>From the trace buffer it seems that the local port (emlxs in this case) is >causing a logout. I know emlxs driver has its own tracing which can be enabled >to get more data. Probably someone from Emulex can help here.
Sumit -----Original Message----- From: Matty [mailto:[email protected]] Sent: Mon 5/4/2009 6:05 PM To: Sumit Gupta Cc: [email protected] Subject: Re: [storage-discuss] Does COMSTAR support ACTIVE/ACTIVE paths? On Mon, May 4, 2009 at 1:00 PM, Sumit Gupta <[email protected]> wrote: > > All the paths on COMSTAR are always active-active. I am not sure why the > paths are showing up as faulty. Try resetting the ports (stmfadm > offline-target wwn.... followed by stmfadm online-target www.....) or > rebooting the initiator to see if that makes any difference. You can get the > stmf trace buffer as: > > echo '*stmf_trace_buf/s' |mdb -k > stmftrace.txt > > This might reveal something. Hey Sumit, After rebooting the host, I ran the following dd command on the Linux client: $ dd if=/dev/mapper/mpath2 of=/dev/null bs=1048576 After this ran for a minute or two, one of the paths failed: $ multipath -ll mpath2 (3600144f02d658400000049fba1bf0001) dm-0 SUN,COMSTAR [size=100G][features=0][hwhandler=0][rw] \_ round-robin 0 [prio=2][active] \_ 0:0:0:0 sda 8:0 [active][ready] \_ 1:0:0:0 sdb 8:16 [failed][ready] mpath4 (3600144f02d658400000049ffc0b80001) dm-1 SUN,COMSTAR [size=100G][features=0][hwhandler=0][rw] \_ round-robin 0 [prio=2][active] \_ 0:0:0:1 sdd 8:48 [active][ready] \_ 1:0:0:1 sdf 8:80 [failed][ready] mpath3 (3600144f02d658400000049ffc0bc0002) dm-2 SUN,COMSTAR [size=100G][features=0][hwhandler=0][rw] \_ round-robin 0 [prio=2][active] \_ 0:0:0:2 sde 8:64 [active][ready] \_ 1:0:0:2 sdg 8:96 [failed][ready] And various errors were sent to the messages file: May 5 04:55:16 disarm kernel: sd 1:0:0:2: SCSI error: return code = 0x00020000 May 5 04:55:16 disarm kernel: end_request: I/O error, dev sdg, sector 11630336 May 5 04:55:16 disarm kernel: device-mapper: multipath: Failing path 8:96. May 5 04:56:02 disarm kernel: sd 1:0:0:1: SCSI error: return code = 0x00020000 May 5 04:56:02 disarm kernel: end_request: I/O error, dev sdf, sector 11620608 May 5 04:56:02 disarm kernel: sd 1:0:0:2: SCSI error: return code = 0x00020000 May 5 04:56:02 disarm kernel: end_request: I/O error, dev sdg, sector 11624192 May 5 04:56:02 disarm kernel: sd 1:0:0:2: SCSI error: return code = 0x00020000 May 5 04:56:02 disarm kernel: end_request: I/O error, dev sdg, sector 11625984 May 5 04:56:02 disarm kernel: sd 1:0:0:2: SCSI error: return code = 0x00020000 May 5 04:56:02 disarm kernel: end_request: I/O error, dev sdg, sector 11625728 May 5 04:56:02 disarm kernel: sd 1:0:0:2: SCSI error: return code = 0x00020000 May 5 04:56:02 disarm kernel: end_request: I/O error, dev sdg, sector 11626240 May 5 04:56:02 disarm kernel: sd 1:0:0:2: SCSI error: return code = 0x00020000 May 5 04:56:02 disarm kernel: end_request: I/O error, dev sdg, sector 11623808 May 5 04:56:02 disarm kernel: sd 1:0:0:2: SCSI error: return code = 0x00020000 May 5 04:56:02 disarm kernel: end_request: I/O error, dev sdg, sector 11625344 May 5 04:56:02 disarm kernel: sd 1:0:0:2: SCSI error: return code = 0x00020000 May 5 04:56:02 disarm kernel: end_request: I/O