[OmniOS-discuss] Omios, hvm and AWS

[Al Slater]

On 31/07/17 11:58, Al Slater wrote:
> On 31/07/2017 11:39, Peter Tribble wrote:
>>
>>
>> On Mon, Jul 31, 2017 at 11:09 AM, Al Slater <al.slater at scluk.com
>> <mailto:al.slater at scluk.com>> wrote:
>>
>>     On 31/07/2017 11:07, Al Slater wrote:
>>     > On 30/07/2017 20:15, Peter Tribble wrote:
>>     >>     > The following should get you going:
>>     >>     >
>>     >>     > https://www.prakashsurya.com/post/2017-02-06-creating-a-custom-amazon-ec2-ami-from-iso/
>>     <https://www.prakashsurya.com/post/2017-02-06-creating-a-custom-amazon-ec2-ami-from-iso/>
>>     >>     <https://www.prakashsurya.com/post/2017-02-06-creating-a-custom-amazon-ec2-ami-from-iso/
>>     <https://www.prakashsurya.com/post/2017-02-06-creating-a-custom-amazon-ec2-ami-from-iso/>>
>>     >
>>     > OK, I followed the above procedure and have produced an AMI.
>>     >
>>     > When I create an instance and try to boot it, I get the following in the
>>     > system log:
>>
>>     SunOS Release 5.11 Version omnios-r151022-f9693432c2 64-bit
>>
>>     Copyright (c) 1983, 2010, Oracle and/or its affiliates. All rights
>>     reserved.
>>
>>     NOTICE: Cannot read the pool label from '/xpvd/xdf at 51728:a'
>>     NOTICE: spa_import_rootpool: error 5
>>
>>     Cannot mount root on /xpvd/xdf at 51728:a fstype zfs
>>     panic[cpu0]/thread=fffffffffbc38560: vfs_mountroot: cannot mount root
>>     Warning - stack not written to the dump buffer
>>     fffffffffbc7ad70 genunix:vfs_mountroot+39b ()
>>     fffffffffbc7adb0 genunix:main+138 ()
>>     fffffffffbc7adc0 unix:_locore_start+90 ()
>>
>>
>>     How can I fix this?
>>
>>
>> You're likely the first person down this path.
>>
>> Generically, this means that the device paths embedded in the pool
>> don't match those provided by the "hardware" you're booting on.
>>
>> So the system thinks it should have a disk at /xpvd/xdf at 51728:a
>>
>> On my instance, I have:
>>
>> /dev/rdsk/c2t0d0s0 -> ../../devices/xpvd/xdf at 51712:a,raw
>>
>> In other words, 51712 not 51728.
>>
>> For this to work, you have to set up your xen instance to exactly mirror
>> what EC2 provides. Somehow it's gotten mixed up. In your configuration,
>> did you use xvda? I think 51728 is what you get if you use xvdb for the
>> disk,
>> which won't work. I had:
>>
>> disk=[  'file:/home/ptribble/iso/tribblix-0m20.1.iso,hdb:cdrom,r',
>>         'file:/root/ami-template.img,xvda,w' ]
>>
> 
> Thanks Peter,  I see what happened...
> 
> I started off with the instructions from
> https://wiki.openindiana.org/oi/Creating+OpenIndiana+EC2+image
> 
> Then changed to following the instructions at
> https://www.prakashsurya.com/post/2017-02-06-creating-a-custom-amazon-ec2-ami-from-iso/
> 
> while neglecting to change the disks line in my xen config.
> 
> Oh well, starting again...

An update for the list...

With the correct disks line in the xen config file, the procedure above,
with Peter's workaround for the xen networking panic, works fine.  I
have a working AMI now.

One more question though, is there any way to enable an SMF service for
the next reboot, but not immediately.  Specifically, I want to enable
the initial-boot service with a .initialboot file in place, then create
a new AMI.

I wist to use .initialboot to grab the instance configuration from
amazon (hostname, root keys etc) and configure appropriately when the
new instance starts.

-- 
Al Slater