while the server is running. > > I am moving several boxes into one quieter zoned box that I would like > to have on 24/7, and the Samba server is intended to go into one of > those zones. I will eventually swap the external array over to > internal disks, but it isn't going to happen straight away; so what I > would like to ask is: > Is there a way for me to be able to turn the disk array on and off as > necessary (it will become the noisy part of the setup), without me > having to reboot the main box (with all the zones, etc) every time? Each FC equipment I know of will issue a FC reset upon booting and OmniOS will pick that up. However, there is of course also a means of having the OmniOS box perform a FC reset on the bus which would also spurr a disk discovery, after which ZFS will happily import your formerly exportet(!) zpool without fuss. Cheers From graham at thestephensdomain.com Sat Jun 13 14:36:39 2015 From: graham at thestephensdomain.com (Graham Stephens) Date: Sat, 13 Jun 2015 15:36:39 +0100 Subject: [OmniOS-discuss] Adding a disk array after boot In-Reply-To: <557C37B5.8050001@jvm.de> References: <557C27EB.2060804@thestephensdomain.com> <557C37B5.8050001@jvm.de> Message-ID: <557C3FF7.5020001@thestephensdomain.com> On 13/06/2015 15:01, Stephan Budach wrote: > Am 13.06.15 um 14:54 schrieb Graham Stephens: >> Perhaps another dumb question, but here goes... >> >> I currently have a FC disk array attached to a server acting, among >> other things, as a Samba file server. I don't need the files serving >> all the time, so mainly start the machine (it is normally off >> overnight due to the noise) without the disk array turned on unless I >> know I will need the files in advance. ZFS doesn't seem to mind as >> long as the disks are attached/not attached at boot, and I don't try >> turning the array on while the server is running. >> >> I am moving several boxes into one quieter zoned box that I would like >> to have on 24/7, and the Samba server is intended to go into one of >> those zones. I will eventually swap the external array over to >> internal disks, but it isn't going to happen straight away; so what I >> would like to ask is: >> Is there a way for me to be able to turn the disk array on and off as >> necessary (it will become the noisy part of the setup), without me >> having to reboot the main box (with all the zones, etc) every time? > > Each FC equipment I know of will issue a FC reset upon booting and > OmniOS will pick that up. However, there is of course also a means of > having the OmniOS box perform a FC reset on the bus which would also > spurr a disk discovery, after which ZFS will happily import your > formerly exportet(!) zpool without fuss. > > Cheers > Ah! I hadn't thought to export the zpools first. If that's all I need to do then I'll be a very happy chap! Thanks. From peter.tribble at gmail.com Sat Jun 13 16:24:23 2015 From: peter.tribble at gmail.com (Peter Tribble) Date: Sat, 13 Jun 2015 17:24:23 +0100 Subject: [OmniOS-discuss] bug in powertop In-Reply-To: <20150610201221.2337ccf1@sleipner.datanom.net> References: <20150609174513.66050f42@sleipner.datanom.net> <20150610201221.2337ccf1@sleipner.datanom.net> Message-ID: On Wed, Jun 10, 2015 at 7:12 PM, Michael Rasmussen wrote: > On Wed, 10 Jun 2015 17:04:20 +0100 > Peter Tribble wrote: > > > > > Delete that line in powertop.c and all should be well. > > > FYI, I've looged illumos bug #6003 https://www.illumos.org/issues/6003 > Do I need to build the entire illumos-gate to recompile powertop? > No. It can often be awkward to build individual tools, but as this is a too that started out life elsewhere it's pretty easy. You don't even need to pull the whole of the gate. svn co https://github.com/illumos/illumos-gate/trunk/usr/src/cmd/powertop cd powertop/common [edit powertop.c] gcc -m64 -c *.c ../amd64/pt_amd64.c gcc -m64 -o powertop *.o -ldtrace -lkstat -lcurses > (I get the following as well: > > powertop: failed to compile P-states (frequencies) program > > powertop: failed to compile C-states (idle power) program > > so I'm not sure it's actually functional.) > > > You need to run powertop either as root or using sudo. > That's not it. Run it without adequate privileges and it will explicitly tell you so. Tracing this a little further (this isn't on an omnios box, by the way, but powertop *should* be the same) and attempting to run the dtrace it creates by hand, I get. "/usr/lib/dtrace/scsi.d", line 46: translator member ic_cdb definition uses incompatible types: "uint8_t *" = "user_desc_t" At which point I gave up, because (a) I don't understand dtrace well enough, and (b) the types used in the code look fine to me, and (c) other than the fact that it fails to compile, whether this is actually the compilation failure being reported by powertop is anyone's guess. -- -Peter Tribble http://www.petertribble.co.uk/ - http://ptribble.blogspot.com/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From daleg at omniti.com Mon Jun 15 07:27:40 2015 From: daleg at omniti.com (Dale Ghent) Date: Mon, 15 Jun 2015 03:27:40 -0400 Subject: [OmniOS-discuss] Adding a disk array after boot In-Reply-To: <557C3FF7.5020001@thestephensdomain.com> References: <557C27EB.2060804@thestephensdomain.com> <557C37B5.8050001@jvm.de> <557C3FF7.5020001@thestephensdomain.com> Message-ID: <903403E2-159F-43B4-A546-47161EED5196@omniti.com> > On Jun 13, 2015, at 10:36 AM, Graham Stephens wrote: > > On 13/06/2015 15:01, Stephan Budach wrote: >> Am 13.06.15 um 14:54 schrieb Graham Stephens: >>> Perhaps another dumb question, but here goes... >>> >>> I currently have a FC disk array attached to a server acting, among >>> other things, as a Samba file server. I don't need the files serving >>> all the time, so mainly start the machine (it is normally off >>> overnight due to the noise) without the disk array turned on unless I >>> know I will need the files in advance. ZFS doesn't seem to mind as >>> long as the disks are attached/not attached at boot, and I don't try >>> turning the array on while the server is running. >>> >>> I am moving several boxes into one quieter zoned box that I would like >>> to have on 24/7, and the Samba server is intended to go into one of >>> those zones. I will eventually swap the external array over to >>> internal disks, but it isn't going to happen straight away; so what I >>> would like to ask is: >>> Is there a way for me to be able to turn the disk array on and off as >>> necessary (it will become the noisy part of the setup), without me >>> having to reboot the main box (with all the zones, etc) every time? >> >> Each FC equipment I know of will issue a FC reset upon booting and >> OmniOS will pick that up. However, there is of course also a means of >> having the OmniOS box perform a FC reset on the bus which would also >> spurr a disk discovery, after which ZFS will happily import your >> formerly exportet(!) zpool without fuss. >> >> Cheers >> > > Ah! I hadn't thought to export the zpools first. If that's all I need to do then I'll be a very happy chap! Yes, do make sure that the zpools are first exported (which implies an unmount of its constituent filesystems) so that the disks are properly quiesced. In FC-land and on the device level, there are a few old commands you can issue to get the view of things once you unplug your array, or plug it back in: # view configured FC drives and controllers cfgadm -al # force a re-probe of a specific (FC) controller on the driver level cfgadm -c configure # clean up no-longer present disk device links under /dev and /devices devfsadm -Cv disks # forcibly re-create them if for some reason they aren?t around devfsadm -v disks # if (a) drive(s) fails to show up on a known controller, send the whole thing a LIP (Loop Init Protocol) command fcinfo force-lip (the old old OLD solaris command to do this was/is luxadm -e forecelip) /dale -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 494 bytes Desc: Message signed with OpenPGP using GPGMail URL: From javier.lopez at trixter.de Mon Jun 15 12:22:49 2015 From: javier.lopez at trixter.de (Lopez, Javier) Date: Mon, 15 Jun 2015 14:22:49 +0200 (CEST) Subject: [OmniOS-discuss] Problem updating ms-omniti local repo In-Reply-To: <881378653.44352.1434370926084.JavaMail.root@de001cs001.trixter.de> Message-ID: <1270943343.44358.1434370969909.JavaMail.root@de001cs001.trixter.de> Hello, I've created a local mirror of the ms.omniti.com repository with the following command: #> pkgrepo create /path/ms.omniti.com/ #> pkgrecv -s http://pkg.omniti.com/omniti-ms/ -d /path/ms.omniti.com/ '*' Everything were fine but when I tried to update the repo I get this error message #> pkgrecv -s http://pkg.omniti.com/omniti-ms/ -d /path/ms.omniti.com/ '*' Processing packages for publisher omniti-ms ... Retrieving catalog ' http://pkg.omniti.com/omniti-ms/ '... Unable to retrieve package data for publisher 'omniti-ms' from one of the following origin(s): http://pkg.omniti.com/omniti-ms/ The catalog retrieved from one of the origin(s) listed above only contains package data for: ms.omniti.com. This is either a result of invalid origin information being provided for publisher 'omniti-ms', or because the wrong publisher name was provided when this publisher was added. Retrieving and evaluating 809 package(s)... PROCESS ITEMS GET (MB) SEND (MB) omniti/runtime/nodejs 1/809 6.6/1805.6 0.2/5478.3 pkgrecv: 'open' failed for transaction ID '1430420933_pkg%3A%2F%2Fms.omniti.com%2Fomniti%2Fruntime%2Fnodejs%400.10.21%2C5.11-0.151014%3A20150430T190853Z': The specified FMRI, 'pkg://ms.omniti.com/omniti/runtime/nodejs at 0.10.21,5.11-0.151014:20150430T190853Z', already exists or has been restricted. pkgrecv: Cached files were preserved in the following directory: /var/tmp/pkgrecv-hLEfRQ Use pkgrecv -c to resume the interrupted download. I'm stucked here. Any help or pointers will be apreciated. Regards. Javier. -------------- next part -------------- An HTML attachment was scrubbed... URL: From vab at bb-c.de Mon Jun 15 12:58:31 2015 From: vab at bb-c.de (Volker A. Brandt) Date: Mon, 15 Jun 2015 14:58:31 +0200 Subject: [OmniOS-discuss] Problem updating ms-omniti local repo In-Reply-To: <1270943343.44358.1434370969909.JavaMail.root@de001cs001.trixter.de> References: <881378653.44352.1434370926084.JavaMail.root@de001cs001.trixter.de> <1270943343.44358.1434370969909.JavaMail.root@de001cs001.trixter.de> Message-ID: <21886.52215.782384.393147@glaurung.bb-c.de> Hi Javier! > #> pkgrepo create /path/ms.omniti.com/ > #> pkgrecv -s http://pkg.omniti.com/omniti-ms/ -d /path/ms.omniti.com/ '*' This is not enough. You need to explicitly set the publisher for your new local repository. The correct publisher name is "ms.omniti.com". So, to create a local repository, do: # pkgrepo create /path/ms.omniti.com/ # pkgrepo add-publisher -s /path/ms.omniti.com/ ms.omniti.com # pkgrecv -s http://pkg.omniti.com/omniti-ms/ -d /path/ms.omniti.com/ '*' Regards -- Volker -- ------------------------------------------------------------------------ Volker A. Brandt Consulting and Support for Oracle Solaris Brandt & Brandt Computer GmbH WWW: http://www.bb-c.de/ Am Wiesenpfad 6, 53340 Meckenheim, GERMANY Email: vab at bb-c.de Handelsregister: Amtsgericht Bonn, HRB 10513 Schuhgr??e: 46 Gesch?ftsf?hrer: Rainer J.H. Brandt und Volker A. Brandt "When logic and proportion have fallen sloppy dead" From tobi at oetiker.ch Tue Jun 16 07:44:28 2015 From: tobi at oetiker.ch (Tobias Oetiker) Date: Tue, 16 Jun 2015 09:44:28 +0200 (CEST) Subject: [OmniOS-discuss] 014 and X540 on intel s2600 In-Reply-To: References: Message-ID: Friday Tobias Oetiker wrote: > We are trying to get an intel 10G network interface x540 to work > on omnios r014. We do see the interface with > > $ dladm show-phys > > but it does not show any reaction when we plug an ethernet cable > leading to a 1Gb switch port. It stays 'down' to resolve this ... it turnes out the customer had neglected to plug the network cable in ... and had steadfastly claimed that everything was plugged propperly when we called him. Only today when we called again and another member of staff went to check, he found that no cable was connected to the port ... To to recap ... x540 10 Gigabit Copper Ethernet ports work on omnios r014 out of the box. IF the cable is plugged in. cheers tobi -- Tobi Oetiker, OETIKER+PARTNER AG, Aarweg 15 CH-4600 Olten, Switzerland www.oetiker.ch tobi at oetiker.ch +41 62 775 9902 From danmcd at omniti.com Tue Jun 16 12:35:35 2015 From: danmcd at omniti.com (Dan McDonald) Date: Tue, 16 Jun 2015 08:35:35 -0400 Subject: [OmniOS-discuss] 014 and X540 on intel s2600 In-Reply-To: References: Message-ID: <274FFAE9-674A-4DC8-9F59-D51FF6AB1638@omniti.com> > On Jun 16, 2015, at 3:44 AM, Tobias Oetiker wrote: > > > To to recap ... x540 10 Gigabit Copper Ethernet ports work on > omnios r014 out of the box. IF the cable is plugged in. Phew. I'm glad this is one less thing for me to deal with after a long weekend trip. Thanks! Dan From danmcd at omniti.com Tue Jun 16 13:14:57 2015 From: danmcd at omniti.com (Dan McDonald) Date: Tue, 16 Jun 2015 09:14:57 -0400 Subject: [OmniOS-discuss] Hello bloody folks! Check out preliminary gcc5.1 support In-Reply-To: References: <5FF7750C-3CFC-434C-A563-2AD1A882096A@omniti.com> Message-ID: <5108F2F9-2901-4639-86E3-9FB574132007@omniti.com> > On Jun 10, 2015, at 10:00 AM, Eric Sproul wrote: > > On Wed, Jun 10, 2015 at 8:46 AM, Dan McDonald wrote: >> >>> On Jun 8, 2015, at 10:46 AM, Eric Sproul wrote: >>> >>> On Fri, Jun 5, 2015 at 3:46 PM, Dan McDonald wrote: >>>> Any code-minded people in the community should have a look at the above webrev, and tell me what you think. Unlike the last compiler change done for r151008, this is a jump from 4 to 5, so "g{cc,++}-N-runtime" has a new value of N, namely 5. The prior versions of gcc runtime are in place (.so.6.0.XX), but the package name has been changed. I've use the "renamed" attribute for gcc/++-4-runtime to aid in this regard. >>> >>> Dan, >>> Great work so far. Mostly looks good to me, though I'm curious if >>> pkglint is/will be complaining about redundant require dependencies. >> >> Remember -- we don't run pkglint by default. (Some existing package naming schemes make that difficult/impossible.) > > Ah, I just looked at the make_package() function. Out of (perhaps > morbid) curiosity, what are the obstacles? -bash-4.3$ pkg list nfs NAME (PUBLISHER) VERSION IFO service/file-system/nfs 0.5.11-0.151015 i-- system/file-system/nfs 0.5.11-0.151015 i-- -bash-4.3$ Though those are in ON, to be fair. >> The whole DEPENDS_IPS/BUILD_DEPENDS_IPS/PKG_DEPENDS_IPS is screwy to me. A bit of history, and a suggested what's-right course of action may be in order. > > OK, so to go _really_ far back into the Dark Ages, our build system > was originally created to make SVR4 packages for Solaris 10 in > OmniTI's hosting environments. No fancy auto-dependency tools there, > so the build had to declare all runtime dependencies manually. That > persisted through the transition to SXCE, which was still SVR4. > Fast-forward to IPS and OmniOS, which grew organically from the SVR4 > builds, and the habit of specifying all deps manually was already > deeply ingrained. Inherited S10/SVR4 stuff -- got it! > So now, in the all-IPS world, with pkgdepend, we generally don't need > to specify run-time dependencies. pkgdepend can handle ELF, Java, and > some scripts (Python, Perl, shell), and maybe a few other minor > things. It can't do Perl modules, that's one glaring issue that comes > to mind, so there are still cases where you need to specify *some* > runtime deps. BUT, since we don't run pkglint, we haven't been good > about cleaning up the unnecessary DEPENDS_IPS items, and our packages > will end up with either duplicate deps or unnecessary ones, e.g. > > $ pkg contents -t depend -o type,fmri bash > TYPE FMRI > require pkg:/SUNWcs at 0.5.11-0.151014 > require pkg:/system/library at 0.5.11-0.151014 > require system/library/gcc-4-runtime > > The first two came from pkgdepend, but bash doesn't actually need > gcc-4-runtime (verified with ldd). It may have at one time, or > someone was just over-zealous in applying an assumed dependency > without validation. > > This is mostly harmless, but as in your case here, it's creating > unnecessary work for you to change to gcc5. Since I've *done* the work already, however, are you okay with the existing changes. I've been running a gcc5'ed bloody on&off for the past 1-2 weeks. I also have a gcc5 install ISO which I need to confirm, but I think I'm good. If you're good, I'll mark you as a reviewer, integrate, rebuild the world, and see if a conventional "pkg update" works on my pristine bloody system. Thanks, Dan From eric.sproul at circonus.com Tue Jun 16 13:31:35 2015 From: eric.sproul at circonus.com (Eric Sproul) Date: Tue, 16 Jun 2015 09:31:35 -0400 Subject: [OmniOS-discuss] Hello bloody folks! Check out preliminary gcc5.1 support In-Reply-To: <5108F2F9-2901-4639-86E3-9FB574132007@omniti.com> References: <5FF7750C-3CFC-434C-A563-2AD1A882096A@omniti.com> <5108F2F9-2901-4639-86E3-9FB574132007@omniti.com> Message-ID: On Tue, Jun 16, 2015 at 9:14 AM, Dan McDonald wrote: >> Ah, I just looked at the make_package() function. Out of (perhaps >> morbid) curiosity, what are the obstacles? > > -bash-4.3$ pkg list nfs > NAME (PUBLISHER) VERSION IFO > service/file-system/nfs 0.5.11-0.151015 i-- > system/file-system/nfs 0.5.11-0.151015 i-- > -bash-4.3$ > > Though those are in ON, to be fair. Yeah, illumos-gate packages are kind of their own thing-- if manifests there are wonky, they should be fixed upstream. pkglint might be of help there, nonetheless. IM(V)HO we should be pkglinting all the omnios-build bits though. In saying that, I realize I should be doing the same for the repo *I* manage. ;) /me goes to made a todo item... >> This is mostly harmless, but as in your case here, it's creating >> unnecessary work for you to change to gcc5. > > Since I've *done* the work already, however, are you okay with the existing changes. I've been running a gcc5'ed bloody on&off for the past 1-2 weeks. I also have a gcc5 install ISO which I need to confirm, but I think I'm good. If you're good, I'll mark you as a reviewer, integrate, rebuild the world, and see if a conventional "pkg update" works on my pristine bloody system. Yeah, let's possibly take it as a followup item to fix unnecessary dependencies. +1 from me. Eric From javier.lopez at trixter.de Tue Jun 16 15:48:35 2015 From: javier.lopez at trixter.de (Lopez, Javier) Date: Tue, 16 Jun 2015 17:48:35 +0200 (CEST) Subject: [OmniOS-discuss] Problem updating ms-omniti local repo In-Reply-To: <1270943343.44358.1434370969909.JavaMail.root@de001cs001.trixter.de> Message-ID: <452674050.46461.1434469715764.JavaMail.root@de001cs001.trixter.de> T hank you for the answer Volker , I've recreated the repo following your indications and I still have the same problem when I try to update the repo. thank you for your time regards ----- Original Message ----- From: "Javier Lopez" To: omnios-discuss at lists.omniti.com Cc: "Trixter IT Dept" Sent: Monday, June 15, 2015 2:22:49 PM Subject: Problem updating ms-omniti local repo Hello, I've created a local mirror of the ms.omniti.com repository with the following command: #> pkgrepo create /path/ms.omniti.com/ #> pkgrecv -s http://pkg.omniti.com/omniti-ms/ -d /path/ms.omniti.com/ '*' Everything were fine but when I tried to update the repo I get this error message #> pkgrecv -s http://pkg.omniti.com/omniti-ms/ -d /path/ms.omniti.com/ '*' Processing packages for publisher omniti-ms ... Retrieving catalog ' http://pkg.omniti.com/omniti-ms/ '... Unable to retrieve package data for publisher 'omniti-ms' from one of the following origin(s): http://pkg.omniti.com/omniti-ms/ The catalog retrieved from one of the origin(s) listed above only contains package data for: ms.omniti.com. This is either a result of invalid origin information being provided for publisher 'omniti-ms', or because the wrong publisher name was provided when this publisher was added. Retrieving and evaluating 809 package(s)... PROCESS ITEMS GET (MB) SEND (MB) omniti/runtime/nodejs 1/809 6.6/1805.6 0.2/5478.3 pkgrecv: 'open' failed for transaction ID '1430420933_pkg%3A%2F%2Fms.omniti.com%2Fomniti%2Fruntime%2Fnodejs%400.10.21%2C5.11-0.151014%3A20150430T190853Z': The specified FMRI, 'pkg://ms.omniti.com/omniti/runtime/nodejs at 0.10.21,5.11-0.151014:20150430T190853Z', already exists or has been restricted. pkgrecv: Cached files were preserved in the following directory: /var/tmp/pkgrecv-hLEfRQ Use pkgrecv -c to resume the interrupted download. I'm stucked here. Any help or pointers will be apreciated. Regards. Javier. -------------- next part -------------- An HTML attachment was scrubbed... URL: From jimklimov at cos.ru Tue Jun 16 17:24:12 2015 From: jimklimov at cos.ru (Jim Klimov) Date: Tue, 16 Jun 2015 19:24:12 +0200 Subject: [OmniOS-discuss] Adding a disk array after boot In-Reply-To: <903403E2-159F-43B4-A546-47161EED5196@omniti.com> References: <557C27EB.2060804@thestephensdomain.com> <557C37B5.8050001@jvm.de> <557C3FF7.5020001@thestephensdomain.com> <903403E2-159F-43B4-A546-47161EED5196@omniti.com> Message-ID: 15 ???? 