So I've encountered more fun FC problems and was wondering if anyone could help 
troubleshoot.

I've got three Sun V210's booting off LP10000DC's with a single fiber 
connection through a Brocade Silkworm 3900 to a PC with a COMSTAR FC target 
running on snv_134 with an LP10000DCe card.

Under normal circumstances I see no problems but when provisioning Oracle I get 
the following errors:
Aug 30 18:31:29 sl10-rac079-03 scsi: WARNING: 
/p...@1d,700000/e...@1/f...@0,0/s...@w10000000c94222ec,e (ssd2):
Aug 30 18:31:29 sl10-rac079-03  SCSI transport failed: reason 'tran_err': 
retrying command

s...@w10000000c94222ec,e is the root disk for the system, and the specific 
error always refers to the root disk in the system experiencing the problem 
(s...@w10000000c94222ec,e for the 3rd node, s...@w10000000c94222ec,d for the 
second node, and s...@w10000000c94222ec,c for the first node). On the first 
node this has caused the file system to go read-only. On the second and third 
systems I get the above error but eventually the provisioning process seems to 
finish fine.

The initiators are running either 1.91a or 1.92a firmware. The card in the head 
is running 1.92a. ZFS does not appear to hang. The only error I see in COMSTAR 
is:
emlxs0:17212312: fct_gid_cb: GPN_ID is not accepted by NS - 
icmd/ffffff0c1c2d0650

That said- the systems have clearly booted and been working fine. I see no 
errors on the switch ports.

Any troubleshooting recommendations would be welcome.

It's possible this is a problem on the V210's but the failure scenario is 
almost identical in every case and I've tried swapping out the switch, cables 
and firmware on the initiators.

Thanks,
-Don
-- 
This message posted from opensolaris.org
_______________________________________________
storage-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/storage-discuss

Reply via email to