From: Michael Pye (michael@ulimit.org)
Date: Thu Oct 25 2007 - 08:03:02 EDT
Hi managers,
I have a problem with a 3310 and using LUNs > 0.
Storedge 3310 serial number firmware: 415G saf-te firmware: rev 1180,
configured as single bus.
Connected to a Sun Fire v210 using vhdci scsi cable to integrated v210
scsi port. v210 is runs 5.10 Generic_118833-36 kernel.
We have created 3 logical drives on the 3310, which map to LUNs
0,1,2 on controller c2t0d0s0. We can access the drive on LUN0
successfully, but as is known with solaris to get to LUNs 1 and 2 you
have to edit /kernel/drv/sd.conf and add:
name="sd" class="scsi" target0 lun=1;
name="sd" class="scsi" target0 lun=2;
I have done this and than ran:
update_drv -f sd
devfsadm
and I can then access all 3 logical drives, see them in format, use them
as normal.
However, upon a reboot, I can no longer access any of the logical drives
and I see errors:
Oct 25 09:25:57 red-d-0 scsi: [ID 107833 kern.warning] WARNING:
/pci@1c,600000/scsi@2,1 (glm1):
Oct 25 09:25:57 red-d-0 MSGREJ not expected
Oct 25 09:25:57 red-d-0 genunix: [ID 408822 kern.info] NOTICE: glm1:
fault detected in device; service still available
Oct 25 09:25:57 red-d-0 genunix: [ID 611667 kern.info] NOTICE: glm1:
MSGREJ not expected
Oct 25 09:25:57 red-d-0 scsi: [ID 107833 kern.warning] WARNING:
/pci@1c,600000/scsi@2,1 (glm1):
Oct 25 09:25:57 red-d-0 got SCSI bus reset
The only way to get access to all 3 drives again is to remove the
entries in sd.conf, reboot, then I can access lun0 but of course not
lun1 and 2.
I have looked around on sunsolve and can see various suggestions that
are similar to this problem, but not the exact problem.
I have updated to mpt driver patch 120061-02 with no success, and I have
also put this into /kernel/drv/mpt.conf
device-type-scsi-options-list =
"SUN StorEdge 3310", "SE3310-scsi-options";
SE3310-scsi-options = 0x41ff8;
which is supposed to force scsi speed negotiation to 160mb, as some
sunsolve documents suggested the problem could be a speed negotiation one.
I'm certain this is a solaris bug and that perhaps there is a patch that
fixes it, as the logical drives are usable before the reboot.
Michael Pye
_______________________________________________
sunmanagers mailing list
sunmanagers@sunmanagers.org
http://www.sunmanagers.org/mailman/listinfo/sunmanagers
This archive was generated by hypermail 2.1.7 : Wed Apr 09 2008 - 23:42:27 EDT