2015??. 9:27:40 CEST, Dale Ghent ?????: > >> On Jun 13, 2015, at 10:36 AM, Graham Stephens > wrote: >> >> On 13/06/2015 15:01, Stephan Budach wrote: >>> Am 13.06.15 um 14:54 schrieb Graham Stephens: >>>> Perhaps another dumb question, but here goes... >>>> >>>> I currently have a FC disk array attached to a server acting, among >>>> other things, as a Samba file server. I don't need the files >serving >>>> all the time, so mainly start the machine (it is normally off >>>> overnight due to the noise) without the disk array turned on unless >I >>>> know I will need the files in advance. ZFS doesn't seem to mind as >>>> long as the disks are attached/not attached at boot, and I don't >try >>>> turning the array on while the server is running. >>>> >>>> I am moving several boxes into one quieter zoned box that I would >like >>>> to have on 24/7, and the Samba server is intended to go into one of >>>> those zones. I will eventually swap the external array over to >>>> internal disks, but it isn't going to happen straight away; so what >I >>>> would like to ask is: >>>> Is there a way for me to be able to turn the disk array on and off >as >>>> necessary (it will become the noisy part of the setup), without me >>>> having to reboot the main box (with all the zones, etc) every time? >>> >>> Each FC equipment I know of will issue a FC reset upon booting and >>> OmniOS will pick that up. However, there is of course also a means >of >>> having the OmniOS box perform a FC reset on the bus which would also >>> spurr a disk discovery, after which ZFS will happily import your >>> formerly exportet(!) zpool without fuss. >>> >>> Cheers >>> >> >> Ah! I hadn't thought to export the zpools first. If that's all I need >to do then I'll be a very happy chap! > >Yes, do make sure that the zpools are first exported (which implies an >unmount of its constituent filesystems) so that the disks are properly >quiesced. > >In FC-land and on the device level, there are a few old commands you >can issue to get the view of things once you unplug your array, or plug >it back in: > ># view configured FC drives and controllers >cfgadm -al > ># force a re-probe of a specific (FC) controller on the driver level >cfgadm -c configure > ># clean up no-longer present disk device links under /dev and /devices >devfsadm -Cv disks > ># forcibly re-create them if for some reason they aren?t around >devfsadm -v disks > ># if (a) drive(s) fails to show up on a known controller, send the >whole thing a LIP (Loop Init Protocol) command >fcinfo force-lip (the old old OLD solaris >command to do this was/is luxadm -e forecelip) > >/dale > > >------------------------------------------------------------------------ > >_______________________________________________ >OmniOS-discuss mailing list >OmniOS-discuss at lists.omniti.com >http://lists.omniti.com/mailman/listinfo/omnios-discuss You might also want to SMFize your pool instances and stuff that depends on the pool (zones, samba, etc.) along the lines of https://github.com/jimklimov/illumos-smf-zfspools and https://github.com/jimklimov/illumos-smf-zones - so you can just 'svcadm disable -ts myfcarray' and when that's done - unplug it. Grab an ePDU and power it off/onn programmatically for bonus nerd points ;) HTH, Jim Klimov -- Typos courtesy of K-9 Mail on my Samsung Android From danmcd at omniti.com Wed Jun 17 02:16:37 2015 From: danmcd at omniti.com (Dan McDonald) Date: Tue, 16 Jun 2015 22:16:37 -0400 Subject: [OmniOS-discuss] FLAG DAY --> do not build omnios-build on bloody until the binaries are updated Message-ID: <2D03B99D-BEE9-4A51-9F58-8AD971990C07@omniti.com> I've pushed back the changes for gcc51 into omnios-build. I now need to build it a final time, and test upgrading. If you pull changes into omnios-build, you will be able to build most of the world, but not the gcc51 portions of it. Watch this list for a followup announcement when the bloody binaries on IPS are updated. Thank you for your patience, Dan From vab at bb-c.de Wed Jun 17 07:06:34 2015 From: vab at bb-c.de (Volker A. Brandt) Date: Wed, 17 Jun 2015 09:06:34 +0200 Subject: [OmniOS-discuss] Problem updating ms-omniti local repo In-Reply-To: <452674050.46461.1434469715764.JavaMail.root@de001cs001.trixter.de> References: <1270943343.44358.1434370969909.JavaMail.root@de001cs001.trixter.de> <452674050.46461.1434469715764.JavaMail.root@de001cs001.trixter.de> Message-ID: <21889.7290.282238.425163@glaurung.bb-c.de> Hi Javier! > T hank you for the answer Volker You are welcome. > I've recreated the repo following your indications and I still have > the same problem when I try to update the repo. Don't worry, you will get there. Here is exactly what I did (I wanted a local copy anyway :-): | # zfs create -o mountpoint=/pkg/ms.omniti.com dpool/pkg/ms.omniti.com | # zfs set atime=off dpool/pkg/ms.omniti.com | # zfs set setuid=off dpool/pkg/ms.omniti.com | # zfs set devices=off dpool/pkg/ms.omniti.com | # zfs set exec=off dpool/pkg/ms.omniti.com | # pkgrepo create /pkg/ms.omniti.com | # pkgrepo add-publisher -s/pkg/ms.omniti.com ms.omniti.com Now you have a choice. You can clone the repository (= make an exact copy): | # pkgrecv -p ms.omniti.com -s http://pkg.omniti.com/omniti-ms/ -d /pkg/ms.omniti.com/ --clone | Processing packages for publisher ms.omniti.com ... | Retrieving and evaluating 1473 package(s)... | DOWNLOAD PKGS FILES XFER (MB) SPEED | .../postgresql-936/pg_buffercache 1473/1473 617732/617732 6552/6552 718k/s | | Verifying repository contents. | Initiating repository verification. | pkg://ms.omniti.com/omniti/monitoring/nad 8/1474 -Traceback (most recent call last): | File "/usr/lib/python2.6/vendor-packages/pkg/server/repository.py", line 2391, in verify | trust_anchors, sig_required_names, use_crls): | File "/usr/lib/python2.6/vendor-packages/pkg/server/repository.py", line 2288, in __gen_verify | hashes, errors = self.__get_hashes(path, pfmri) | File "/usr/lib/python2.6/vendor-packages/pkg/server/repository.py", line 2005, in __get_hashes | fnames = fname.split() | AttributeError: 'NoneType' object has no attribute 'split' | Traceback (most recent call last): | File "/usr/bin/pkgrepo", line 1689, in handle_errors | __ret = func(*args, **kwargs) | File "/usr/bin/pkgrepo", line 1665, in main_func | return func(conf, pargs) | File "/usr/bin/pkgrepo", line 1512, in subcmd_verify | for verify_tuple in repo.verify(pub=xpub, progtrack=progtrack): | File "/usr/lib/python2.6/vendor-packages/pkg/server/repository.py", line 2396, in verify | raise apx._convert_error(e) | AttributeError: 'NoneType' object has no attribute 'split' | | | pkgrepo: This is an internal error in pkg(5) version 1427212657. Please log a | Service Request about this issue including the information above and this | message. | | pkgrecv: Pkgrepo verify found errors in the updated repository. | The original package catalog has been restored. | The clone operation can be retried; package content that has already been retrieved will not be downloaded again. This is a crash in "pkgrepo verify", which will run automatically after the actual copy. But the repository is there. Or you can retrieve just the latest packages: | # pkgrecv -m latest -s http://pkg.omniti.com/omniti-ms/ -d /pkg/ms.omniti.com/ '*' | Processing packages for publisher omniti-ms ... | Retrieving catalog 1/1 omniti-ms 114.48 kB | Unable to retrieve package data for publisher 'omniti-ms' from one | of the following origin(s): | | http://pkg.omniti.com/omniti-ms/ | | The catalog retrieved from one of the origin(s) listed above only | contains package data for: ms.omniti.com. | | This is either a result of invalid origin information being provided | for publisher 'omniti-ms', or because the wrong publisher name was | provided when this publisher was added. | | Retrieving and evaluating 809 package(s)... | Download Manifests ( 36/809) - Again, there is an error message (because either the FMRI with "omniti-ms" confuses pkgrecv, or there really is an empty publisher entry for "omniti-ms" in the original repository), but the data will be there. Regards -- Volker -- ------------------------------------------------------------------------ Volker A. Brandt Consulting and Support for Oracle Solaris Brandt & Brandt Computer GmbH WWW: http://www.bb-c.de/ Am Wiesenpfad 6, 53340 Meckenheim, GERMANY Email: vab at bb-c.de Handelsregister: Amtsgericht Bonn, HRB 10513 Schuhgr??e: 46 Gesch?ftsf?hrer: Rainer J.H. Brandt und Volker A. Brandt "When logic and proportion have fallen sloppy dead" From javier.lopez at trixter.de Wed Jun 17 12:59:18 2015 From: javier.lopez at trixter.de (Lopez, Javier) Date: Wed, 17 Jun 2015 14:59:18 +0200 (CEST) Subject: [OmniOS-discuss] Problem updating ms-omniti local repo In-Reply-To: <21889.7290.282238.425163@glaurung.bb-c.de> Message-ID: <1983432428.47247.1434545958611.JavaMail.root@de001cs001.trixter.de> Hi Volker, Thank you very much for your help. I think we had a small misunderstanding though: I don't have a problem with creating the repo, I in fact having it working now. I used your advice from your second email: # pkgrepo create /path/ms.omniti.com/ # pkgrepo add-publisher -s /path/ms.omniti.com/ ms.omniti.com # pkgrecv -s http://pkg.omniti.com/omniti-ms/ -d /path/ms.omniti.com/ '*' However, from the beginning, I've had a problem with updating it. I continue to receive this error message: #> pkgrecv -s http://pkg.omniti.com/omniti-ms/ -d /path/ms.omniti.com/ '*' Processing packages for publisher omniti-ms ... Retrieving catalog ' http://pkg.omniti.com/omniti-ms/ '... Unable to retrieve package data for publisher 'omniti-ms' from one of the following origin(s): http://pkg.omniti.com/omniti-ms/ The catalog retrieved from one of the origin(s) listed above only contains package data for: ms.omniti.com. This is either a result of invalid origin information being provided for publisher 'omniti-ms', or because the wrong publisher name was provided when this publisher was added. I noticed in your last email you used a different method of cloning the repo: | # pkgrecv -p ms.omniti.com -s http://pkg.omniti.com/omniti-ms/ -d /pkg/ms.omniti.com/ --clone We ran into a problem with this command-- "pkgrecv: Illegal option --p". Any idea what is going on here? We are running Your pkgrecv command as written specifies the publisher (I guess with the -p option): is this perhaps what is giving us a problem with updating? If that is the case, is there a way to fix what we already without having to download the entire repo again? (Just to give a bit of background, the reason why we asked this question in the first place is because we want to have our repo management server update our omnios repos on a cron job) Regards Javier Lopez System Administrator phone: mobile: fax: mail: web: +49 89 287018 0 +49 89 287018 18 javier.lopez at trixter.de http://www.trixter.de TRIXTER Film GmbH Amalienstrasse 67 80799 Munich, Germany TRIXTER FILM GMBH, Sitz Muenchen, Registergericht Muenchen, HRB 125141, Geschaeftsfuehrer: Simone Kraus, Michael Coldewey, Christian Sommer TRIXTER FILM GMBH, Place of business Munich, Register Court Munich, HRB 125141, Managing Directors: Simone Kraus, Michael Coldewey, Christian Sommer ----- Original Message ----- From: "Volker A. Brandt" To: "Javier Lopez" Cc: omnios-discuss at lists.omniti.com, "Trixter IT Dept" Sent: Wednesday, June 17, 2015 9:06:34 AM Subject: Re: [OmniOS-discuss] Problem updating ms-omniti local repo Hi Javier! > T hank you for the answer Volker You are welcome. > I've recreated the repo following your indications and I still have > the same problem when I try to update the repo. Don't worry, you will get there. Here is exactly what I did (I wanted a local copy anyway :-): | # zfs create -o mountpoint=/pkg/ms.omniti.com dpool/pkg/ms.omniti.com | # zfs set atime=off dpool/pkg/ms.omniti.com | # zfs set setuid=off dpool/pkg/ms.omniti.com | # zfs set devices=off dpool/pkg/ms.omniti.com | # zfs set exec=off dpool/pkg/ms.omniti.com | # pkgrepo create /pkg/ms.omniti.com | # pkgrepo add-publisher -s/pkg/ms.omniti.com ms.omniti.com Now you have a choice. You can clone the repository (= make an exact copy): | # pkgrecv -p ms.omniti.com -s http://pkg.omniti.com/omniti-ms/ -d /pkg/ms.omniti.com/ --clone | Processing packages for publisher ms.omniti.com ... | Retrieving and evaluating 1473 package(s)... | DOWNLOAD PKGS FILES XFER (MB) SPEED | .../postgresql-936/pg_buffercache 1473/1473 617732/617732 6552/6552 718k/s | | Verifying repository contents. | Initiating repository verification. | pkg://ms.omniti.com/omniti/monitoring/nad 8/1474 -Traceback (most recent call last): | File "/usr/lib/python2.6/vendor-packages/pkg/server/repository.py", line 2391, in verify | trust_anchors, sig_required_names, use_crls): | File "/usr/lib/python2.6/vendor-packages/pkg/server/repository.py", line 2288, in __gen_verify | hashes, errors = self.__get_hashes(path, pfmri) | File "/usr/lib/python2.6/vendor-packages/pkg/server/repository.py", line 2005, in __get_hashes | fnames = fname.split() | AttributeError: 'NoneType' object has no attribute 'split' | Traceback (most recent call last): | File "/usr/bin/pkgrepo", line 1689, in handle_errors | __ret = func(*args, **kwargs) | File "/usr/bin/pkgrepo", line 1665, in main_func | return func(conf, pargs) | File "/usr/bin/pkgrepo", line 1512, in subcmd_verify | for verify_tuple in repo.verify(pub=xpub, progtrack=progtrack): | File "/usr/lib/python2.6/vendor-packages/pkg/server/repository.py", line 2396, in verify | raise apx._convert_error(e) | AttributeError: 'NoneType' object has no attribute 'split' | | | pkgrepo: This is an internal error in pkg(5) version 1427212657. Please log a | Service Request about this issue including the information above and this | message. | | pkgrecv: Pkgrepo verify found errors in the updated repository. | The original package catalog has been restored. | The clone operation can be retried; package content that has already been retrieved will not be downloaded again. This is a crash in "pkgrepo verify", which will run automatically after the actual copy. But the repository is there. Or you can retrieve just the latest packages: | # pkgrecv -m latest -s http://pkg.omniti.com/omniti-ms/ -d /pkg/ms.omniti.com/ '*' | Processing packages for publisher omniti-ms ... | Retrieving catalog 1/1 omniti-ms 114.48 kB | Unable to retrieve package data for publisher 'omniti-ms' from one | of the following origin(s): | | http://pkg.omniti.com/omniti-ms/ | | The catalog retrieved from one of the origin(s) listed above only | contains package data for: ms.omniti.com. | | This is either a result of invalid origin information being provided | for publisher 'omniti-ms', or because the wrong publisher name was | provided when this publisher was added. | | Retrieving and evaluating 809 package(s)... | Download Manifests ( 36/809) - Again, there is an error message (because either the FMRI with "omniti-ms" confuses pkgrecv, or there really is an empty publisher entry for "omniti-ms" in the original repository), but the data will be there. Regards -- Volker -- ------------------------------------------------------------------------ Volker A. Brandt Consulting and Support for Oracle Solaris Brandt & Brandt Computer GmbH WWW: http://www.bb-c.de/ Am Wiesenpfad 6, 53340 Meckenheim, GERMANY Email: vab at bb-c.de Handelsregister: Amtsgericht Bonn, HRB 10513 Schuhgr??e: 46 Gesch?ftsf?hrer: Rainer J.H. Brandt und Volker A. Brandt "When logic and proportion have fallen sloppy dead" -------------- next part -------------- An HTML attachment was scrubbed... URL: From javier.lopez at trixter.de Wed Jun 17 13:02:49 2015 From: javier.lopez at trixter.de (Lopez, Javier) Date: Wed, 17 Jun 2015 15:02:49 +0200 (CEST) Subject: [OmniOS-discuss] Problem updating ms-omniti local repo In-Reply-To: <1983432428.47247.1434545958611.JavaMail.root@de001cs001.trixter.de> Message-ID: <1560707420.47257.1434546169153.JavaMail.root@de001cs001.trixter.de> Sorry I forgot to specify we are running 151014 Javier Lopez System Administrator phone: mobile: fax: mail: web: +49 89 287018 0 +49 89 287018 18 javier.lopez at trixter.de http://www.trixter.de TRIXTER Film GmbH Amalienstrasse 67 80799 Munich, Germany TRIXTER FILM GMBH, Sitz Muenchen, Registergericht Muenchen, HRB 125141, Geschaeftsfuehrer: Simone Kraus, Michael Coldewey, Christian Sommer TRIXTER FILM GMBH, Place of business Munich, Register Court Munich, HRB 125141, Managing Directors: Simone Kraus, Michael Coldewey, Christian Sommer ----- Original Message ----- From: "Javier Lopez" To: "Volker A. Brandt" Cc: omnios-discuss at lists.omniti.com, "Trixter IT Dept" Sent: Wednesday, June 17, 2015 2:59:18 PM Subject: Re: [OmniOS-discuss] Problem updating ms-omniti local repo Hi Volker, Thank you very much for your help. I think we had a small misunderstanding though: I don't have a problem with creating the repo, I in fact having it working now. I used your advice from your second email: # pkgrepo create /path/ms.omniti.com/ # pkgrepo add-publisher -s /path/ms.omniti.com/ ms.omniti.com # pkgrecv -s http://pkg.omniti.com/omniti-ms/ -d /path/ms.omniti.com/ '*' However, from the beginning, I've had a problem with updating it. I continue to receive this error message: #> pkgrecv -s http://pkg.omniti.com/omniti-ms/ -d /path/ms.omniti.com/ '*' Processing packages for publisher omniti-ms ... Retrieving catalog ' http://pkg.omniti.com/omniti-ms/ '... Unable to retrieve package data for publisher 'omniti-ms' from one of the following origin(s): http://pkg.omniti.com/omniti-ms/ The catalog retrieved from one of the origin(s) listed above only contains package data for: ms.omniti.com. This is either a result of invalid origin information being provided for publisher 'omniti-ms', or because the wrong publisher name was provided when this publisher was added. I noticed in your last email you used a different method of cloning the repo: | # pkgrecv -p ms.omniti.com -s http://pkg.omniti.com/omniti-ms/ -d /pkg/ms.omniti.com/ --clone We ran into a problem with this command-- "pkgrecv: Illegal option --p". Any idea what is going on here? We are running Your pkgrecv command as written specifies the publisher (I guess with the -p option): is this perhaps what is giving us a problem with updating? If that is the case, is there a way to fix what we already without having to download the entire repo again? (Just to give a bit of background, the reason why we asked this question in the first place is because we want to have our repo management server update our omnios repos on a cron job) Regards Javier Lopez System Administrator phone: mobile: fax: mail: web: +49 89 287018 0 +49 89 287018 18 javier.lopez at trixter.de http://www.trixter.de TRIXTER Film GmbH Amalienstrasse 67 80799 Munich, Germany TRIXTER FILM GMBH, Sitz Muenchen, Registergericht Muenchen, HRB 125141, Geschaeftsfuehrer: Simone Kraus, Michael Coldewey, Christian Sommer TRIXTER FILM GMBH, Place of business Munich, Register Court Munich, HRB 125141, Managing Directors: Simone Kraus, Michael Coldewey, Christian Sommer ----- Original Message ----- From: "Volker A. Brandt" To: "Javier Lopez" Cc: omnios-discuss at lists.omniti.com, "Trixter IT Dept" Sent: Wednesday, June 17, 2015 9:06:34 AM Subject: Re: [OmniOS-discuss] Problem updating ms-omniti local repo Hi Javier! > T hank you for the answer Volker You are welcome. > I've recreated the repo following your indications and I still have > the same problem when I try to update the repo. Don't worry, you will get there. Here is exactly what I did (I wanted a local copy anyway :-): | # zfs create -o mountpoint=/pkg/ms.omniti.com dpool/pkg/ms.omniti.com | # zfs set atime=off dpool/pkg/ms.omniti.com | # zfs set setuid=off dpool/pkg/ms.omniti.com | # zfs set devices=off dpool/pkg/ms.omniti.com | # zfs set exec=off dpool/pkg/ms.omniti.com | # pkgrepo create /pkg/ms.omniti.com | # pkgrepo add-publisher -s/pkg/ms.omniti.com ms.omniti.com Now you have a choice. You can clone the repository (= make an exact copy): | # pkgrecv -p ms.omniti.com -s http://pkg.omniti.com/omniti-ms/ -d /pkg/ms.omniti.com/ --clone | Processing packages for publisher ms.omniti.com ... | Retrieving and evaluating 1473 package(s)... | DOWNLOAD PKGS FILES XFER (MB) SPEED | .../postgresql-936/pg_buffercache 1473/1473 617732/617732 6552/6552 718k/s | | Verifying repository contents. | Initiating repository verification. | pkg://ms.omniti.com/omniti/monitoring/nad 8/1474 -Traceback (most recent call last): | File "/usr/lib/python2.6/vendor-packages/pkg/server/repository.py", line 2391, in verify | trust_anchors, sig_required_names, use_crls): | File "/usr/lib/python2.6/vendor-packages/pkg/server/repository.py", line 2288, in __gen_verify | hashes, errors = self.