[OmniOS-discuss] Bloody update for December 11th
Guenther Alka
alka at hfg-gmuend.de
Tue Dec 15 11:31:00 UTC 2015
I have updated the pdf with results from OmniOS bloody.
Main resultsfor 10G Ethernet on OSX 10.11 and Windows 8.1
- OS version and client network driver is very criticalfor 10G
on some configs or with some driver releases 10G is not faster than
1G (mostly on reads)
- From Windows, performance to Solaris is similar than to OmniOS (at a
lower level than with OSX)
- From OSX, SMB2 to Solaris is faster than to OmniOS
- OSX is faster than Windows on SMB2 reads and writes out of the box
SMB2 perfomance on OSX goes up to > 600 MB/s on writes and > 800
MB/s on reads
SMB perfomance on Windows goes up to > 300 MB/s on writes and > 600
MB/s on reads
This is a quick "out of the box" check with SMB2 and Jumboframes as the
only special settingson OSX.
On Windows 8.1 defaults + mtu 9000 are used. Maybe we need some
additional tweakings on Windows
Am 12.12.2015 um 22:07 schrieb Bob Friesenhahn:
> On Fri, 11 Dec 2015, Günther Alka wrote:
>
>> Many Thanks to Nexenta
>> and to OmniTi for this december bloody with SMB 2
>>
>> I have just done some tests on OSX under Solaris 11.3 to check some
>> configuration
>> options for a ZFS video editing storage server for my Mac Pros.
>
> Do you plan to add tests with the implementation in OmniOS bloody? The
> Nexenta implementation might be quite a lot different than the Oracle
> Solaris one. Perhaps it might even fail with your tests.
>
>> There are two must have principles: SMB2 and Jumboframes
>> see http://napp-it.org/doc/downloads/performance_smb2.pdf
>
> I was surprised to see the huge improvement with jumbo frames.
>
> Bob
--
H f G
Hochschule für Gestaltung
university of design
Schwäbisch Gmünd
Rektor-Klaus Str. 100
73525 Schwäbisch Gmünd
Guenther Alka, Dipl.-Ing. (FH)
Leiter des Rechenzentrums
head of computer center
Tel 07171 602 627
Fax 07171 69259
guenther.alka at hfg-gmuend.de
http://rz.hfg-gmuend.de
More information about the OmniOS-discuss
mailing list