Hi Bruce
Seeing as your question has, unfortunately,
not attracted many replies, I will give you an opinion!

Which database are you proposing to use?
What iscsi initiator software are you proposing to use?

There are a few 'issues' with the Solaris iscsi target.
It's beginning to look like you may have more
issues with Solaris 10 than Nevada, as it seems
the bugs are not being automatically fixed in Solaris 10.

There are currently some memory leak issues.
If you run the target for long enough it may use
up all your memory, giving sluggish performance.
Of course, a restart of the target daemon with reset this.
These memory leaks should be fixed in Nevada soon.
http://bugs.opensolaris.org/view_bug.do?bug_id=6657591

Also under certain unusual circumstance, the target can core dump.
A fix has recently been putback into the Nevada code base for this:
http://bugs.opensolaris.org/view_bug.do?bug_id=6593606

Is your are using the Microsoft Windows iscsi initiator.
and you are using Solaris 10u4 (8/07) for the target
then you are going to hit the bug discussed recently
on this storage-forum.  Hopefully a patch will be available
soon. (This has been fixed in Nevada for along time).
http://bugs.opensolaris.org/view_bug.do?bug_id=6521425

Also it is reported that the target fails on the Microsoft
iscsi compliance test, but this MAY be more academic,
than a practical concern.
http://bugs.opensolaris.org/view_bug.do?bug_id=6641983

I would recommend testing the target for yourself,
in your environment, with whatever initiator
and database you plan to use.

For anyone contemplating Microsoft's SQL server database,
they do have a stress testing tool, 'SQLIOSim '
with which you can perform testing.
http://support.microsoft.com/kb/q231619/
Regards
Nigel Smith
 
 
This message posted from opensolaris.org
_______________________________________________
storage-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/storage-discuss

Reply via email to