__get_hashes(path, pfmri) | File "/usr/lib/python2.6/vendor-packages/pkg/server/repository.py", line 2005, in __get_hashes | fnames = fname.split() | AttributeError: 'NoneType' object has no attribute 'split' | Traceback (most recent call last): | File "/usr/bin/pkgrepo", line 1689, in handle_errors | __ret = func(*args, **kwargs) | File "/usr/bin/pkgrepo", line 1665, in main_func | return func(conf, pargs) | File "/usr/bin/pkgrepo", line 1512, in subcmd_verify | for verify_tuple in repo.verify(pub=xpub, progtrack=progtrack): | File "/usr/lib/python2.6/vendor-packages/pkg/server/repository.py", line 2396, in verify | raise apx._convert_error(e) | AttributeError: 'NoneType' object has no attribute 'split' | | | pkgrepo: This is an internal error in pkg(5) version 1427212657. Please log a | Service Request about this issue including the information above and this | message. | | pkgrecv: Pkgrepo verify found errors in the updated repository. | The original package catalog has been restored. | The clone operation can be retried; package content that has already been retrieved will not be downloaded again. This is a crash in "pkgrepo verify", which will run automatically after the actual copy. But the repository is there. Or you can retrieve just the latest packages: | # pkgrecv -m latest -s http://pkg.omniti.com/omniti-ms/ -d /pkg/ms.omniti.com/ '*' | Processing packages for publisher omniti-ms ... | Retrieving catalog 1/1 omniti-ms 114.48 kB | Unable to retrieve package data for publisher 'omniti-ms' from one | of the following origin(s): | | http://pkg.omniti.com/omniti-ms/ | | The catalog retrieved from one of the origin(s) listed above only | contains package data for: ms.omniti.com. | | This is either a result of invalid origin information being provided | for publisher 'omniti-ms', or because the wrong publisher name was | provided when this publisher was added. | | Retrieving and evaluating 809 package(s)... | Download Manifests ( 36/809) - Again, there is an error message (because either the FMRI with "omniti-ms" confuses pkgrecv, or there really is an empty publisher entry for "omniti-ms" in the original repository), but the data will be there. Regards -- Volker -- ------------------------------------------------------------------------ Volker A. Brandt Consulting and Support for Oracle Solaris Brandt & Brandt Computer GmbH WWW: http://www.bb-c.de/ Am Wiesenpfad 6, 53340 Meckenheim, GERMANY Email: vab at bb-c.de Handelsregister: Amtsgericht Bonn, HRB 10513 Schuhgr??e: 46 Gesch?ftsf?hrer: Rainer J.H. Brandt und Volker A. Brandt "When logic and proportion have fallen sloppy dead" -------------- next part -------------- An HTML attachment was scrubbed... URL: From javier.lopez at trixter.de Wed Jun 17 13:07:16 2015 From: javier.lopez at trixter.de (Lopez, Javier) Date: Wed, 17 Jun 2015 15:07:16 +0200 (CEST) Subject: [OmniOS-discuss] Problem updating ms-omniti local repo In-Reply-To: <1560707420.47257.1434546169153.JavaMail.root@de001cs001.trixter.de> Message-ID: <1193252866.47263.1434546436120.JavaMail.root@de001cs001.trixter.de> I also noticed that you referred to this version of pkg: 1427212657 however I am running # pkg version 1382373239 which came with our install of OmniOS (which I believe to be the latest version). Are you perhaps using the Bloody? ----- Original Message ----- From: "Javier Lopez" To: "Volker A. Brandt" Cc: omnios-discuss at lists.omniti.com, "Trixter IT Dept" Sent: Wednesday, June 17, 2015 3:02:49 PM Subject: Re: [OmniOS-discuss] Problem updating ms-omniti local repo Sorry I forgot to specify we are running 151014 Javier Lopez System Administrator phone: mobile: fax: mail: web: +49 89 287018 0 +49 89 287018 18 javier.lopez at trixter.de http://www.trixter.de TRIXTER Film GmbH Amalienstrasse 67 80799 Munich, Germany TRIXTER FILM GMBH, Sitz Muenchen, Registergericht Muenchen, HRB 125141, Geschaeftsfuehrer: Simone Kraus, Michael Coldewey, Christian Sommer TRIXTER FILM GMBH, Place of business Munich, Register Court Munich, HRB 125141, Managing Directors: Simone Kraus, Michael Coldewey, Christian Sommer ----- Original Message ----- From: "Javier Lopez" To: "Volker A. Brandt" Cc: omnios-discuss at lists.omniti.com, "Trixter IT Dept" Sent: Wednesday, June 17, 2015 2:59:18 PM Subject: Re: [OmniOS-discuss] Problem updating ms-omniti local repo Hi Volker, Thank you very much for your help. I think we had a small misunderstanding though: I don't have a problem with creating the repo, I in fact having it working now. I used your advice from your second email: # pkgrepo create /path/ms.omniti.com/ # pkgrepo add-publisher -s /path/ms.omniti.com/ ms.omniti.com # pkgrecv -s http://pkg.omniti.com/omniti-ms/ -d /path/ms.omniti.com/ '*' However, from the beginning, I've had a problem with updating it. I continue to receive this error message: #> pkgrecv -s http://pkg.omniti.com/omniti-ms/ -d /path/ms.omniti.com/ '*' Processing packages for publisher omniti-ms ... Retrieving catalog ' http://pkg.omniti.com/omniti-ms/ '... Unable to retrieve package data for publisher 'omniti-ms' from one of the following origin(s): http://pkg.omniti.com/omniti-ms/ The catalog retrieved from one of the origin(s) listed above only contains package data for: ms.omniti.com. This is either a result of invalid origin information being provided for publisher 'omniti-ms', or because the wrong publisher name was provided when this publisher was added. I noticed in your last email you used a different method of cloning the repo: | # pkgrecv -p ms.omniti.com -s http://pkg.omniti.com/omniti-ms/ -d /pkg/ms.omniti.com/ --clone We ran into a problem with this command-- "pkgrecv: Illegal option --p". Any idea what is going on here? We are running Your pkgrecv command as written specifies the publisher (I guess with the -p option): is this perhaps what is giving us a problem with updating? If that is the case, is there a way to fix what we already without having to download the entire repo again? (Just to give a bit of background, the reason why we asked this question in the first place is because we want to have our repo management server update our omnios repos on a cron job) Regards Javier Lopez System Administrator phone: mobile: fax: mail: web: +49 89 287018 0 +49 89 287018 18 javier.lopez at trixter.de http://www.trixter.de TRIXTER Film GmbH Amalienstrasse 67 80799 Munich, Germany TRIXTER FILM GMBH, Sitz Muenchen, Registergericht Muenchen, HRB 125141, Geschaeftsfuehrer: Simone Kraus, Michael Coldewey, Christian Sommer TRIXTER FILM GMBH, Place of business Munich, Register Court Munich, HRB 125141, Managing Directors: Simone Kraus, Michael Coldewey, Christian Sommer ----- Original Message ----- From: "Volker A. Brandt" To: "Javier Lopez" Cc: omnios-discuss at lists.omniti.com, "Trixter IT Dept" Sent: Wednesday, June 17, 2015 9:06:34 AM Subject: Re: [OmniOS-discuss] Problem updating ms-omniti local repo Hi Javier! > T hank you for the answer Volker You are welcome. > I've recreated the repo following your indications and I still have > the same problem when I try to update the repo. Don't worry, you will get there. Here is exactly what I did (I wanted a local copy anyway :-): | # zfs create -o mountpoint=/pkg/ms.omniti.com dpool/pkg/ms.omniti.com | # zfs set atime=off dpool/pkg/ms.omniti.com | # zfs set setuid=off dpool/pkg/ms.omniti.com | # zfs set devices=off dpool/pkg/ms.omniti.com | # zfs set exec=off dpool/pkg/ms.omniti.com | # pkgrepo create /pkg/ms.omniti.com | # pkgrepo add-publisher -s/pkg/ms.omniti.com ms.omniti.com Now you have a choice. You can clone the repository (= make an exact copy): | # pkgrecv -p ms.omniti.com -s http://pkg.omniti.com/omniti-ms/ -d /pkg/ms.omniti.com/ --clone | Processing packages for publisher ms.omniti.com ... | Retrieving and evaluating 1473 package(s)... | DOWNLOAD PKGS FILES XFER (MB) SPEED | .../postgresql-936/pg_buffercache 1473/1473 617732/617732 6552/6552 718k/s | | Verifying repository contents. | Initiating repository verification. | pkg://ms.omniti.com/omniti/monitoring/nad 8/1474 -Traceback (most recent call last): | File "/usr/lib/python2.6/vendor-packages/pkg/server/repository.py", line 2391, in verify | trust_anchors, sig_required_names, use_crls): | File "/usr/lib/python2.6/vendor-packages/pkg/server/repository.py", line 2288, in __gen_verify | hashes, errors = self.__get_hashes(path, pfmri) | File "/usr/lib/python2.6/vendor-packages/pkg/server/repository.py", line 2005, in __get_hashes | fnames = fname.split() | AttributeError: 'NoneType' object has no attribute 'split' | Traceback (most recent call last): | File "/usr/bin/pkgrepo", line 1689, in handle_errors | __ret = func(*args, **kwargs) | File "/usr/bin/pkgrepo", line 1665, in main_func | return func(conf, pargs) | File "/usr/bin/pkgrepo", line 1512, in subcmd_verify | for verify_tuple in repo.verify(pub=xpub, progtrack=progtrack): | File "/usr/lib/python2.6/vendor-packages/pkg/server/repository.py", line 2396, in verify | raise apx._convert_error(e) | AttributeError: 'NoneType' object has no attribute 'split' | | | pkgrepo: This is an internal error in pkg(5) version 1427212657. Please log a | Service Request about this issue including the information above and this | message. | | pkgrecv: Pkgrepo verify found errors in the updated repository. | The original package catalog has been restored. | The clone operation can be retried; package content that has already been retrieved will not be downloaded again. This is a crash in "pkgrepo verify", which will run automatically after the actual copy. But the repository is there. Or you can retrieve just the latest packages: | # pkgrecv -m latest -s http://pkg.omniti.com/omniti-ms/ -d /pkg/ms.omniti.com/ '*' | Processing packages for publisher omniti-ms ... | Retrieving catalog 1/1 omniti-ms 114.48 kB | Unable to retrieve package data for publisher 'omniti-ms' from one | of the following origin(s): | | http://pkg.omniti.com/omniti-ms/ | | The catalog retrieved from one of the origin(s) listed above only | contains package data for: ms.omniti.com. | | This is either a result of invalid origin information being provided | for publisher 'omniti-ms', or because the wrong publisher name was | provided when this publisher was added. | | Retrieving and evaluating 809 package(s)... | Download Manifests ( 36/809) - Again, there is an error message (because either the FMRI with "omniti-ms" confuses pkgrecv, or there really is an empty publisher entry for "omniti-ms" in the original repository), but the data will be there. Regards -- Volker -- ------------------------------------------------------------------------ Volker A. Brandt Consulting and Support for Oracle Solaris Brandt & Brandt Computer GmbH WWW: http://www.bb-c.de/ Am Wiesenpfad 6, 53340 Meckenheim, GERMANY Email: vab at bb-c.de Handelsregister: Amtsgericht Bonn, HRB 10513 Schuhgr??e: 46 Gesch?ftsf?hrer: Rainer J.H. Brandt und Volker A. Brandt "When logic and proportion have fallen sloppy dead" -------------- next part -------------- An HTML attachment was scrubbed... URL: From vab at bb-c.de Wed Jun 17 14:15:26 2015 From: vab at bb-c.de (Volker A. Brandt) Date: Wed, 17 Jun 2015 16:15:26 +0200 Subject: [OmniOS-discuss] Problem updating ms-omniti local repo In-Reply-To: <1560707420.47257.1434546169153.JavaMail.root@de001cs001.trixter.de> References: <1983432428.47247.1434545958611.JavaMail.root@de001cs001.trixter.de> <1560707420.47257.1434546169153.JavaMail.root@de001cs001.trixter.de> Message-ID: <21889.33022.997158.188273@glaurung.bb-c.de> Hi Javier! > Sorry I forgot to specify we are running 151014 Same here: # uname -a SunOS nfs 5.11 omnios-170cea2 i86pc i386 i86pc # cat /etc/release OmniOS v11 r151014 Copyright 2015 OmniTI Computer Consulting, Inc. All rights reserved. Use is subject to license terms. # pkg version 1427212657 > I think we had a small misunderstanding though: I don't have a > problem with creating the repo, I in fact having it working now. I > used your advice from your second email: Ah, that is good. > However, from the beginning, I've had a problem with updating it. I > continue to receive this error message: > > #> pkgrecv -s http://pkg.omniti.com/omniti-ms/ -d /path/ms.omniti.com/ '*' > Processing packages for publisher omniti-ms ... > Retrieving catalog ' http://pkg.omniti.com/omniti-ms/ '... > Unable to retrieve package data for publisher 'omniti-ms' from one > of the following origin(s): > http://pkg.omniti.com/omniti-ms/ > The catalog retrieved from one of the origin(s) listed above only > contains package data for: ms.omniti.com. > This is either a result of invalid origin information being provided > for publisher 'omniti-ms', or because the wrong publisher name was > provided when this publisher was added. Yes, this message is annoying. I tried to explain what I think why this message appears. However, the repository is good despite the message. Just ignore it. > | # pkgrecv -p ms.omniti.com -s http://pkg.omniti.com/omniti-ms/ -d > /pkg/ms.omniti.com/ --clone > > We ran into a problem with this command-- "pkgrecv: Illegal option > --p". Any idea what is going on here? Maybe your pkgrecv simply does not have the -p option. Your version of pkg(5) might be too old. This is the version I have in my 151014 system: # pkg info pkg Name: package/pkg Summary: Image Packaging System Description: The Image Packaging System (IPS), or pkg(5), is the software delivery system used on Oracle Solaris. This package contains the core command-line components and pkg.depotd server. Category: System/Packaging State: Installed Publisher: omnios Version: 0.5.11 Branch: 0.151014 Packaging Date: Thu Apr 2 18:43:07 2015 Size: 11.75 MB FMRI: pkg://omnios/package/pkg at 0.5.11-0.151014:20150402T184307Z What does "pkg info pkg" show on your box? Regards -- Volker -- ------------------------------------------------------------------------ Volker A. Brandt Consulting and Support for Oracle Solaris Brandt & Brandt Computer GmbH WWW: http://www.bb-c.de/ Am Wiesenpfad 6, 53340 Meckenheim, GERMANY Email: vab at bb-c.de Handelsregister: Amtsgericht Bonn, HRB 10513 Schuhgr??e: 46 Gesch?ftsf?hrer: Rainer J.H. Brandt und Volker A. Brandt "When logic and proportion have fallen sloppy dead" From danmcd at omniti.com Thu Jun 18 05:37:47 2015 From: danmcd at omniti.com (Dan McDonald) Date: Thu, 18 Jun 2015 01:37:47 -0400 Subject: [OmniOS-discuss] OmniOS bloody updated --> now with gcc51 for non-illumos bits In-Reply-To: <2D03B99D-BEE9-4A51-9F58-8AD971990C07@omniti.com> References: <2D03B99D-BEE9-4A51-9F58-8AD971990C07@omniti.com> Message-ID: <8DB20F2A-007F-4F9A-952F-6C63387794B4@omniti.com> Earlier, I said... > On Jun 16, 2015, at 10:16 PM, Dan McDonald wrote: > > I've pushed back the changes for gcc51 into omnios-build. I now need to build it a final time, and test upgrading. > > If you pull changes into omnios-build, you will be able to build most of the world, but not the gcc51 portions of it. Watch this list for a followup announcement when the bloody binaries on IPS are updated. I've now updated the IPS binaries on http://pkg.omniti.com/omnios/bloody/. The old "gcc/g++-4-runtime" has been marked as "renamed" and "pkg update" will download the new "gcc/g++-5-runtime" instead. These packages, however, will need to be installed explicitly to supplant their gcc48 predecessors: libgmp-gcc51 libmpfr-gcc51 libmpc-gcc51 gcc51 You can resume builds of omnios-build's non-illumos bits after this update and the installation of the gcc51 packages. An important safety tip for gcc51 --> its default standard's mode is "gnu11". Some software assumes "gnu89", so you may need to add "-std=gnu89" to your CFLAGS. You'll note some of the gcc51 changes in omnios-build were to do just that for certain pieces of software. I've also updated the ISO, USB-dd, and Kayak .zfs.bz2 images. Details can be found on the http://omnios.omniti.com/wiki.php/Installation page. Thank you! Dan From danmcd at omniti.com Thu Jun 18 14:53:09 2015 From: danmcd at omniti.com (Dan McDonald) Date: Thu, 18 Jun 2015 10:53:09 -0400 Subject: [OmniOS-discuss] OmniOS bloody updated --> now with gcc51 for non-illumos bits In-Reply-To: <8DB20F2A-007F-4F9A-952F-6C63387794B4@omniti.com> References: <2D03B99D-BEE9-4A51-9F58-8AD971990C07@omniti.com> <8DB20F2A-007F-4F9A-952F-6C63387794B4@omniti.com> Message-ID: <4FC3BF11-A62E-4E38-AA24-182FD53C0E8D@omniti.com> Forgot to mention some per-release tidbits: 1.) illumos-omnios master branch, commit 4f84bf5 2.) omnios-build master branch, commit 248fbc8 Dan From graham at thestephensdomain.com Fri Jun 19 15:14:27 2015 From: graham at thestephensdomain.com (Graham Stephens) Date: Fri, 19 Jun 2015 16:14:27 +0100 Subject: [OmniOS-discuss] How do I rename a zone (including pool paths)? Message-ID: <558431D3.5050106@thestephensdomain.com> Seems a simple question, but I've tried: zoneadm move which doesn't - it leaves the old zonepath and creates a link which causes me no end of confusion as I want another zone called what this used to be called... If I try to copy, using: zoneadm -z yyyy clone -m copy xxxx it tells me that there is no "-m copy" option for an ipkg zone (then the help kindly informs me I can use "copy" as a -m option). ;) I don't want to use a snapshot because I want to delete the old zone info afterwards. Once again I'm afraid I'm lost (not unusual)...! From jdg117 at elvis.arl.psu.edu Fri Jun 19 16:12:26 2015 From: jdg117 at elvis.arl.psu.edu (John D Groenveld) Date: Fri, 19 Jun 2015 12:12:26 -0400 Subject: [OmniOS-discuss] How do I rename a zone (including pool paths)? In-Reply-To: Your message of "Fri, 19 Jun 2015 16:14:27 BST." <558431D3.5050106@thestephensdomain.com> References: <558431D3.5050106@thestephensdomain.com> Message-ID: <201506191612.t5JGCQPE022549@elvis.arl.psu.edu> In message <558431D3.5050106 at thestephensdomain.com>, Graham Stephens writes: >Seems a simple question, but I've tried: Try kludging it via the zone migration recipe. # zoneadm -z foo detach # zonecfg -z foo export >/root/bar.zonecfg ... fix zonepath # zfs set zoned=off zones/foo/ROOT # zfs rename zones/foo zones/bar # zfs set zoned=on zones/bar/ROOT # zfs set mountpoint=/zones/bar zones/bar # zfs set mountpoint=/zones/bar/root zones/bar/ROOT/zbe-1 # zonecfg -z bar -f /root/bar.zonecfg # zoneadm -z bar attach John groenveld at acm.org From graham at thestephensdomain.com Fri Jun 19 20:35:35 2015 From: graham at thestephensdomain.com (Graham Stephens) Date: Fri, 19 Jun 2015 21:35:35 +0100 Subject: [OmniOS-discuss] How do I rename a zone (including pool paths)? In-Reply-To: <201506191612.t5JGCQPE022549@elvis.arl.psu.edu> References: <558431D3.5050106@thestephensdomain.com> <201506191612.t5JGCQPE022549@elvis.arl.psu.edu> Message-ID: <55847D17.8020203@thestephensdomain.com> On 19/06/2015 17:12, John D Groenveld wrote: > In message <558431D3.5050106 at thestephensdomain.com>, Graham Stephens writes: >> Seems a simple question, but I've tried: > > Try kludging it via the zone migration recipe. > # zoneadm -z foo detach > # zonecfg -z foo export >/root/bar.zonecfg > ... fix zonepath > # zfs set zoned=off zones/foo/ROOT > # zfs rename zones/foo zones/bar > # zfs set zoned=on zones/bar/ROOT > # zfs set mountpoint=/zones/bar zones/bar > # zfs set mountpoint=/zones/bar/root zones/bar/ROOT/zbe-1 > # zonecfg -z bar -f /root/bar.zonecfg > # zoneadm -z bar attach > > John > groenveld at acm.org > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss > Thanks John. I did have some trouble with the mountpoint - I don't know why - but retrying the commands a few times got there eventually. Cheers. From Josh.Barton at usurf.usu.edu Fri Jun 19 23:10:29 2015 From: Josh.Barton at usurf.usu.edu (Josh Barton) Date: Fri, 19 Jun 2015 23:10:29 +0000 Subject: [OmniOS-discuss] HP Smart Array Controller Drivers Message-ID: I am trying to install some hp drivers (CPQary3) on omnios, currently without the drivers the install cd does not see the drives. I managed to load the cd with the drivers and tried to install them using the terminal option available right before install but because /usr is read only it didn't work. Any suggestions? Incidentally what is the best way to mount cd's...and find their mountpoint? Thanks! Josh Barton -------------- next part -------------- An HTML attachment was scrubbed... URL: From jdg117 at elvis.arl.psu.edu Sat Jun 20 12:50:29 2015 From: jdg117 at elvis.arl.psu.edu (John D Groenveld) Date: Sat, 20 Jun 2015 08:50:29 -0400 Subject: [OmniOS-discuss] HP Smart Array Controller Drivers In-Reply-To: Your message of "Fri, 19 Jun 2015 23:10:29 -0000." References: Message-ID: <201506201250.t5KCoTDQ014763@elvis.arl.psu.edu> In message , Josh Barton writes: >Incidentally what is the best way to mount cd's...and find their mountpoint= >? Does cfgadm(1M) report the CD/DVD reader? John groenveld at acm.org From Josh.Barton at usurf.usu.edu Mon Jun 22 22:13:12 2015 From: Josh.Barton at usurf.usu.edu (Josh Barton) Date: Mon, 22 Jun 2015 22:13:12 +0000 Subject: [OmniOS-discuss] illumos Compatible Server Configurations Message-ID: I am looking for information on possible server configurations that will work with OmniOS. One server I have looked at is the: Power Edge R320 by Dell, but they don't always provide all the information needed to determine if illumos will run on it. Thanks, Josh Barton Utah State University Research Foundation -------------- next part -------------- An HTML attachment was scrubbed... URL: From jdg117 at elvis.arl.psu.edu Tue Jun 23 00:04:10 2015 From: jdg117 at elvis.arl.psu.edu (John D Groenveld) Date: Mon, 22 Jun 2015 20:04:10 -0400 Subject: [OmniOS-discuss] illumos Compatible Server Configurations In-Reply-To: Your message of "Mon, 22 Jun 2015 22:13:12 -0000." References: Message-ID: <201506230004.t5N04AQh015939@elvis.arl.psu.edu> In message , Josh Barton writes: >I am looking for information on possible server configurations that will wo= >rk with OmniOS. One server I have looked at is the: Power Edge R320 by Dell= >, but they don't always provide all the information needed to determine if = >illumos will run on it. Tell Dell's salescritter you want the Intel NIC in place of the stock Broadcom and the LSI SAS HBA in place of the stock H310 hardware RAID controller. I don't have part numbers but the LSI HBA's specsheet is here: The H310 will work, but you'll be buying extraneous silicon and firmware that you'll be reflashing to IT mode. If Dell's salecritter is clueless, tell him you want to talk to a sales engineer who worked the Nexenta partnership. Happy shopping, John groenveld at acm.org From danmcd at omniti.com Tue Jun 23 00:20:43 2015 From: danmcd at omniti.com (Dan McDonald) Date: Mon, 22 Jun 2015 20:20:43 -0400 Subject: [OmniOS-discuss] illumos Compatible Server Configurations In-Reply-To: <201506230004.t5N04AQh015939@elvis.arl.psu.edu> References: <201506230004.t5N04AQh015939@elvis.arl.psu.edu> Message-ID: <40D47E39-893F-42FF-857B-8EAAE817E3B4@omniti.com> Thanks John for answering this! > On Jun 22, 2015, at 8:04 PM, John D Groenveld wrote: > > In message , Josh Barton > writes: >> I am looking for information on possible server configurations that will wo= >> rk with OmniOS. One server I have looked at is the: Power Edge R320 by Dell= >> , but they don't always provide all the information needed to determine if = >> illumos will run on it. > > Tell Dell's salescritter you want the Intel NIC in place of the > stock Broadcom The Broadcoms actually are somewhat better-supported now thanks to recent pushes in illumos-gate, which are also in r151014. Still, generally speaking the Intel NICs are better if you can get 'em. > and the LSI SAS HBA in place of the stock H310 > hardware RAID controller. Amen here --> And the last bit: > If Dell's salecritter is clueless, tell him you want to talk to > a sales engineer who worked the Nexenta partnership. A Nexenta-happy box will run OmniOS just fine as well. Dan From lslusser at gmail.com Tue Jun 23 08:20:07 2015 From: lslusser at gmail.com (Liam Slusser) Date: Tue, 23 Jun 2015 01:20:07 -0700 Subject: [OmniOS-discuss] illumos Compatible Server Configurations In-Reply-To: <40D47E39-893F-42FF-857B-8EAAE817E3B4@omniti.com> References: <201506230004.t5N04AQh015939@elvis.arl.psu.edu> <40D47E39-893F-42FF-857B-8EAAE817E3B4@omniti.com> Message-ID: Dell has partnered with Nexenta and provided blessed hardware which was designed to run Illumos. You can ask your Dell sales person to spec hardware to "Nexenta Configurations." I've ordered a few "Nexenta Configuration" type systems and they've all worked great with OmniOS. http://www.dell.com/learn/us/en/04/campaigns/dell-nexenta-storage If you're having a problem finding somebody to help you, email me privately and I can connect with you with somebody over at Dell that is in the know. thanks! liam On Mon, Jun 22, 2015 at 5:20 PM, Dan McDonald wrote: > Thanks John for answering this! > > > On Jun 22, 2015, at 8:04 PM, John D Groenveld > wrote: > > > > In message , > Josh Barton > > writes: > >> I am looking for information on possible server configurations that > will wo= > >> rk with OmniOS. One server I have looked at is the: Power Edge R320 by > Dell= > >> , but they don't always provide all the information needed to determine > if = > >> illumos will run on it. > > > > Tell Dell's salescritter you want the Intel NIC in place of the > > stock Broadcom > > The Broadcoms actually are somewhat better-supported now thanks to recent > pushes in illumos-gate, which are also in r151014. > > Still, generally speaking the Intel NICs are better if you can get 'em. > > > and the LSI SAS HBA in place of the stock H310 > > hardware RAID controller. > > Amen here --> And the last bit: > > > If Dell's salecritter is clueless, tell him you want to talk to > > a sales engineer who worked the Nexenta partnership. > > A Nexenta-happy box will run OmniOS just fine as well. > > Dan > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss > -------------- next part -------------- An HTML attachment was scrubbed... URL: From javier.lopez at trixter.de Tue Jun 23 16:54:13 2015 From: javier.lopez at trixter.de (Lopez, Javier) Date: Tue, 23 Jun 2015 18:54:13 +0200 (CEST) Subject: [OmniOS-discuss] Problem updating ms-omniti local repo In-Reply-To: <21889.33022.997158.188273@glaurung.bb-c.de> Message-ID: <208911579.53641.1435078452952.JavaMail.root@de001cs001.trixter.de> Hello again, I didn't notice that the server managing the updating process was running r151012 ... I'm sorry for that. I've upgraded it to the r151014. So now: >>> # pkg info pkg Name: package/pkg Summary: Image Packaging System Description: The Image Packaging System (IPS), or pkg(5), is the software delivery system used on Oracle Solaris. This package contains the core command-line components and pkg.depotd server. Category: System/Packaging State: Installed Publisher: omnios Version: 0.5.11 Branch: 0.151014 Packaging Date: April 2, 2015 06:43:07 PM Size: 11.75 MB FMRI: pkg://omnios/package/pkg at 0.5.11-0.151014:20150402T184307Z <<< I've tried again following your instructions. The problem updating the repo is still here, I get the message: >>> # pkgrecv -s http://pkg.omniti.com/omniti-ms/ -d /mnt/rpm-repo/omnios/ms.omniti.com/ -m all-timestamps '*' Processing packages for publisher omniti-ms ... Retrieving catalog 1/1 omniti-ms 115.63 kB Unable to retrieve package data for publisher 'omniti-ms' from one of the following origin(s): http://pkg.omniti.com/omniti-ms/ The catalog retrieved from one of the origin(s) listed above only contains package data for: ms.omniti.com. This is either a result of invalid origin information being provided for publisher 'omniti-ms', or because the wrong publisher name was provided when this publisher was added. Retrieving and evaluating 1492 package(s)... PROCESS ITEMS GET (MB) SEND (MB) network/iftop 1/1492 0/6591 0/19490 pkgrecv: 'open' failed for transaction ID '1376680458_pkg%3A%2F%2Fms.omniti.com%2Fnetwork%2Fiftop%401.0.2%2C5.11-0.151006%3A20130816T191418Z': The specified FMRI, 'pkg://ms.omniti.com/network/iftop at 1.0.2,5.11-0.151006:20130816T191418Z', already exists or has been restricted. pkgrecv: Cached files were preserved in the following directory: /var/tmp/pkgrecv-T0Njj3 Use pkgrecv -c to resume the interrupted download. <<< As you can see, the initial error is the same as the one I sent to you earlier. This is the error you told me to ignore earlier, and the process does still run but is stopped by the error above as it tries to retrieve packages. Sometimes it is able to retrieve and evaluate a number of packages before it halts. Are you able to get all the packages even when you get the initial error when you're trying to update the repo? This is the big question for me. In case this helps, I did the same process with another repo and it works fine: >>> # pkgrecv -s http://pkg.omniti.com/omnios/r151014/ -d /mnt/rpm-repo/omnios/r151014/ -m all-timestamps '*' Processing packages for publisher omnios ... Retrieving and evaluating 1882 package(s)... PROCESS ITEMS GET (MB) SEND (MB) Completed 865/865 225.9/225.9 541.2/541.2 <<< If I try again nothing happens (because it is already updated?) >>> # pkgrecv -s http://pkg.omniti.com/omnios/r151014/ -d /mnt/rpm-repo/omnios/r151014/ -m all-timestamps '*' Processing packages for publisher omnios ... Retrieving and evaluating 1882 package(s)... <<< Before: >>> # pkgrepo info -s /mnt/rpm-repo/omnios/r151014/ PUBLISHER PACKAGES STATUS UPDATED omnios 1016 online 2015-06-14T16:01:01.659806Z <<< After: >>> # pkgrepo info -s /mnt/rpm-repo/omnios/r151014/ PUBLISHER PACKAGES STATUS UPDATED omnios 1016 online 2015-06-23T16:19:00.901415Z <<< Thank you so much for your time. :) Regards Javier. ----- Original Message ----- From: "Volker A. Brandt" To: "Javier Lopez" Cc: omnios-discuss at lists.omniti.com, "Trixter IT Dept" Sent: Wednesday, June 17, 2015 4:15:26 PM Subject: Re: [OmniOS-discuss] Problem updating ms-omniti local repo Hi Javier! > Sorry I forgot to specify we are running 151014 Same here: pkg in# uname -a SunOS nfs 5.11 omnios-170cea2 i86pc i386 i86pc # cat /etc/release OmniOS v11 r151014 Copyright 2015 OmniTI Computer Consulting, Inc. All rights reserved. Use is subject to license terms. # pkg version 1427212657 > I think we had a small misunderstanding though: I don't have a > problem with creating the repo, I in fact having it working now. I > used your advice from your second email: Ah, that is good. > However, from the beginning, I've had a problem with updating it. I > continue to receive this error message: > > #> pkgrecv -s http://pkg.omniti.com/omniti-ms/ -d /path/ms.omniti.com/ '*' > Processing packages for publisher omniti-ms ... > Retrieving catalog ' http://pkg.omniti.com/omniti-ms/ '... > Unable to retrieve package data for publisher 'omniti-ms' from one > of the following origin(s): > http://pkg.omniti.com/omniti-ms/ > The catalog retrieved from one of the origin(s) listed above only > contains package data for: ms.omniti.com. > This is either a result of invalid origin information being provided > for publisher 'omniti-ms', or because the wrong publisher name was > provided when this publisher was added. Yes, this message is annoying. I tried to explain what I think why this message appears. However, the repository is good despite the message. Just ignore it. > | # pkgrecv -p ms.omniti.com -s http://pkg.omniti.com/omniti-ms/ -d > /pkg/ms.omniti.com/ --clone > > We ran into a problem with this command-- "pkgrecv: Illegal option > --p". Any idea what is going on here? Maybe your pkgrecv simply does not have the -p option. Your version of pkg(5) might be too old. This is the version I have in my 151014 system: # pkg info pkg Name: package/pkg Summary: Image Packaging System Description: The Image Packaging System (IPS), or pkg(5), is the software delivery system used on Oracle # uname -a SunOS nfs 5.11 omnios-170cea2 i86pc i386 i86pc # cat /etc/release OmniOS v11 r151014> However, from the beginning, I've had a problem with updating it. I > continue to receive this error message: > > #> pkgrecv -s http://pkg.omniti.com/omniti-ms/ -d /path/ms.omniti.com/ '*' > Processing packages for publisher omniti-ms ... > Retrieving catalog ' http://pkg.omniti.com/omniti-ms/ '... > Unable to retrieve package data for publisher 'omniti-ms' from one Copyright 2015 OmniTI Computer Consulting, Inc. All rights reserved. Use is subject to license terms. # pkg version 1427212657 > I think we had a small misunderstanding though: I don't have a > problem with creating the repo, I in fact having it working now. I > used your advice from your second email: Ah, that is good. > However, from the beginning, I've had a problem with updating it. I > continue to receive this error message: > > #> pkgrecv -s http://pkg.omniti.com/omniti-ms/ -d /path/ms.omniti.com/ '*' > Processing packages for publisher omniti-ms ... > Retrieving catalog ' http://pkg.omniti.com/omniti-ms/ '... > Unable to retrieve package data for publisher 'omniti-ms' from one > of the following origin(s): > http://pkg.omniti.com/omniti-ms/ > The catalog retrieved from one of the origin(s) listed above only > contains package data for: ms.omniti.com. > This is either a result of invalid origin information being provided > for publisher 'omniti-ms', or because the wrong publisher name was > provided when this publisher was added. Solaris. This package contains the core command-line components and pkg.depotd server. Category: System/Packaging State: Installed Publisher: omnios Version: 0.5.11 Branch: 0.151014 Packaging Date: Thu Apr 2 18:43:07 2015 Size: 11.75 MB FMRI: pkg://omnios/package/pkg at 0.5.11-0.151014:20150402T184307Z What does "pkg info pkg" show on your box? Regards -- Volker -- ------------------------------------------------------------------------ Volker A. Brandt Consulting and Support for Oracle Solaris Brandt & Brandt Computer GmbH WWW: http://www.bb-c.de/ Am Wiesenpfad 6, 53340 Meckenheim, GERMANY Email: vab at bb-c.de Handelsregister: Amtsgericht Bonn, HRB 10513 Schuhgr??e: 46 Gesch?ftsf?hrer: Rainer J.H. Brandt und Volker A. Brandt "When logic and proportion have fallen sloppy dead" From vab at bb-c.de Tue Jun 23 17:30:49 2015 From: vab at bb-c.de (Volker A. Brandt) Date: Tue, 23 Jun 2015 19:30:49 +0200 Subject: [OmniOS-discuss] Problem updating ms-omniti local repo In-Reply-To: <208911579.53641.1435078452952.JavaMail.root@de001cs001.trixter.de> References: <21889.33022.997158.188273@glaurung.bb-c.de> <208911579.53641.1435078452952.JavaMail.root@de001cs001.trixter.de> Message-ID: <21897.38857.982889.731865@glaurung.bb-c.de> Hello Javier! > Unable to retrieve package data for publisher 'omniti-ms' from one > of the following origin(s): > > http://pkg.omniti.com/omniti-ms/ > > The catalog retrieved from one of the origin(s) listed above only > contains package data for: ms.omniti.com. I guess that this could be the result of an error in the IPS software. Maybe the "omniti-ms" in the pkg origin URI makes pkgrecv think that this should be the publisher name. It could also be because at one time the repository hosted at "http://pkg.omniti.com/omniti-ms/" really did contain packages for the publisher "ms-omniti". FWIW a "pkg list -g http://pkg.omniti.com/omniti-ms/" will give a similar but slightly longer message: Unable to retrieve package data for publisher 'omniti-ms' from one of the following origin(s): http://pkg.omniti.com/omniti-ms/ The catalog retrieved from one of the origin(s) listed above only contains package data for: ms.omniti.com. To resolve this issue, correct the origin information provided for publisher 'omniti-ms' using the pkg set-publisher subcommand, or re-add the publisher using the correct name and remove the 'omniti-ms' publisher. To re-add this publisher with the correct name, execute the following commands as a privileged user: pkg set-publisher -P -g http://pkg.omniti.com/omniti-ms/ ms.omniti.com pkg unset-publisher omniti-ms This advice is bogus because the publisher was never set to "omniti-ms" in the first place. :-( Anyway, you can safely ignore this message because "ms.omniti.com" is the publisher you want. > PROCESS ITEMS GET (MB) SEND (MB) > network/iftop 1/1492 0/6591 0/19490 > pkgrecv: 'open' failed for transaction ID '1376680458_pkg%3A%2F%2Fms.omniti.com%2Fnetwork%2Fiftop%401.0.2%2C5.11-0.151006%3A20130816T191418Z': The specified FMRI, 'pkg://ms.omniti.com/network/iftop at 1.0.2,5.11-0.151006:20130816T191418Z', already exists or has been restricted. > [...] > As you can see, the initial error is the same as the one I sent to you > earlier. This is the error you told me to ignore earlier Hmmm... I think I said to ignore the other error above. The "'open failed'" error is a bug in the repo and can only be fixed locally. I cannot really tell what "restricted" means. The package has a duplicate package name in the metadata: # pkg contents -m -g http://pkg.omniti.com/omniti-ms/ network/iftop | fgrep pkg.fmri [...error message omitted...] set name=pkg.fmri value=pkg://ms.omniti.com/network/iftop at 1.0.2,5.11-0.151006:20130816T191418Z set name=pkg.fmri value=pkg://ms.omniti.com/omniti/network/iftop at 1.0.2,5.11-0.151006:20130816T191228Z The second entry should be removed. That is really the name of the new package the old "network/iftop" package was renamed to. Unfortunately it is not possible to tell pkgrecv "give me everything *except* "network/iftop at 1.0.2,5.11-0.151006:20130816T191418Z". You could retrieve a list of the latest version all packages in the repo and ask for those packages explicitly. This could be scripted. Unfortunately, I do not have time for that at the moment. > Are you able to get all the packages even when you get the initial error > when you're trying to update the repo? This is the big question for me. No, I get exactly the same error. Dan & Eric, can you help? Thanks -- Volker -- ------------------------------------------------------------------------ Volker A. Brandt Consulting and Support for Oracle Solaris Brandt & Brandt Computer GmbH WWW: http://www.bb-c.de/ Am Wiesenpfad 6, 53340 Meckenheim, GERMANY Email: vab at bb-c.de Handelsregister: Amtsgericht Bonn, HRB 10513 Schuhgr??e: 46 Gesch?ftsf?hrer: Rainer J.H. Brandt und Volker A. Brandt "When logic and proportion have fallen sloppy dead" From vab at bb-c.de Tue Jun 23 17:38:02 2015 From: vab at bb-c.de (Volker A. Brandt) Date: Tue, 23 Jun 2015 19:38:02 +0200 Subject: [OmniOS-discuss] Problem updating ms-omniti local repo In-Reply-To: <21897.38857.982889.731865@glaurung.bb-c.de> References: <21889.33022.997158.188273@glaurung.bb-c.de> <208911579.53641.1435078452952.JavaMail.root@de001cs001.trixter.de> <21897.38857.982889.731865@glaurung.bb-c.de> Message-ID: <21897.39290.530683.454117@glaurung.bb-c.de> I wrote: > The package has a > duplicate package name in the metadata: > > # pkg contents -m -g http://pkg.omniti.com/omniti-ms/ network/iftop | fgrep pkg.fmri > [...error message omitted...] > set name=pkg.fmri value=pkg://ms.omniti.com/network/iftop at 1.0.2,5.11-0.151006:20130816T191418Z > set name=pkg.fmri value=pkg://ms.omniti.com/omniti/network/iftop at 1.0.2,5.11-0.151006:20130816T191228Z This is wrong. I should have asked for the full version: # pkg contents -m -g http://pkg.omniti.com/omniti-ms/ network/iftop at 1.0.2,5.11-0.151006:20130816T191418Z [...error message omitted...] set name=pkg.fmri value=pkg://ms.omniti.com/network/iftop at 1.0.2,5.11-0.151006:20130816T191418Z set name=pkg.renamed value=true set name=variant.opensolaris.zone value=global value=nonglobal depend fmri=pkg://ms.omniti.com/omniti/network/iftop type=require This is just a normal "renamed" package. I can't really see what the problem is here. Regards -- Volker -- ------------------------------------------------------------------------ Volker A. Brandt Consulting and Support for Oracle Solaris Brandt & Brandt Computer GmbH WWW: http://www.bb-c.de/ Am Wiesenpfad 6, 53340 Meckenheim, GERMANY Email: vab at bb-c.de Handelsregister: Amtsgericht Bonn, HRB 10513 Schuhgr??e: 46 Gesch?ftsf?hrer: Rainer J.H. Brandt und Volker A. Brandt "When logic and proportion have fallen sloppy dead" From jim at cos.ru Wed Jun 24 12:04:17 2015 From: jim at cos.ru (Jim Klimov) Date: Wed, 24 Jun 2015 14:04:17 +0200 Subject: [OmniOS-discuss] Fix to VirtualBox installer under OI/OmniOS In-Reply-To: References: Message-ID: Hello all, Earlier in March I posted a patch that fixes installation of VirtualBox on non-Oracle Solaris hosts (e.g. OpenIndiana and OmniOS). Lest it be lost in the list archives, I moved the patch (now licensed as GPLv2 to match VirtualBox licensing) and an accompanying README to GitHub: https://github.com/jimklimov/vboxconfig_sh Feel free to add the changes into the distributions (if you maintain packages/repositories), so such hackery won't be needed ;) Thanks, Jim Klimov ----- ???????? ????????? ----- ??: Jim Klimov ????: Monday, March 16, 2015 9:30 ????: [OmniOS-discuss] Fix to VirtualBox installer under OI/OmniOS ???? (To): openindiana-discuss at openindiana.org, omnios-discuss ????? (Cc): thomas at espix.net > Hello all, > Earlier this month I posted to the VirtualBox lists my patch for vboxconfig.sh to better support > installation on non-Sun/Oracle Solaris-like hosts, tested on OI (Hipster) and OmniOS (bloody). > It may support SXCE as well (has the code, but was not practically tested lately) since there are > aspects where SXCE was like Solaris 10 (SVR4 packaging) and yet more advanced than that > (CrossBow, etc.). > Unfortunately, so far the post fell through the cracks... so for possible integration into packages > provided by the illumos distributions (e.g. to help the work done by Thomas Gouverneur to turn > SVR releases into IPS, or to support proper GPL-source builds of VirtualBox), I'd like to repost > these fixes here too. Feel free to build on them, to make VB updates hiccup-less ;) > In my practice, it sufficed to "pkgtrans" the Oracle VB package into a directory, patch the script > "vboxconfig.sh" provided in the installation image, update the "pkgmap" with new values for > size and checksum of the changed file, (run "pkgrm" to remove an obsoleted installation) and > "pkgadd" to install the package from the directory. > > --- > The reposted part: > http://article.gmane.org/gmane.comp.emulators.virtualbox.devel/7390 > Date: 2015-03-09 15:09:25 GMT > Hello all, > I submit a patch which should simplify installation of modern VirtualBox on some > non-Oracle derivate distributions of Solaris which are known to host VirtualBox > just fine. Most of the patch deals with "proper" detection of major and minor OS > version numbers on those distributions. Also there is now a touchable filename > to enforce installation of USB filters and corresponding UNIX group accounts, > as a workaround for further distros not detected as supported by even new code. > I submit this patch under the terms of MIT license. > This was last tested with VirtualBox 4.3.24 and OpenIndiana Hipster and OmniOS > Bloody, all updated today to the most current states available. > Note this was not yet tested in practice with SXCE although snippets were developed > that should work there. I don't expect there are many of those installations left beside my > closet, but those are all pleased with old VBox versions for now... ain't broke, you know ;) > Hope this helps, > // Jim Klimov > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss -- +============================================================+ | | | ?????? ???????, Jim Klimov | | ??????????? ???????? CTO | | ??? "??? ? ??" JSC COS&HT | | | | +7-903-7705859 (cellular) mailto:jimklimov at cos.ru | | CC:admin at cos.ru,jimklimov at gmail.com | +============================================================+ | () ascii ribbon campaign - against html mail | | /\ - against microsoft attachments | +============================================================+ -------------- next part -------------- An HTML attachment was scrubbed... URL: From moo at wuffers.net Wed Jun 24 14:43:22 2015 From: moo at wuffers.net (wuffers) Date: Wed, 24 Jun 2015 10:43:22 -0400 Subject: [OmniOS-discuss] Array returning queue full (SCSI 0x28) running Veeam jobs Message-ID: After introducing Veeam into our environment (which has been the backend for both Hyper-V and VMware hosts for 2+ years with minimal issues), I've been trying to deal with a new intermittent issue. SAN traffic is via IB/SRP presented by COMSTAR and I am currently running r151012. What happens is that when my Veeam backup jobs run (at this point, I had one big job of ~60 VMs), latency on the datastore that the repository is on deteriorates. When I first saw this, it seemed it was only affecting the VM running the Veeam proxy/repository (esxtop was only showing high QAVG on that VM). I would vMotion this VM to another host and everything would work normally again (I would put the host into maintenance mode and reboot it to get it out of this "state"). This suggested that it was not a SAN issue and isolated to the VM and/or host. After updating to the latest Veeam B&R version (update 2), the issue persisted. Then I updated to vSphere 5.5 update 2e, and the issue went away for about 2 weeks (and my Veeam incident was closed). At this point, my backup job tried to merge the incremental backup back into the full, and that's when the issue reappeared. I was able to vMotion the VM to another host and get the job completed. I also noticed after update 2e that the latency appeared on any datastore that the VM had access to (had hot-added), and subsequently noticed via CDM benchmarks that it was all VMs on the affected host. I reopened the case at Veeam, and took their recommendation to split the job up into several smaller jobs. This seemed to work for a few days until I caught it doing it again. This time I opened a case with VMware - while I was on hold the server got out of the condition by itself (no more latency - I was actually in the middle of doing CrystalDiskMark benchmarks on a few VMs, on and off the affected host). The VMware engineer took a look at the vmkernel logs and highlighted the following entries: 2015-06-23T16:11:07.942Z cpu19:33135)<7>ib_srp: TGT[vmhba_mlx4_0.1.1 :: 1] Abort cmd[2a] SN 0x9276395 req=7a done=0 tsk_status=255 req_lim=73 tx_head=106811312 tx_tail=106811184 2015-06-23T16:11:07.942Z cpu19:33135)<7>ib_srp: TGT[vmhba_mlx4_0.1.1 :: 1] Abort cmd[2a] SN 0x927639a req=65 on lun 0 state 0 2015-06-23T16:11:07.942Z cpu19:33135)<7>ib_srp: TGT[vmhba_mlx4_0.1.1 :: 1] Abort cmd[2a] SN 0x927639a req=65 done=0 tsk_status=255 req_lim=72 tx_head=106811313 tx_tail=106811185 2015-06-23T16:11:07.942Z cpu19:33135)<7>ib_srp: TGT[vmhba_mlx4_0.1.1 :: 1] Abort cmd[2a] SN 0x9276396 req=4f on lun 11 state 0 2015-06-23T16:11:07.942Z cpu19:33135)<7>ib_srp: TGT[vmhba_mlx4_0.1.1 :: 1] Abort cmd[2a] SN 0x9276396 req=4f done=0 tsk_status=255 req_lim=71 tx_head=106811314 tx_tail=106811186 2015-06-23T16:11:07.942Z cpu19:33135)<7>ib_srp: TGT[vmhba_mlx4_0.1.1 :: 1] Abort cmd[2a] SN 0x9276397 req=36 on lun 10 state 0 2015-06-23T16:11:08.016Z cpu19:33135)<7>ib_srp: TGT[vmhba_mlx4_0.1.1 :: 1] Abort cmd[2a] SN 0x9276399 req=31 on lun 6 state 0 2015-06-23T16:05:16.110Z cpu26:32839)ScsiDeviceIO: 2325: Cmd(0x4136c28516c0) 0x88, CmdSN 0x8000001a from world 35997 to dev "naa.600144f09084e251000052919b460003" failed H:0x0 D:0x8 P:0x0 Possible sense data: 0x0 0x0 0x0. 2015-06-23T16:05:18.253Z cpu20:36138)ScsiDeviceIO: 2325: Cmd(0x4136c0428d40) 0x2a, CmdSN 0xa389000 from world 35970 to dev "naa.600144f09084e251000052919b560004" failed H:0x0 D:0x8 P:0x0 Possible sense data: 0x0 0x0 0x0. 2015-06-23T16:05:19.382Z cpu8:32821)ScsiDeviceIO: 2325: Cmd(0x412ece968480) 0x28, CmdSN 0x80000038 from world 36101 to dev "naa.600144f09084e251000052919b460003" failed H:0x0 D:0x8 P:0x0 Possible sense data: 0x0 0x0 0x0. 2015-06-23T16:15:06.963Z cpu24:33485)<7>ib_srp: TGT[vmhba_mlx4_0.1.1 :: 1] New target id_ext=0x0002c903000bc0bc ioc_guid=0x0002c903000bc0bc pkey=0xffff service_id=0x0002c903000bc0bc dgid=fe80:0000:0000:0000:0002:c903:000b:c0bd - req_lim 128 2015-06-23T16:15:06.963Z cpu23:33475)<7>ib_srp: TGT[vmhba_mlx4_0.1.1 :: 1] ALIVE!!! 2015-06-23T16:15:06.972Z cpu15:33730)NMP: nmp_ThrottleLogForDevice:2322: Cmd 0x16 (0x412ecb3737c0, 0) to dev "naa.600144f09084e25100005286cf240003" on path "vmhba_mlx4_0.1.1:C0:T1:L2 " Failed: H:0x0 D:0x2 P:0x0 Valid sense data: 0x6 0x29 0x0. Act:NONE 2015-06-23T16:15:06.972Z cpu15:33730)ScsiDeviceIO: 2325: Cmd(0x412ecb3737c0) 0x16, CmdSN 0x9f1f from world 0 to dev "naa.600144f09084e25100005286cf240003" failed H:0x0 D:0x2 P:0x0 Va lid sense data: 0x6 0x29 0x0. 2015-06-23T16:15:06.972Z cpu15:33730)ScsiCore: 1486: Power-on Reset occurred on naa.600144f09084e25100005286cf240003 2015-06-23T16:15:05.029Z cpu6:34170)World: 14302: VC opID hostd-8fa8 maps to vmkernel opID ba341ef0 2015-06-23T16:15:06.941Z cpu13:33135)<7>ib_srp: TGT[vmhba_mlx4_0.1.1 :: 1] Abort cmd[12] SN 0x92766c9 req=63 on lun 6 state 0 2015-06-23T16:15:06.941Z cpu13:33135)<7>ib_srp: TGT[vmhba_mlx4_0.1.1 :: 1] active_abort 1 lim 1 used_req 126 tag=0000000000000163 tsk_tag=0000000000000063 2015-06-23T16:15:06.941Z cpu13:33135)<7>ib_srp: TGT[vmhba_mlx4_0.1.1 :: 1] Abort cmd[12] SN 0x92766c9 req=63 done=0 tsk_status=255 req_lim=0 tx_head=106811505 tx_tail=106811377 2015-06-23T16:15:06.941Z cpu13:33135)WARNING: LinScsi: SCSILinuxAbortCommands:1843: Failed, Driver ib_srp, for vmhba_mlx4_0.1.1 2015-06-23T16:15:06.959Z cpu4:33135)<7>ib_srp: TGT[vmhba_mlx4_0.1.1 :: 1] Abort cmd[1a] SN 0x92766f2 req=5a on lun 10 state 0 2015-06-23T16:15:06.959Z cpu4:33135)<3>ib_srp: TGT[vmhba_mlx4_0.1.1 :: 1] tsk_mgmt no TX_IU head 106811505 tail 106811378 lim 0 2015-06-23T16:15:06.959Z cpu4:33135)WARNING: LinScsi: SCSILinuxAbortCommands:1843: Failed, Driver ib_srp, for vmhba_mlx4_0.1.1 2015-06-23T16:15:06.959Z cpu21:33475)<7>ib_srp: TGT[vmhba_mlx4_0.1.1 :: 1] Sending DREQ cm_id 0x410b54a06d00 qp_num 0x55 2015-06-23T16:15:06.960Z cpu25:36234)NMP: nmp_ThrottleLogForDevice:2322: Cmd 0x12 (0x4136c4245a40, 0) to dev "naa.600144f09084e251000052919b340002" on path "vmhba_mlx4_0.1.1:C0:T1:L6 " Failed: H:0x8 D:0x0 P:0x0 Possible sense data: 0x0 0x0 0x0. Act:EVAL Of particular interest is the 0x28 SCSI device message returned by the SAN: VMK_SCSI_DEVICE_QUEUE_FULL (TASK SET FULL) = 0x28 This status is returned when the LUN prevents accepting SCSI commands from initiators due to lack of resources, namely the queue depth on the array. (outlined in this KB: http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=1030381) The engineer said that when under load the SAN cannot service the requests, and gets into this state. He asked me to check with the storage vendor (hah!) and check with corresponding timestamps to see what is happening. What doesn't make sense to me is that it got into this state at around 3am (that's when the performance dropped, when no other load other than backup was running, and it had been running for 6+ hours at that point), and then magically recovered at around noon (9+ hours). It's isolated to the host because other VMs experience the same thing (although you can't really see it on esxtop until you do sequential reads/writes). I've lowered the max concurrent tasks for the Veeam proxy (from 4 to 3) for now, and will try to change the transport mode as suggested by the Veeam tech (to see if it's a hot-add thing). Is there a tunable I can check/change, or some other troubleshooting I can do? Or is this one of those, "Oh, it's COMSTAR.. YMMV" answers? :) What is strange is that after updating vSphere my initial "big" job ran, whereas previously it would fail after so many hours consistently. Is it a driver thing (running Mellanox 1.8.2, latest that supports SRP)? -------------- next part -------------- An HTML attachment was scrubbed... URL: From vab at bb-c.de Thu Jun 25 07:26:26 2015 From: vab at bb-c.de (Volker A. Brandt) Date: Thu, 25 Jun 2015 09:26:26 +0200 Subject: [OmniOS-discuss] Problem updating ms-omniti local repo In-Reply-To: <208911579.53641.1435078452952.JavaMail.root@de001cs001.trixter.de> References: <21889.33022.997158.188273@glaurung.bb-c.de> <208911579.53641.1435078452952.JavaMail.root@de001cs001.trixter.de> Message-ID: <21899.44322.2195.481653@glaurung.bb-c.de> Hello Javier! Hi Javier! I have a quick and dirty workaround for you now. I did this manually as a proof of concept, but it can be scripted easily. Step 1: Create a temporary directory to hold a "raw" copy of the repository content. You can delete the directory after you are finished. Then, create the repository as you showed in your previous mails. In the following example, I use /RAW for the temp directory, and /REPO for the repository. Step 2: Get the raw repository contents (I use just the latest, and only packages with the "omniti/" prefix in their name): # /usr/bin/pkgrecv -m latest -s http://pkg.omniti.com/omniti-ms/ -d /RAW --raw 'pkg://ms.omniti.com/omniti/*' This will give you a directory tree with file objects and metadata. Here is the directory for pkg://ms.omniti.com/omniti/viewer/xpdf as an example: # cd /RAW ; ls -golFR omniti%2Fviewer%2Fxpdf/3.4%2C5.11-0.151006%3A20141016T190442Z omniti%2Fviewer%2Fxpdf/3.4%2C5.11-0.151006%3A20141016T190442Z: total 20369 -rw-r--r-- 1 9348 Jun 25 08:45 09c7c05f1c49e6bc2aeea6355ca6e45243cfe21c -rw-r--r-- 1 1882480 Jun 25 08:45 0c966a51f4d9c3e2e3afb984cda715891f1414ac [...] -rw-r--r-- 1 1768656 Jun 25 08:45 fd77184cdab518416fe71a920575a36d363c65f8 -rw-r--r-- 1 7486 Jun 25 08:44 manifest -rw-r--r-- 1 433 Jun 25 08:44 manifest.depend -rw-r--r-- 1 514 Jun 25 08:44 manifest.dir -rw-r--r-- 1 248 Jun 25 08:44 manifest.dircache -rw-r--r-- 1 6212 Jun 25 08:44 manifest.file -rw-r--r-- 1 0 Jun 25 08:44 manifest.mediatorcache -rw-r--r-- 1 446 Jun 25 08:44 manifest.set Step 3: Create the missing subdirectories in the repository. When you are finished, it should look like this: # cd /REPO/publisher/ms.omniti.com ; ls -golF total 366 drwxr-xr-x 2 6 Jun 25 08:38 catalog/ drwxr-xr-x 258 258 Jun 25 08:25 file/ drwxr-xr-x 2 38 Jun 25 08:40 index/ drwxr-xr-x 807 807 Jun 25 08:34 pkg/ -rw-r--r-- 1 171 Jun 25 08:24 pub.p5i drwxr-xr-x 2 3 Jun 25 08:38 tmp/ Step 4: Create the hash directories in the "file" directory: # cd file # mkdir 00 01 ... fe ff Look at an existing repository with data in it to see what it should look like. Step 5: Move the file objects from /RAW to /REPO into the correct "file/XX" directory: For example (really one line for each file): # mv /RAW/omniti%2Fviewer%2Fxpdf/3.4%2C5.11-0.151006%3A20141016T190442Z/09c7c05f1c49e6bc2aeea6355ca6e45243cfe21c \ /REPO/publisher/ms.omniti.com/file/09/ # mv /RAW/omniti%2Fviewer%2Fxpdf/3.4%2C5.11-0.151006%3A20141016T190442Z/0c966a51f4d9c3e2e3afb984cda715891f1414ac \ /REPO/publisher/ms.omniti.com/file/0c/ Do this for all file objects in all package subdirectories. I love shell loops. :-) Step 6: Copy the manifest of each package into a subdirectory under "pkg" with the package version as the file name: # mkdir /REPO/publisher/ms.omniti.com/pkg/omniti%2Fviewer%2Fxpdf # mv /RAW/omniti%2Fviewer%2Fxpdf/3.4%2C5.11-0.151006%3A20141016T190442Z/manifest \ omniti%2Fviewer%2Fxpdf/3.4%2C5.11-0.151006%3A20141016T190442Z Do this for all manifests in all package subdirectories. Ignore the other "manifest.*" files in the package subdirectory. Step 7: Rebuild the repository catalog, index, etc.: # pkgrepo rebuild -s /REPO You can now use /REPO as an origin for the ms.omniti.com publisher. Regards -- Volker -- ------------------------------------------------------------------------ Volker A. Brandt Consulting and Support for Oracle Solaris Brandt & Brandt Computer GmbH WWW: http://www.bb-c.de/ Am Wiesenpfad 6, 53340 Meckenheim, GERMANY Email: vab at bb-c.de Handelsregister: Amtsgericht Bonn, HRB 10513 Schuhgr??e: 46 Gesch?ftsf?hrer: Rainer J.H. Brandt und Volker A. Brandt "When logic and proportion have fallen sloppy dead" From mark0x01 at gmail.com Thu Jun 25 08:11:49 2015 From: mark0x01 at gmail.com (Mark) Date: Thu, 25 Jun 2015 20:11:49 +1200 Subject: [OmniOS-discuss] snmp fm extension issue Message-ID: <558BB7C5.7050905@gmail.com> I have encountered an issue when including the fmd extension into the snmp config. I have older openindiana systems running this configuration (all 64 bit installs) and have not had a problem. This is the only x86 system I have - not sure why it picked x86, as it should run 64bit - although the hardware is about four years old- so it might be specific to the x86 flavour. An snmpwalk will eventually timeout when the oid reaches the extension range. dlmod sunFM /usr/lib/fm/libfmd_snmp.so.1 SunOS nzakl3nas05 5.11 omnios-a708424 i86pc i386 i86pc Is this a known problem ? Mark. From paladinemishakal at gmail.com Thu Jun 25 13:11:49 2015 From: paladinemishakal at gmail.com (Lawrence Giam) Date: Thu, 25 Jun 2015 21:11:49 +0800 Subject: [OmniOS-discuss] DFS Replication with CIFS shares Message-ID: Hi All, I am looking at setting up DFS replication with CIFS shares where I have a Windows DFS namespace and I have multiple target folder, is it possible to get DFS Replication working? Googling on the web shows results that say the DFS target have to be windows filesystem https://social.technet.microsoft.com/Forums/windowsserver/en-US/e37d0a9a-1759-487c-a560-b5ba60220892/dfs-replication-wont-work-with-2008-dfs-front-end-and-2-cifs-shares-on-netapp?forum=winserverfiles So my setup is that the SAN/NAS is running OmniOS that provides the CIFS to the Windows server and with that I have setup DFS namespace with a single target folder. When I tried to add another target folder to the DFS namespace,the system prompted me if I want to create a replication group. Proceeding forward, I got the error saying "The member is running a down level operating system". Any help? Thanks & Regards. -------------- next part -------------- An HTML attachment was scrubbed... URL: From danmcd at omniti.com Thu Jun 25 15:03:49 2015 From: danmcd at omniti.com (Dan McDonald) Date: Thu, 25 Jun 2015 11:03:49 -0400 Subject: [OmniOS-discuss] snmp fm extension issue In-Reply-To: <558BB7C5.7050905@gmail.com> References: <558BB7C5.7050905@gmail.com> Message-ID: > On Jun 25, 2015, at 4:11 AM, Mark wrote: > > I have encountered an issue when including the fmd extension into the snmp config. > I have older openindiana systems running this configuration (all 64 bit installs) and have not had a problem. > > This is the only x86 system I have - not sure why it picked x86, as it should run 64bit - although the hardware is about four years old- so it might be specific to the x86 flavour. > > An snmpwalk will eventually timeout when the oid reaches the extension range. Grab the pstack(1) of the process that's timing out. > dlmod sunFM /usr/lib/fm/libfmd_snmp.so.1 That's the 32-bit process library: bloody(~/ws/mdi-patch)[1]% ls /usr/lib/fm/amd64/*snmp* /usr/lib/fm/amd64/libfmd_snmp.so@ /usr/lib/fm/amd64/llib-lfmd_snmp.ln /usr/lib/fm/amd64/libfmd_snmp.so.1* bloody(~/ws/mdi-patch)[0]% > SunOS nzakl3nas05 5.11 omnios-a708424 i86pc i386 i86pc > > Is this a known problem ? This is the first I've heard about it, but I don't use SNMP/libsnmp with FMA. Dan From jim.oltman at gmail.com Thu Jun 25 17:18:40 2015 From: jim.oltman at gmail.com (Jim Oltman) Date: Thu, 25 Jun 2015 11:18:40 -0600 Subject: [OmniOS-discuss] DFS Replication with CIFS shares In-Reply-To: References: Message-ID: On Thu, Jun 25, 2015 at 7:11 AM, Lawrence Giam wrote: > Hi All, > > I am looking at setting up DFS replication with CIFS shares where I have a > Windows DFS namespace and I have multiple target folder, is it possible to > get DFS Replication working? > > Googling on the web shows results that say the DFS target have to be > windows filesystem > > https://social.technet.microsoft.com/Forums/windowsserver/en-US/e37d0a9a-1759-487c-a560-b5ba60220892/dfs-replication-wont-work-with-2008-dfs-front-end-and-2-cifs-shares-on-netapp?forum=winserverfiles > > So my setup is that the SAN/NAS is running OmniOS that provides the CIFS > to the Windows server and with that I have setup DFS namespace with a > single target folder. When I tried to add another target folder to the DFS > namespace,the system prompted me if I want to create a replication group. > Proceeding forward, I got the error saying "The member is running a down > level operating system". > > Any help? > > Thanks & Regards. > When I was faced with this, I chose to share out zVols via iSCSI to the Windows machine. It made DFS so much simpler. Jim -------------- next part -------------- An HTML attachment was scrubbed... URL: From paladinemishakal at gmail.com Fri Jun 26 02:10:50 2015 From: paladinemishakal at gmail.com (Lawrence Giam) Date: Fri, 26 Jun 2015 10:10:50 +0800 Subject: [OmniOS-discuss] DFS Replication with CIFS shares In-Reply-To: References: Message-ID: Hi Jim, Can you explain more on this? You share out the zVols via iSCSI to the Windows machine, so on the windows machine, you attach that zVols as a disk? But in my situation, I have alot of smb shares serve from the NAS and doing it your way, would I be then face with drive letter limitation? Regards. On Fri, Jun 26, 2015 at 1:18 AM, Jim Oltman wrote: > On Thu, Jun 25, 2015 at 7:11 AM, Lawrence Giam > wrote: > >> Hi All, >> >> I am looking at setting up DFS replication with CIFS shares where I have >> a Windows DFS namespace and I have multiple target folder, is it possible >> to get DFS Replication working? >> >> Googling on the web shows results that say the DFS target have to be >> windows filesystem >> >> https://social.technet.microsoft.com/Forums/windowsserver/en-US/e37d0a9a-1759-487c-a560-b5ba60220892/dfs-replication-wont-work-with-2008-dfs-front-end-and-2-cifs-shares-on-netapp?forum=winserverfiles >> >> So my setup is that the SAN/NAS is running OmniOS that provides the CIFS >> to the Windows server and with that I have setup DFS namespace with a >> single target folder. When I tried to add another target folder to the DFS >> namespace,the system prompted me if I want to create a replication group. >> Proceeding forward, I got the error saying "The member is running a down >> level operating system". >> >> Any help? >> >> Thanks & Regards. >> > > When I was faced with this, I chose to share out zVols via iSCSI to the > Windows machine. It made DFS so much simpler. > > Jim > -------------- next part -------------- An HTML attachment was scrubbed... URL: From graham at thestephensdomain.com Fri Jun 26 09:18:53 2015 From: graham at thestephensdomain.com (Graham Stephens) Date: Fri, 26 Jun 2015 10:18:53 +0100 Subject: [OmniOS-discuss] KVM vnics not starting Message-ID: <558D18FD.20103@thestephensdomain.com> Guys, Does anyone else here have problems trying to start a kvm due to vnics not initializing properly? It seems to me that the more vnics you add, the more likely the problem is to occur. The firewall that I'm trying to build currently has four vnics attached, and I'm intending to add more before I'm finished - but frequently when I try to start it one will fail to 'catch' (like starting an old car) and kvm fails. It's usually the third or fourth vnic, but not always. I have to keep retrying the start script, and eventually it will work but it can take five minutes for the darn thing to come up. Is there a setting I have missed? From johan.kragsterman at capvert.se Fri Jun 26 13:37:59 2015 From: johan.kragsterman at capvert.se (Johan Kragsterman) Date: Fri, 26 Jun 2015 15:37:59 +0200 Subject: [OmniOS-discuss] Ang: KVM vnics not starting In-Reply-To: <558D18FD.20103@thestephensdomain.com> References: <558D18FD.20103@thestephensdomain.com> Message-ID: Hi! -----"OmniOS-discuss" skrev: ----- Till: omnios-discuss at lists.omniti.com Fr?n: Graham Stephens S?nt av: "OmniOS-discuss" Datum: 2015-06-26 11:20 ?rende: [OmniOS-discuss] KVM vnics not starting Guys, Does anyone else here have problems trying to start a kvm due to vnics not initializing properly? It seems to me that the more vnics you add, the more likely the problem is to occur. The firewall that I'm trying to build currently has four vnics attached, and I'm intending to add more before I'm finished - but frequently when I try to start it one will fail to 'catch' (like starting an old car) and kvm fails. It's usually the third or fourth vnic, but not always. I have to keep retrying the start script, and eventually it will work but it can take five minutes for the darn thing to come up. Is there a setting I have missed? I've never seen this! I have used VM firewalls/routers with 5 vnic's or more, never seen it. Johan _______________________________________________ OmniOS-discuss mailing list OmniOS-discuss at lists.omniti.com http://lists.omniti.com/mailman/listinfo/omnios-discuss From graham at thestephensdomain.com Fri Jun 26 14:24:09 2015 From: graham at thestephensdomain.com (Graham Stephens) Date: Fri, 26 Jun 2015 15:24:09 +0100 Subject: [OmniOS-discuss] Ang: KVM vnics not starting In-Reply-To: References: <558D18FD.20103@thestephensdomain.com> Message-ID: <558D6089.9030307@thestephensdomain.com> On 26/06/2015 14:37, Johan Kragsterman wrote: > Hi! > > > -----"OmniOS-discuss" skrev: ----- > Till: omnios-discuss at lists.omniti.com > Fr?n: Graham Stephens > S?nt av: "OmniOS-discuss" > Datum: 2015-06-26 11:20 > ?rende: [OmniOS-discuss] KVM vnics not starting > > Guys, > > Does anyone else here have problems trying to start a kvm due to vnics > not initializing properly? It seems to me that the more vnics you add, > the more likely the problem is to occur. > The firewall that I'm trying to build currently has four vnics attached, > and I'm intending to add more before I'm finished - but frequently when > I try to start it one will fail to 'catch' (like starting an old car) > and kvm fails. It's usually the third or fourth vnic, but not always. I > have to keep retrying the start script, and eventually it will work but > it can take five minutes for the darn thing to come up. > > Is there a setting I have missed? > > > > I've never seen this! > > I have used VM firewalls/routers with 5 vnic's or more, never seen it. > > Johan > Hi Johan, Can I ask some questions (speculations)? Are you running the VMs from inside zones? Are they attached to physical nics or virtual ones? Are you using hyperthreaded CPUs? Perhaps one of these might make a difference. Graham. From johan.kragsterman at capvert.se Fri Jun 26 14:30:56 2015 From: johan.kragsterman at capvert.se (Johan Kragsterman) Date: Fri, 26 Jun 2015 16:30:56 +0200 Subject: [OmniOS-discuss] Ang: Re: Ang: KVM vnics not starting In-Reply-To: <558D6089.9030307@thestephensdomain.com> References: <558D6089.9030307@thestephensdomain.com>, <558D18FD.20103@thestephensdomain.com> Message-ID: Hi! -----Graham Stephens skrev: ----- Till: Johan Kragsterman Fr?n: Graham Stephens Datum: 2015-06-26 16:24 Kopia: omnios-discuss at lists.omniti.com ?rende: Re: Ang: [OmniOS-discuss] KVM vnics not starting On 26/06/2015 14:37, Johan Kragsterman wrote: > Hi! > > > -----"OmniOS-discuss" skrev: ----- > Till: omnios-discuss at lists.omniti.com > Fr?n: Graham Stephens > S?nt av: "OmniOS-discuss" > Datum: 2015-06-26 11:20 > ?rende: [OmniOS-discuss] KVM vnics not starting > > Guys, > > Does anyone else here have problems trying to start a kvm due to vnics > not initializing properly? It seems to me that the more vnics you add, > the more likely the problem is to occur. > The firewall that I'm trying to build currently has four vnics attached, > and I'm intending to add more before I'm finished - but frequently when > I try to start it one will fail to 'catch' (like starting an old car) > and kvm fails. It's usually the third or fourth vnic, but not always. I > have to keep retrying the start script, and eventually it will work but > it can take five minutes for the darn thing to come up. > > Is there a setting I have missed? > > > > I've never seen this! > > I have used VM firewalls/routers with 5 vnic's or more, never seen it. > > Johan > Hi Johan, Can I ask some questions (speculations)? Are you running the VMs from inside zones? No VM's inside zone's, so far.... Are they attached to physical nics or virtual ones? That I don't understand, VM's needs virtual nics? Perhaps you mean vnic attached to etherstubs? I use both sorts, both to etherstubs and to phys nics. Are you using hyperthreaded CPUs? Yes, I use HT. Johan Perhaps one of these might make a difference. Graham. From bmx1955 at gmail.com Fri Jun 26 14:59:19 2015 From: bmx1955 at gmail.com (Mick Burns) Date: Fri, 26 Jun 2015 10:59:19 -0400 Subject: [OmniOS-discuss] OmniOS considered in-memory OS ? Message-ID: As per current wikipedia entry for SmartOS, we can read the following : "SmartOS is an in-memory operating system and boots directly into random access memory. It supports various boot mechanisms such as booting off of universal serial bus thumbdrive, ISO Image, or over the network via PXE boot." Does this statement applies to OmniOS as well ? Ultimately, what I need to know is if OmniOS is good fit to run OS off a pair of either internal SDcards or good quality USB sticks ? thank you. From graham at thestephensdomain.com Fri Jun 26 15:07:51 2015 From: graham at thestephensdomain.com (Graham Stephens) Date: Fri, 26 Jun 2015 16:07:51 +0100 Subject: [OmniOS-discuss] Ang: KVM vnics not starting In-Reply-To: References: <558D6089.9030307@thestephensdomain.com>, <558D18FD.20103@thestephensdomain.com> Message-ID: <558D6AC7.4000604@thestephensdomain.com> On 26/06/2015 15:30, Johan Kragsterman wrote: > >> -----"OmniOS-discuss" skrev: ----- >> Till: omnios-discuss at lists.omniti.com >> Fr?n: Graham Stephens >> S?nt av: "OmniOS-discuss" >> Datum: 2015-06-26 11:20 >> ?rende: [OmniOS-discuss] KVM vnics not starting >> >> Guys, >> >> Does anyone else here have problems trying to start a kvm due to vnics >> not initializing properly? It seems to me that the more vnics you add, >> the more likely the problem is to occur. >> The firewall that I'm trying to build currently has four vnics attached, >> and I'm intending to add more before I'm finished - but frequently when >> I try to start it one will fail to 'catch' (like starting an old car) >> and kvm fails. It's usually the third or fourth vnic, but not always. I >> have to keep retrying the start script, and eventually it will work but >> it can take five minutes for the darn thing to come up. >> >> Is there a setting I have missed? >> >> >> >> I've never seen this! >> >> I have used VM firewalls/routers with 5 vnic's or more, never seen it. >> >> Johan >> > > Hi Johan, > > Can I ask some questions (speculations)? > > Are you running the VMs from inside zones? > > > > No VM's inside zone's, so far.... > Maybe that's where my problems lie - but I don't know what to do about it :( > > Are they attached to physical nics or virtual ones? > > > That I don't understand, VM's needs virtual nics? Perhaps you mean vnic attached to etherstubs? I use both sorts, both to etherstubs and to phys nics. Sorry, yes, I meant the vnics virtually attached or physically, as you said. > > Johan > From danmcd at omniti.com Fri Jun 26 15:15:00 2015 From: danmcd at omniti.com (Dan McDonald) Date: Fri, 26 Jun 2015 11:15:00 -0400 Subject: [OmniOS-discuss] OmniOS considered in-memory OS ? In-Reply-To: References: Message-ID: > On Jun 26, 2015, at 10:59 AM, Mick Burns wrote: > > As per current wikipedia entry for SmartOS, we can read the following : > > "SmartOS is an in-memory operating system and boots directly into > random access memory. It supports various boot mechanisms such as > booting off of universal serial bus thumbdrive, ISO Image, or over the > network via PXE boot." > > Does this statement applies to OmniOS as well ? No. We boot off of traditional media. > Ultimately, what I need to know is if OmniOS is good fit to run OS off > a pair of either internal SDcards or good quality USB sticks ? As shipped, no. Dan From john.barfield at bissinc.com Fri Jun 26 15:47:23 2015 From: john.barfield at bissinc.com (John Barfield) Date: Fri, 26 Jun 2015 15:47:23 +0000 Subject: [OmniOS-discuss] OmniOS considered in-memory OS ? In-Reply-To: References: Message-ID: I?ve been interested in configuring omnios to run in memory off of a ram disk myself. Does anyone know where you could find a good guide for booting Solaris(h) kernel into memory with a ramdisk? John Barfield / Sr Principal Engineer +1 (214) 425-0783/ john.barfield at bissinc.com BISS, Inc. Office: +1 (214) 506-8354 4925 Greenville Ave Suite 900 Dallas, TX 75206 support.bissinc.com On 6/26/15, 10:15 AM, "Dan McDonald" wrote: > >> On Jun 26, 2015, at 10:59 AM, Mick Burns wrote: >> >> As per current wikipedia entry for SmartOS, we can read the following : >> >> "SmartOS is an in-memory operating system and boots directly into >> random access memory. It supports various boot mechanisms such as >> booting off of universal serial bus thumbdrive, ISO Image, or over the >> network via PXE boot." >> >> Does this statement applies to OmniOS as well ? > >No. We boot off of traditional media. > >> Ultimately, what I need to know is if OmniOS is good fit to run OS off >> a pair of either internal SDcards or good quality USB sticks ? > >As shipped, no. > >Dan > >_______________________________________________ >OmniOS-discuss mailing list >OmniOS-discuss at lists.omniti.com >http://lists.omniti.com/mailman/listinfo/omnios-discuss From alka at hfg-gmuend.de Fri Jun 26 18:12:07 2015 From: alka at hfg-gmuend.de (=?utf-8?Q?G=C3=BCnther_Alka?=) Date: Fri, 26 Jun 2015 20:12:07 +0200 Subject: [OmniOS-discuss] OmniOS considered in-memory OS ? In-Reply-To: References: Message-ID: <18F25612-AE74-4625-9966-0E2B16649646@hfg-gmuend.de> booting from a good pair of 32 GB USB sticks is possible. (Turn off atime) But I would prefer a pair of 32 GB DOMs like http://www.supermicro.com/products/nfo/SATADOM.cfm > Am 26.06.2015 um 17:47 schrieb John Barfield : > > I?ve been interested in configuring omnios to run in memory off of a ram > disk myself. > > Does anyone know where you could find a good guide for booting Solaris(h) > kernel into memory with a ramdisk? > > > > > John Barfield / Sr Principal Engineer > +1 (214) 425-0783/ john.barfield at bissinc.com > BISS, Inc. Office: +1 (214) 506-8354 > > 4925 Greenville Ave Suite 900 > Dallas, TX 75206 > support.bissinc.com > > > > > > > On 6/26/15, 10:15 AM, "Dan McDonald" wrote: > >> >>> On Jun 26, 2015, at 10:59 AM, Mick Burns wrote: >>> >>> As per current wikipedia entry for SmartOS, we can read the following : >>> >>> "SmartOS is an in-memory operating system and boots directly into >>> random access memory. It supports various boot mechanisms such as >>> booting off of universal serial bus thumbdrive, ISO Image, or over the >>> network via PXE boot." >>> >>> Does this statement applies to OmniOS as well ? >> >> No. We boot off of traditional media. >> >>> Ultimately, what I need to know is if OmniOS is good fit to run OS off >>> a pair of either internal SDcards or good quality USB sticks ? >> >> As shipped, no. >> >> Dan >> >> _______________________________________________ >> OmniOS-discuss mailing list >> OmniOS-discuss at lists.omniti.com >> http://lists.omniti.com/mailman/listinfo/omnios-discuss > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss -------------- next part -------------- An HTML attachment was scrubbed... URL: From jdg117 at elvis.arl.psu.edu Fri Jun 26 18:42:51 2015 From: jdg117 at elvis.arl.psu.edu (John D Groenveld) Date: Fri, 26 Jun 2015 14:42:51 -0400 Subject: [OmniOS-discuss] KVM vnics not starting In-Reply-To: Your message of "Fri, 26 Jun 2015 10:18:53 BST." <558D18FD.20103@thestephensdomain.com> References: <558D18FD.20103@thestephensdomain.com> Message-ID: <201506261842.t5QIgpQC006523@elvis.arl.psu.edu> In message <558D18FD.20103 at thestephensdomain.com>, Graham Stephens writes: >Does anyone else here have problems trying to start a kvm due to vnics >not initializing properly? It seems to me that the more vnics you add, >the more likely the problem is to occur. No. What's your reproducable configuration? What version OmniOS? How many VMs and VNICs? What's your KVM invocations? John groenveld at acm.org From jim.oltman at gmail.com Fri Jun 26 20:59:46 2015 From: jim.oltman at gmail.com (Jim Oltman) Date: Fri, 26 Jun 2015 14:59:46 -0600 Subject: [OmniOS-discuss] DFS Replication with CIFS shares In-Reply-To: References: Message-ID: On Thu, Jun 25, 2015 at 8:10 PM, Lawrence Giam wrote: > Hi Jim, > > Can you explain more on this? You share out the zVols via iSCSI to the > Windows machine, so on the windows machine, you attach that zVols as a disk? > > But in my situation, I have alot of smb shares serve from the NAS and > doing it your way, would I be then face with drive letter limitation? > > Regards. > I would create one or 2 large zVols, share them via iSCSI to your Windows machines. Then I'd format those as NTFS. You can then create folders in those disk LUNs. Create something like this: D:\shares\share1 D:\shares\share2 D:\shares\etc E:\shares\share3 etc Then, CIFS share each one of the "shareX" folders. You can then tell DFS which shares on your server to use for a DFS namespace: \\server\share1 \\server\share2 \\server\share3 Point DFS at those CIFS shares. -------------- next part -------------- An HTML attachment was scrubbed... URL: From danmcd at omniti.com Fri Jun 26 21:01:40 2015 From: danmcd at omniti.com (Dan McDonald) Date: Fri, 26 Jun 2015 17:01:40 -0400 Subject: [OmniOS-discuss] Say hello to very-larval i40e... Message-ID: <75353284-14DD-47C9-9BCD-E8DC41D32F7C@omniti.com> Just so y'all know, this is the very earliest intro to the i40e driver (Intel 40Gig Ethernet, aka. Fortville, aka. the XL710 board). There are LOTS of KEBE comments (my work-in-progress markers) with more to come, there are TONS of missing pieces of functionality, and I haven't even tried attaching it to a real piece of hardware yet. I *do* have real hardware, however, thanks to Intel - a pair of XL710s that I will connect back-to-back for testing. This project started with the "ixl" driver from FreeBSD common-from-Intel code, and I will be attempting to wrap a GLDv3 driver around this common code, much like ixgbe(7D) is. In fact, the initial directions I'm going are largely derived from ixgbe(7D). I have the very large programmer's reference manual (which you can download from Intel for free) up for when I get that far, as well as the ixl code from FreeBSD. I figure developing this in the open would help keep me honest, as well as allow anyone in the community to inspect, or even help, with things. (I've Bcc:ed the very helpful contact from Intel, so Intel knows things as well.) Here's the very larval webrev: http://kebe.com/~danmcd/webrevs/i40e/ I could turn this into an actual git repo if I wanted, and there's enough interest. Currently I'm working in a child of illumos-omnios, but there's nothing here that won't just drop into illumos-gate, or your favorite other child of illumos-gate (modulo perhaps packaging, but the .mf is already in place). This won't happen overnight - but I hope it will happen relatively quickly. FYI, Dan From bfriesen at simple.dallas.tx.us Sat Jun 27 02:23:23 2015 From: bfriesen at simple.dallas.tx.us (Bob Friesenhahn) Date: Fri, 26 Jun 2015 21:23:23 -0500 (CDT) Subject: [OmniOS-discuss] OmniOS considered in-memory OS ? In-Reply-To: <18F25612-AE74-4625-9966-0E2B16649646@hfg-gmuend.de> References: <18F25612-AE74-4625-9966-0E2B16649646@hfg-gmuend.de> Message-ID: On Fri, 26 Jun 2015, G?nther Alka wrote: > booting from a good pair of 32 GB USB sticks is possible. > (Turn off atime) > > But I would prefer a pair of 32 GB DOMs like > http://www.supermicro.com/products/nfo/SATADOM.cfm The next big thing seems to be M.2 PCIe cards, which are a whole lot faster. Hopefully they will work with OmniOS. I have two Samsung SM951s on the way for use as boot drives (over-kill) ... Bob -- Bob Friesenhahn bfriesen at simple.dallas.tx.us, http://www.simplesystems.org/users/bfriesen/ GraphicsMagick Maintainer, http://www.GraphicsMagick.org/ From graham at thestephensdomain.com Sat Jun 27 11:02:44 2015 From: graham at thestephensdomain.com (Graham Stephens) Date: Sat, 27 Jun 2015 12:02:44 +0100 Subject: [OmniOS-discuss] KVM vnics not starting In-Reply-To: <201506261842.t5QIgpQC006523@elvis.arl.psu.edu> References: <558D18FD.20103@thestephensdomain.com> <201506261842.t5QIgpQC006523@elvis.arl.psu.edu> Message-ID: <558E82D4.5050203@thestephensdomain.com> On 26/06/2015 19:42, John D Groenveld wrote: > In message <558D18FD.20103 at thestephensdomain.com>, Graham Stephens writes: >> Does anyone else here have problems trying to start a kvm due to vnics >> not initializing properly? It seems to me that the more vnics you add, >> the more likely the problem is to occur. > > No. > > What's your reproducable configuration? > What version OmniOS? > How many VMs and VNICs? > What's your KVM invocations? > > John > groenveld at acm.org Hi John. I'm on 151014 2cac697. There are a couple of other zones defined but not running at the time. The problem happens with just one zone running and the kvm inside. In the zone: show-link: LINK CLASS MTU STATE BRIDGE OVER fw2z0 vnic 9000 up -- ? fwo0 vnic 1500 up -- ? fwl0 vnic 1500 down -- ? fwl1 vnic 1500 up -- ? fwl2 vnic 1500 down -- ? fwl3 vnic 1500 down -- ? fwo0 is attached to an rge0 fwl0-3 are attached to a four-port e1000g fw2zo is attached to an etherstub. show-addr: ADDROBJ TYPE STATE ADDR lo0/v4 static ok 127.0.0.1/8 fwl0/temp static inaccessible 10.0.0.1/24 lo0/v6 static ok ::1/128 kvm startup script: #!/usr/bin/bash # configuration VNICW0=fwo0 VNICL1=fwl1 VNICL2=fwl2 VNICL3=fwl3 # Sample zvol path. HDD=/dev/zvol/dsk/rpool/zones/eppifw01/kvm CD=/root/OpenBSD_x64_57.iso VNC=5 # Memory for the KVM instance, in Megabytes (2^20 bytes). MEM=1024 macw0=`dladm show-vnic -po macaddress $VNICW0` macl1=`dladm show-vnic -po macaddress $VNICL1` macl2=`dladm show-vnic -po macaddress $VNICL2` macl3=`dladm show-vnic -po macaddress $VNICL3` /usr/bin/qemu-system-x86_64 \ -name "$(basename $CD)" \ -boot cd \ -enable-kvm \ -vnc 10.0.0.1:$VNC -k en-gb \ -smp 2 \ -m $MEM \ -no-hpet \ -localtime \ -drive file=$HDD,if=ide,index=0 \ -drive file=$CD,media=cdrom,if=ide,index=2 \ -net nic,vlan=0,name=net0,model=e1000,macaddr=$macw0 \ -net vnic,vlan=0,name=net0,ifname=$VNICW0 \ -net nic,vlan=1,name=net0,model=e1000,macaddr=$macl1 \ -net vnic,vlan=1,name=net0,ifname=$VNICL1 \ -net nic,vlan=2,name=net0,model=e1000,macaddr=$macl2 \ -net vnic,vlan=2,name=net0,ifname=$VNICL2 \ -net nic,vlan=3,name=net0,model=e1000,macaddr=$macl3 \ -net vnic,vlan=3,name=net0,ifname=$VNICL3 \ -vga std \ -daemonize if [ $? -gt 0 ]; then echo "Failed to start VM" fi # UDP port for VNC connections to the KVM instance. 5900 is added in the command. port=`expr 5900 + $VNC` public_nic=$(dladm show-vnic|grep fwl0|awk '{print $2}') public_ip=$(ifconfig fwl0|grep inet|awk '{print $2}') #public_ip=$(ifconfig $public_nic|grep inet|awk '{print $2}') echo "Started VM:" echo "Public: ${public_ip}:${port}" --- I took the macaddr part off the vnic description, because at the time it seemed to help reduce the issue - but no longer. The usual error message I get from one of the vnics is "failed to bind interface xxxx". ...and please don't assume I know what I'm doing - that's probably not the case. ;) From graham at thestephensdomain.com Sat Jun 27 11:08:46 2015 From: graham at thestephensdomain.com (Graham Stephens) Date: Sat, 27 Jun 2015 12:08:46 +0100 Subject: [OmniOS-discuss] KVM vnics not starting In-Reply-To: <201506261842.t5QIgpQC006523@elvis.arl.psu.edu> References: <558D18FD.20103@thestephensdomain.com> <201506261842.t5QIgpQC006523@elvis.arl.psu.edu> Message-ID: <558E843E.4070108@thestephensdomain.com> On 26/06/2015 19:42, John D Groenveld wrote: > In message <558D18FD.20103 at thestephensdomain.com>, Graham Stephens writes: >> Does anyone else here have problems trying to start a kvm due to vnics >> not initializing properly? It seems to me that the more vnics you add, >> the more likely the problem is to occur. > > No. > > What's your reproducable configuration? > What version OmniOS? > How many VMs and VNICs? > What's your KVM invocations? > > John > groenveld at acm.org ------------------- I also get these messages in the global log, but only AFTER the kvm has managed to start: Jun 27 09:13:19 eppigw01 genunix: [ID 420667 kern.info] kvm_lapic_reset: vcpu=ffffff04ffa6d000, id=0, base_msr= fee00100 PRIx64 base_address=fee00000 Jun 27 09:13:19 eppigw01 genunix: [ID 710719 kern.info] vmcs revision_id = f Jun 27 09:13:19 eppigw01 genunix: [ID 420667 kern.info] kvm_lapic_reset: vcpu=ffffff04ffa65000, id=1, base_msr= fee00000 PRIx64 base_address=fee00000 Jun 27 09:13:19 eppigw01 genunix: [ID 710719 kern.info] vmcs revision_id = f Jun 27 09:13:21 eppigw01 genunix: [ID 391722 kern.info] unhandled wrmsr: 0xfed8c25b data 53 Jun 27 09:13:21 eppigw01 last message repeated 1 time Jun 27 09:13:21 eppigw01 genunix: [ID 391722 kern.info] unhandled wrmsr: 0xff3fd4a0 data fffffd7fffdfea70 Jun 27 09:13:21 eppigw01 last message repeated 1 time Jun 27 09:13:21 eppigw01 genunix: [ID 391722 kern.info] unhandled wrmsr: 0xffdff078 data fffffd7fffdfe900 Jun 27 09:13:21 eppigw01 last message repeated 1 time Jun 27 09:13:21 eppigw01 genunix: [ID 291337 kern.info] vcpu 1 received sipi with vector # 10 Jun 27 09:13:21 eppigw01 genunix: [ID 420667 kern.info] kvm_lapic_reset: vcpu=ffffff04ffa65000, id=1, base_msr= fee00800 PRIx64 base_address=fee00000 Jun 27 09:13:37 eppigw01 genunix: [ID 713435 kern.info] unhandled rdmsr: 0x756e6547 Jun 27 09:13:37 eppigw01 genunix: [ID 391722 kern.info] unhandled wrmsr: 0x0 data 0 Jun 27 09:13:44 eppigw01 genunix: [ID 291337 kern.info] vcpu 1 received sipi with vector # 10 Jun 27 09:13:44 eppigw01 genunix: [ID 420667 kern.info] kvm_lapic_reset: vcpu=ffffff04ffa65000, id=1, base_msr= fee00800 PRIx64 base_address=fee00000 Jun 27 09:13:44 eppigw01 genunix: [ID 713435 kern.info] unhandled rdmsr: 0x756e6547 Jun 27 09:13:44 eppigw01 genunix: [ID 391722 kern.info] unhandled wrmsr: 0x0 data 0 Jun 27 09:13:44 eppigw01 genunix: [ID 713435 kern.info] unhandled rdmsr: 0x756e6547 Jun 27 09:13:44 eppigw01 genunix: [ID 391722 kern.info] unhandled wrmsr: 0x0 data 0 (sorry for the email client formatting) From tobi at oetiker.ch Sat Jun 27 11:24:31 2015 From: tobi at oetiker.ch (Tobias Oetiker) Date: Sat, 27 Jun 2015 13:24:31 +0200 (CEST) Subject: [OmniOS-discuss] allocation throttle Message-ID: I am just watching OpenZFS Conference Videos. George Wilson just showed off his allocation throttle work ... is this in omnios already ? cheers tobi -- Tobi Oetiker, OETIKER+PARTNER AG, Aarweg 15 CH-4600 Olten, Switzerland www.oetiker.ch tobi at oetiker.ch +41 62 775 9902 From john.barfield at bissinc.com Sat Jun 27 14:28:23 2015 From: john.barfield at bissinc.com (John Barfield) Date: Sat, 27 Jun 2015 14:28:23 +0000 Subject: [OmniOS-discuss] OmniOS considered in-memory OS ? In-Reply-To: References: <18F25612-AE74-4625-9966-0E2B16649646@hfg-gmuend.de> Message-ID: <28CABC00-70E1-47CA-B435-79F1A094D054@bissinc.com> I?m booting off of a pair SLC USB sticks today. They?re extremely fast boot still not as fast as SmartOS running in memory. I guess I?ll do some googling to get omni running in memory :p John Barfield / Sr Principal Engineer +1 (214) 425-0783/ john.barfield at bissinc.com BISS, Inc. Office: +1 (214) 506-8354 On 6/26/15, 9:23 PM, "Bob Friesenhahn" wrote: >On Fri, 26 Jun 2015, G?nther Alka wrote: > >> booting from a good pair of 32 GB USB sticks is possible. >> (Turn off atime) >> >> But I would prefer a pair of 32 GB DOMs like >> http://www.supermicro.com/products/nfo/SATADOM.cfm > >The next big thing seems to be M.2 PCIe cards, which are a whole lot >faster. Hopefully they will work with OmniOS. I have two Samsung >SM951s on the way for use as boot drives (over-kill) ... > >Bob >-- >Bob Friesenhahn >bfriesen at simple.dallas.tx.us, http://www.simplesystems.org/users/bfriesen/ >GraphicsMagick Maintainer, http://www.GraphicsMagick.org/ From danmcd at omniti.com Sat Jun 27 15:13:56 2015 From: danmcd at omniti.com (Dan McDonald) Date: Sat, 27 Jun 2015 11:13:56 -0400 Subject: [OmniOS-discuss] allocation throttle In-Reply-To: References: Message-ID: <46056636-FB58-4867-9CA1-F4681E9F3EBA@omniti.com> > On Jun 27, 2015, at 7:24 AM, Tobias Oetiker wrote: > > I am just watching OpenZFS Conference Videos. George Wilson just > showed off his allocation throttle work ... is this in omnios > already ? If it's in illumos-gate, it's in at least OmniOS bloody. I believe this work hasn't been upstreamed yet out of DelphixOS. Dan From richard.elling at richardelling.com Sat Jun 27 18:06:19 2015 From: richard.elling at richardelling.com (Richard Elling) Date: Sat, 27 Jun 2015 11:06:19 -0700 Subject: [OmniOS-discuss] allocation throttle In-Reply-To: <46056636-FB58-4867-9CA1-F4681E9F3EBA@omniti.com> References: <46056636-FB58-4867-9CA1-F4681E9F3EBA@omniti.com> Message-ID: <04D97B85-737C-4049-B2F8-66F59741D561@RichardElling.com> it has been in for a year or so -- richard > On Jun 27, 2015, at 8:13 AM, Dan McDonald wrote: > > >> On Jun 27, 2015, at 7:24 AM, Tobias Oetiker wrote: >> >> I am just watching OpenZFS Conference Videos. George Wilson just >> showed off his allocation throttle work ... is this in omnios >> already ? > > If it's in illumos-gate, it's in at least OmniOS bloody. I believe this work hasn't been upstreamed yet out of DelphixOS. > > Dan > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss From danmcd at omniti.com Sat Jun 27 23:03:48 2015 From: danmcd at omniti.com (Dan McDonald) Date: Sat, 27 Jun 2015 19:03:48 -0400 Subject: [OmniOS-discuss] allocation throttle In-Reply-To: <04D97B85-737C-4049-B2F8-66F59741D561@RichardElling.com> References: <46056636-FB58-4867-9CA1-F4681E9F3EBA@omniti.com> <04D97B85-737C-4049-B2F8-66F59741D561@RichardElling.com> Message-ID: > On Jun 27, 2015, at 2:06 PM, Richard Elling wrote: > > it has been in for a year or so > > -- richard Eesh - which commit? I can tell you which OmniOS release it first appeared in Dan Sent from my iPhone (typos, autocorrect, and all) -------------- next part -------------- An HTML attachment was scrubbed... URL: From takashiary at gmail.com Sun Jun 28 17:05:37 2015 From: takashiary at gmail.com (takashi ary) Date: Mon, 29 Jun 2015 02:05:37 +0900 Subject: [OmniOS-discuss] ntpd -x steps clock on leap second Message-ID: Hello, OmniOS ntp version is 4.2.8p2. This version has leap second bug. Bug fix release within this month? ntp version 4.2.8p3 is not yet released.(now 4.2.8p3-RC3) Bug 2745 - ntpd -x steps clock on leap second http://bugzilla.ntp.org/show_bug.cgi?id=2745 fedora xleap.patch http://pkgs.fedoraproject.org/cgit/ntp.git/tree/ntp-4.2.6p5-xleap.patch Oracle Solaris ntp patch https://java.net/projects/solaris-userland/sources/gate/content/components/ntp/patches/70-leapslew.patch Thanks From richard.elling at richardelling.com Sun Jun 28 17:03:28 2015 From: richard.elling at richardelling.com (Richard Elling) Date: Sun, 28 Jun 2015 10:03:28 -0700 Subject: [OmniOS-discuss] OmniOS considered in-memory OS ? In-Reply-To: References: Message-ID: <24D89F84-5EC6-4CBF-8029-2510EB54B828@richardelling.com> > On Jun 26, 2015, at 8:47 AM, John Barfield wrote: > > I?ve been interested in configuring omnios to run in memory off of a ram > disk myself. > > Does anyone know where you could find a good guide for booting Solaris(h) > kernel into memory with a ramdisk? :-) the funny thing is that SunOS has been bootable into ramdisk since... forever. Back in the day, before CDs were invented, we booted from tape or (if you really hated yourself) floppies :-). Net-net, this probably isn't documented anywhere, per se, outside of the building shell scripts :-( The way it works is to produce an image of a file system. Add that to a boot loader and instruct the boot loader to put it in memory and run. The actual work is to produce the image, and this is where the build scripts comes into play. You can certainly take an OmniOS build and produce an in-memory root, however, the missing management part is handling the persistence of identity. For that, it is useful to see how SmartOS handles /var and files normally modified in /etc. -- richard From richard.elling at richardelling.com Sun Jun 28 17:23:02 2015 From: richard.elling at richardelling.com (Richard Elling) Date: Sun, 28 Jun 2015 10:23:02 -0700 Subject: [OmniOS-discuss] allocation throttle In-Reply-To: References: <46056636-FB58-4867-9CA1-F4681E9F3EBA@omniti.com> <04D97B85-737C-4049-B2F8-66F59741D561@RichardElling.com> Message-ID: > On Jun 27, 2015, at 4:03 PM, Dan McDonald wrote: > > On Jun 27, 2015, at 2:06 PM, Richard Elling > wrote: > >> it has been in for a year or so >> >> -- richard > > Eesh - which commit? I can tell you which OmniOS release it first appeared in time flies? 69962b56 Matt Ahrens, 2013-08-26 ? richard > > Dan > > Sent from my iPhone (typos, autocorrect, and all) -- Richard.Elling at RichardElling.com +1-760-896-4422 -------------- next part -------------- An HTML attachment was scrubbed... URL: From john.barfield at bissinc.com Sun Jun 28 21:59:35 2015 From: john.barfield at bissinc.com (John Barfield) Date: Sun, 28 Jun 2015 21:59:35 +0000 Subject: [OmniOS-discuss] Zfs encryption Message-ID: <867309FA36CA349E.8E697B94-2C54-4069-8AFB-6CAAE25252D4@mail.outlook.com> Ive done some googling around and Im not seeing much about zfs encryption support in illumos/omnios. The redmine project is there but empty. What are some of the options that others are using today? Wheres xlofi? Did Oracle take that for their zfs encryption? I was around solaris before oracle, i left, then i came back to illumos so I may have missed some details here and there. -------------- next part -------------- An HTML attachment was scrubbed... URL: From wzmalone at gmail.com Mon Jun 29 00:32:06 2015 From: wzmalone at gmail.com (Zach Malone) Date: Sun, 28 Jun 2015 18:32:06 -0600 Subject: [OmniOS-discuss] Zfs encryption In-Reply-To: <867309FA36CA349E.8E697B94-2C54-4069-8AFB-6CAAE25252D4@mail.outlook.com> References: <867309FA36CA349E.8E697B94-2C54-4069-8AFB-6CAAE25252D4@mail.outlook.com> Message-ID: ZFS encryption happened after Oracle, and last time I checked, no one had built something similar in Illumos. You can use lofiadm with encryption to build block devices, and then build your zpool out of those. That's what I did the last time I needed disk encryption. I think I referenced https://blogs.oracle.com/darren/entry/encrypting_zfs_pools_using_lofi , but it's been a while. -Zach On Sunday, June 28, 2015, John Barfield wrote: > Ive done some googling around and Im not seeing much about zfs > encryption support in illumos/omnios. The redmine project is there but > empty. > > What are some of the options that others are using today? > > Wheres xlofi? Did Oracle take that for their zfs encryption? I was > around solaris before oracle, i left, then i came back to illumos so I may > have missed some details here and there. > -------------- next part -------------- An HTML attachment was scrubbed... URL: From danmcd at omniti.com Mon Jun 29 14:13:26 2015 From: danmcd at omniti.com (Dan McDonald) Date: Mon, 29 Jun 2015 10:13:26 -0400 Subject: [OmniOS-discuss] ntpd -x steps clock on leap second In-Reply-To: References: Message-ID: <2E0D6939-B865-42EE-8A54-6AA293B9C0C4@omniti.com> > On Jun 28, 2015, at 1:05 PM, takashi ary wrote: > > Hello, > > OmniOS ntp version is 4.2.8p2. > This version has leap second bug. > Bug fix release within this month? > ntp version 4.2.8p3 is not yet released.(now 4.2.8p3-RC3) I was hoping p3 would be past the RC stage and actually be *released*. Given how close we are to the leap second, however, I may not be able to do that. Watch this list for an announcement of when I update NTP. I still hope sometime today p3 will jump from RC to Released. Dan From danmcd at omniti.com Mon Jun 29 14:15:18 2015 From: danmcd at omniti.com (Dan McDonald) Date: Mon, 29 Jun 2015 10:15:18 -0400 Subject: [OmniOS-discuss] allocation throttle In-Reply-To: References: <46056636-FB58-4867-9CA1-F4681E9F3EBA@omniti.com> <04D97B85-737C-4049-B2F8-66F59741D561@RichardElling.com> Message-ID: > On Jun 28, 2015, at 1:23 PM, Richard Elling wrote: > >> >> Eesh - which commit? I can tell you which OmniOS release it first appeared in > > time flies? > 69962b56 Matt Ahrens, 2013-08-26 > ? richard Which means it's been in since r151008. Thanks, Dan From bfriesen at simple.dallas.tx.us Mon Jun 29 14:57:55 2015 From: bfriesen at simple.dallas.tx.us (Bob Friesenhahn) Date: Mon, 29 Jun 2015 09:57:55 -0500 (CDT) Subject: [OmniOS-discuss] ntpd -x steps clock on leap second In-Reply-To: <2E0D6939-B865-42EE-8A54-6AA293B9C0C4@omniti.com> References: <2E0D6939-B865-42EE-8A54-6AA293B9C0C4@omniti.com> Message-ID: On Mon, 29 Jun 2015, Dan McDonald wrote: > > I was hoping p3 would be past the RC stage and actually be *released*. Given how close we are to the leap second, however, I may not be able to do that. > > Watch this list for an announcement of when I update NTP. I still hope sometime today p3 will jump from RC to Released. What harm might be caused if ntpd misses the leap second? It is my understanding that the Illumos kernel does not make any provision for leap seconds. Only applications with timers based on absolute time should care. Bob -- Bob Friesenhahn bfriesen at simple.dallas.tx.us, http://www.simplesystems.org/users/bfriesen/ GraphicsMagick Maintainer, http://www.GraphicsMagick.org/ From danmcd at omniti.com Mon Jun 29 14:59:22 2015 From: danmcd at omniti.com (Dan McDonald) Date: Mon, 29 Jun 2015 10:59:22 -0400 Subject: [OmniOS-discuss] ntpd -x steps clock on leap second In-Reply-To: References: <2E0D6939-B865-42EE-8A54-6AA293B9C0C4@omniti.com> Message-ID: <02F9104E-4F13-42F3-8496-C7C96102AAE9@omniti.com> > On Jun 29, 2015, at 10:57 AM, Bob Friesenhahn wrote: > > On Mon, 29 Jun 2015, Dan McDonald wrote: >> >> I was hoping p3 would be past the RC stage and actually be *released*. Given how close we are to the leap second, however, I may not be able to do that. >> >> Watch this list for an announcement of when I update NTP. I still hope sometime today p3 will jump from RC to Released. > > What harm might be caused if ntpd misses the leap second? It is my understanding that the Illumos kernel does not make any provision for leap seconds. Only applications with timers based on absolute time should care. Anyone who runs wallclock-type applications (E-commerce) cares a lot about this. Dan From javier.lopez at trixter.de Mon Jun 29 15:03:18 2015 From: javier.lopez at trixter.de (Lopez, Javier) Date: Mon, 29 Jun 2015 17:03:18 +0200 (CEST) Subject: [OmniOS-discuss] Problem updating ms-omniti local repo In-Reply-To: <21899.44322.2195.481653@glaurung.bb-c.de> Message-ID: <955830378.58392.1435590198116.JavaMail.root@de001cs001.trixter.de> Hello Volker, I'm sorry for my silence, I was out of the office the last week. Thank you so much for your help on this, we will try your suggestions and continue on this problem. I will give you feedback as soon as we have progressed. Thank you so much for your help. Best regards Javier From cks at cs.toronto.edu Mon Jun 29 15:07:08 2015 From: cks at cs.toronto.edu (Chris Siebenmann) Date: Mon, 29 Jun 2015 11:07:08 -0400 Subject: [OmniOS-discuss] Schedule for a new kernel update on OmniOS r151014? Message-ID: <20150629150708.D43E87A059A@apps0.cs.toronto.edu> According to the OmniOS source repository's changelogs, there's a fix in the r151014 branch that we're quite interested in: 3783 Flow control is needed in rpcmod when the NFS server is unable to keep up with the network As far as I know this fix is not in the initial release version of r1501014 (it's dated from after then) or in updates since then. (The wiki is confusing on this. On the one hand the http://omnios.omniti.com/wiki.php/ReleaseNotes/r151014 page says that the April 20th update brings the system up to illumos-omnios branch r151014 at 2cac697. On the other hand, that commit is dated May 12th, well after the nominal April 20th date, and the latest kernel update my r151014 test system shows was applied in early May and has a version date component of '20150417' (and 'uname -a' matches the April 20th kernel version).) So: does anyone know when the next OmniOS kernel update is likely to be released? Is there a schedule for this (vague or otherwise)? Am I just looking in the wrong place to spot an available update? (I've done 'pkg refresh' and checked 'pkg update -nv', with nothing visible.) Thanks in advance. - cks From danmcd at omniti.com Mon Jun 29 15:30:54 2015 From: danmcd at omniti.com (Dan McDonald) Date: Mon, 29 Jun 2015 11:30:54 -0400 Subject: [OmniOS-discuss] Schedule for a new kernel update on OmniOS r151014? In-Reply-To: <20150629150708.D43E87A059A@apps0.cs.toronto.edu> References: <20150629150708.D43E87A059A@apps0.cs.toronto.edu> Message-ID: > On Jun 29, 2015, at 11:07 AM, Chris Siebenmann wrote: > > According to the OmniOS source repository's changelogs, there's a > fix in the r151014 branch that we're quite interested in: > 3783 Flow control is needed in rpcmod when the NFS server is > unable to keep up with the network It has not been pushed out yet to the repo servers. > (The wiki is confusing on this. On the one hand the > http://omnios.omniti.com/wiki.php/ReleaseNotes/r151014 page says that > the April 20th update brings the system up to illumos-omnios branch > r151014 at 2cac697. I fubar'ed the revision, and have now fixed that on the wiki. My fault. > So: does anyone know when the next OmniOS kernel update is likely > to be released? Is there a schedule for this (vague or otherwise)? > Am I just looking in the wrong place to spot an available update? The wiki fubar caused the confusion - I'm very sorry about that. There is no set schedule for updates to existing OmniOS releases. I've held back releasing a reboot-your-box update to LTS because, well, it's a reboot-your-box event for a lot of people. I know how important THIS fix is to you, Chris. There's an NTP update going out very soon (once 4.2.8p3 happens), and since that will be disruptive enough, I don't want to add a reboot-your-box event to that this week. There is also an additional kernel fix that I'd like to include alongside the rpcmod fix (illumos 4051). Because kernel component changes can be tricky (and sometimes leave out missing components that are affected), I would like the next reboot-the-box update to be a full rebuild of illumos-omnios. It would include both the rpcmod fix, AND the additional kernel fix that Chip has been testing. Chip has found additional problems, but they were pre-existing, so an Hope this helps, Dan From richard.elling at richardelling.com Sun Jun 28 17:03:28 2015 From: richard.elling at richardelling.com (Richard Elling) Date: Sun, 28 Jun 2015 10:03:28 -0700 Subject: [OmniOS-discuss] OmniOS considered in-memory OS ? In-Reply-To: References: Message-ID: <24D89F84-5EC6-4CBF-8029-2510EB54B828@richardelling.com> > On Jun 26, 2015, at 8:47 AM, John Barfield wrote: > > I?ve been interested in configuring omnios to run in memory off of a ram > disk myself. > > Does anyone know where you could find a good guide for booting Solaris(h) > kernel into memory with a ramdisk? :-) the funny thing is that SunOS has been bootable into ramdisk since... forever. Back in the day, before CDs were invented, we booted from tape or (if you really hated yourself) floppies :-). Net-net, this probably isn't documented anywhere, per se, outside of the building shell scripts :-( The way it works is to produce an image of a file system. Add that to a boot loader and instruct the boot loader to put it in memory and run. The actual work is to produce the image, and this is where the build scripts comes into play. You can certainly take an OmniOS build and produce an in-memory root, however, the missing management part is handling the persistence of identity. For that, it is useful to see how SmartOS handles /var and files normally modified in /etc. -- richard From michael.carruthers at wildbit.com Mon Jun 29 17:29:47 2015 From: michael.carruthers at wildbit.com (Michael Carruthers) Date: Mon, 29 Jun 2015 10:29:47 -0700 Subject: [OmniOS-discuss] Highly Available ZFS Message-ID: Looking into some options for highly available ZFS. So far all signs point to proprietary products, however I am curious if you fine people have any open source recommendations? -------------- next part -------------- An HTML attachment was scrubbed... URL: From felix.nielsen at gmail.com Mon Jun 29 17:40:37 2015 From: felix.nielsen at gmail.com (Felix Nielsen) Date: Mon, 29 Jun 2015 19:40:37 +0200 Subject: [OmniOS-discuss] Highly Available ZFS In-Reply-To: References: Message-ID: Saso did some great work on that -> http://zfs-create.blogspot.dk/2013/06/building-zfs-storage-appliance-part-1.html I have not tried it myself, but please post any learning's. Thanks Felix On Mon, Jun 29, 2015 at 7:29 PM, Michael Carruthers < michael.carruthers at wildbit.com> wrote: > Looking into some options for highly available ZFS. So far all signs point > to proprietary products, however I am curious if you fine people have any > open source recommendations? > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From doug at will.to Mon Jun 29 18:40:24 2015 From: doug at will.to (Doug Hughes) Date: Mon, 29 Jun 2015 14:40:24 -0400 Subject: [OmniOS-discuss] Highly Available ZFS In-Reply-To: References: Message-ID: For things that require that, we've just been going with dual-head controller boxes. One such example is Seagate (nee Xyratex) which has a dual head box and then you can use something like pacemaker to fail over by force pool import/export. You might need to combine this with something like tcpkill to trigger linux NFS clients to failover soon-ish, but it works. previously I've managed to get as low as 30 second failover under the right circumstances. the SMB stuff just seems to work. Another dual head thing includes stuff with dual attached SAS (e.g. old school Sun J4400s), or echo-streams boxes with dual heads. Just about anything with dual attached back end should work. You do have to force import the zpool, so the setup and reliability of your back end connectivity is rather important. On Mon, Jun 29, 2015 at 1:40 PM, Felix Nielsen wrote: > Saso did some great work on that -> > http://zfs-create.blogspot.dk/2013/06/building-zfs-storage-appliance-part-1.html > > I have not tried it myself, but please post any learning's. > > Thanks > Felix > > On Mon, Jun 29, 2015 at 7:29 PM, Michael Carruthers < > michael.carruthers at wildbit.com> wrote: > >> Looking into some options for highly available ZFS. So far all signs >> point to proprietary products, however I am curious if you fine people have >> any open source recommendations? >> >> _______________________________________________ >> OmniOS-discuss mailing list >> OmniOS-discuss at lists.omniti.com >> http://lists.omniti.com/mailman/listinfo/omnios-discuss >> >> > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From mir at miras.org Mon Jun 29 18:52:09 2015 From: mir at miras.org (Michael Rasmussen) Date: Mon, 29 Jun 2015 20:52:09 +0200 Subject: [OmniOS-discuss] Highly Available ZFS In-Reply-To: References: Message-ID: <20150629205209.1b1502d3@sleipner.datanom.net> Does anybody have an idea of how Nexenta does their HA-setup? My guess is that it must involve something with a constant snapshot of the pool using zfs send combined with forced import. -- Hilsen/Regards Michael Rasmussen Get my public GnuPG keys: michael rasmussen cc http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E mir datanom net http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C mir miras org http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 -------------------------------------------------------------- /usr/games/fortune -es says: Ego sum ens omnipotens. -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From Robert.Brock at 2hoffshore.com Mon Jun 29 18:57:31 2015 From: Robert.Brock at 2hoffshore.com (Robert A. Brock) Date: Mon, 29 Jun 2015 18:57:31 +0000 Subject: [OmniOS-discuss] Highly Available ZFS In-Reply-To: <20150629205209.1b1502d3@sleipner.datanom.net> References: , <20150629205209.1b1502d3@sleipner.datanom.net> Message-ID: <2859482C466CCA42AD9B84B9F5621230102A9804@2H199.2hukwok2.local> RSF-1. Sent from my Android phone using TouchDown (www.nitrodesk.com) -----Original Message----- From: Michael Rasmussen [mir at miras.org] Received: Monday, 29 Jun 2015, 19:52 To: omnios-discuss at lists.omniti.com [omnios-discuss at lists.omniti.com] Subject: Re: [OmniOS-discuss] Highly Available ZFS Does anybody have an idea of how Nexenta does their HA-setup? My guess is that it must involve something with a constant snapshot of the pool using zfs send combined with forced import. -- Hilsen/Regards Michael Rasmussen Get my public GnuPG keys: michael rasmussen cc http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E mir datanom net http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C mir miras org http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 -------------------------------------------------------------- /usr/games/fortune -es says: Ego sum ens omnipotens. -------------- next part -------------- An HTML attachment was scrubbed... URL: From robin.blanchard at gmail.com Mon Jun 29 18:59:00 2015 From: robin.blanchard at gmail.com (Robin Blanchard) Date: Mon, 29 Jun 2015 13:59:00 -0500 Subject: [OmniOS-discuss] Highly Available ZFS In-Reply-To: <20150629205209.1b1502d3@sleipner.datanom.net> References: <20150629205209.1b1502d3@sleipner.datanom.net> Message-ID: Last I knew, HA was offloaded to RSF-1 http://www.high-availability.com/zfs-ha-plugin/ On Mon, Jun 29, 2015 at 1:52 PM, Michael Rasmussen wrote: > Does anybody have an idea of how Nexenta does their HA-setup? > > My guess is that it must involve something with a constant snapshot of > the pool using zfs send combined with forced import. > > -- > Hilsen/Regards > Michael Rasmussen > > Get my public GnuPG keys: > michael rasmussen cc > http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E > mir datanom net > http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C > mir miras org > http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 > -------------------------------------------------------------- > /usr/games/fortune -es says: > Ego sum ens omnipotens. > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss > From chip at innovates.com Mon Jun 29 20:24:34 2015 From: chip at innovates.com (Schweiss, Chip) Date: Mon, 29 Jun 2015 15:24:34 -0500 Subject: [OmniOS-discuss] Highly Available ZFS In-Reply-To: <20150629205209.1b1502d3@sleipner.datanom.net> References: <20150629205209.1b1502d3@sleipner.datanom.net> Message-ID: On Mon, Jun 29, 2015 at 1:52 PM, Michael Rasmussen wrote: > Does anybody have an idea of how Nexenta does their HA-setup? > > My guess is that it must involve something with a constant snapshot of > the pool using zfs send combined with forced import. > Nexenta uses RSF-1 from HighAvailability.com. It is dual servers connected to SAS devices. Pools are exported either gracefully or by force on one host and imported on the other. A floating IP address allows the clients to maintain connectivity. I use RSF-1 with OmniOS. It works well, but HA in general has a steep learning curve and A LOT of gotchas that are not well documented anywhere. It took me about a year of learning until HA started actually increasing my storage availability. The price of RSF-1 is well justified if you don't have a lot experience with HA on ZFS. Eventually I will attempt HA without it, but in the mean time it is serving me well. -Chip > > -- > Hilsen/Regards > Michael Rasmussen > > Get my public GnuPG keys: > michael rasmussen cc > http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E > mir datanom net > http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C > mir miras org > http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 > -------------------------------------------------------------- > /usr/games/fortune -es says: > Ego sum ens omnipotens. > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From doug at will.to Mon Jun 29 21:03:28 2015 From: doug at will.to (Doug Hughes) Date: Mon, 29 Jun 2015 17:03:28 -0400 Subject: [OmniOS-discuss] allocation throttle In-Reply-To: References: <46056636-FB58-4867-9CA1-F4681E9F3EBA@omniti.com> <04D97B85-737C-4049-B2F8-66F59741D561@RichardElling.com> Message-ID: It wasn't entirely clear from the video.. is this enabled automatically, or must an /etc/system variable be set to take advantage of it? (We're probably not target customer since our zpools are setup once and done; no incremental growth/change, except very infrequently upgrade disks in place with all the same) On Mon, Jun 29, 2015 at 10:15 AM, Dan McDonald wrote: > > > On Jun 28, 2015, at 1:23 PM, Richard Elling < > richard.elling at richardelling.com> wrote: > > > >> > >> Eesh - which commit? I can tell you which OmniOS release it first > appeared in > > > > time flies? > > 69962b56 Matt Ahrens, 2013-08-26 > > ? richard > > Which means it's been in since r151008. > > Thanks, > Dan > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss > -------------- next part -------------- An HTML attachment was scrubbed... URL: From danmcd at omniti.com Mon Jun 29 21:15:58 2015 From: danmcd at omniti.com (Dan McDonald) Date: Mon, 29 Jun 2015 17:15:58 -0400 Subject: [OmniOS-discuss] NTP 4.2.8p3 Message-ID: I've put it out on the repo servers for 006, 012, 014, and bloody. Officially it's not out yet, but it's downloadable from ntp.org, and is now on mirrors.omniti.com as well. Thanks, Dan From takashiary at gmail.com Mon Jun 29 22:53:03 2015 From: takashiary at gmail.com (takashi ary) Date: Tue, 30 Jun 2015 07:53:03 +0900 Subject: [OmniOS-discuss] NTP 4.2.8p3 In-Reply-To: References: Message-ID: Hi Dan, Thank you for the quick release! Thanks From danmcd at omniti.com Tue Jun 30 17:30:49 2015 From: danmcd at omniti.com (Dan McDonald) Date: Tue, 30 Jun 2015 13:30:49 -0400 Subject: [OmniOS-discuss] Update your NTPd, please Message-ID: <7DE57A70-5AB4-41E0-B98C-0B75C74D2E5C@omniti.com> Yes, I sent the internal note to the wrong alias (this one), but yes, NTPd is now updated to 4.2.8p3 for ALL supported releases: r151006 (old LTS), r151012 (prev. Stable), r151014 (LTS & Stable), and bloody. Turns out, not ONLY does this revision better handle tonight's leapsecond, it also has a small vulnerability fix: http://support.ntp.org/bin/view/Main/SecurityNotice#June_2015_NTP_Security_Vulnerabi So please utter "pkg update" and restart your NTP service, just to be safe. Thanks! Dan