error, dev sdg, sector 11626496 May 5 04:56:02 disarm kernel: sd 1:0:0:2: SCSI error: return code = 0x00020000 May 5 04:56:02 disarm kernel: end_request: I/O error, dev sdg, sector 11626624 May 5 04:56:02 disarm kernel: sd 1:0:0:2: SCSI error: return code = 0x00020000 May 5 04:56:02 disarm kernel: end_request: I/O error, dev sdg, sector 11626752 May 5 04:56:03 disarm kernel: sd 1:0:0:2: SCSI error: return code = 0x00020000 May 5 04:56:03 disarm kernel: end_request: I/O error, dev sdg, sector 11627520 May 5 04:56:03 disarm kernel: sd 1:0:0:2: SCSI error: return code = 0x00020000 May 5 04:56:03 disarm kernel: end_request: I/O error, dev sdg, sector 11627264 May 5 04:56:03 disarm kernel: sd 1:0:0:2: SCSI error: return code = 0x00020000 May 5 04:56:03 disarm kernel: end_request: I/O error, dev sdg, sector 11627008 May 5 04:56:03 disarm kernel: sd 1:0:0:2: SCSI error: return code = 0x00020000 May 5 04:56:03 disarm kernel: end_request: I/O error, dev sdg, sector 11627648 May 5 04:56:03 disarm kernel: sd 1:0:0:2: SCSI error: return code = 0x00020000 May 5 04:56:03 disarm kernel: end_request: I/O error, dev sdg, sector 11628416 May 5 04:56:03 disarm kernel: sd 1:0:0:2: SCSI error: return code = 0x00020000 May 5 04:56:03 disarm kernel: end_request: I/O error, dev sdg, sector 11628544 May 5 04:56:47 disarm multipathd: sdf: readsector0 checker reports path is down May 5 04:56:47 disarm multipathd: 8:96: mark as failed May 5 04:56:47 disarm multipathd: mpath3: remaining active paths: 1 May 5 04:56:47 disarm multipathd: dm-2: add map (uevent) May 5 04:56:47 disarm multipathd: dm-2: devmap already registered May 5 04:56:54 disarm multipathd: sdg: readsector0 checker reports path is up May 5 04:56:54 disarm multipathd: 8:96: reinstated May 5 04:56:54 disarm multipathd: mpath3: remaining active paths: 2 May 5 04:56:54 disarm multipathd: dm-2: add map (uevent) May 5 04:56:54 disarm multipathd: dm-2: devmap already registered May 5 04:56:56 disarm kernel: sd 1:0:0:2: SCSI error: return code = 0x00020000 May 5 04:56:56 disarm kernel: end_request: I/O error, dev sdg, sector 11658240 May 5 04:56:56 disarm kernel: device-mapper: multipath: Failing path 8:96. May 5 04:57:39 disarm kernel: sd 1:0:0:0: SCSI error: return code = 0x00020000 May 5 04:57:39 disarm kernel: end_request: I/O error, dev sdb, sector 11633792 May 5 04:57:39 disarm kernel: device-mapper: multipath: Failing path 8:16. May 5 04:57:40 disarm kernel: sd 1:0:0:0: SCSI error: return code = 0x00020000 May 5 04:57:40 disarm kernel: end_request: I/O error, dev sdb, sector 11640064 May 5 04:57:49 disarm multipathd: sdf: readsector0 checker reports path is down May 5 04:57:49 disarm multipathd: 8:96: mark as failed May 5 04:57:49 disarm multipathd: mpath3: remaining active paths: 1 May 5 04:57:49 disarm multipathd: dm-2: add map (uevent) May 5 04:57:49 disarm multipathd: dm-2: devmap already registered May 5 04:57:49 disarm multipathd: dm-0: add map (uevent) May 5 04:57:49 disarm multipathd: dm-0: devmap already registered May 5 04:57:49 disarm multipathd: 8:16: mark as failed May 5 04:57:49 disarm multipathd: mpath2: remaining active paths: 1 May 5 04:57:53 disarm multipathd: sdg: readsector0 checker reports path is up May 5 04:57:53 disarm multipathd: 8:96: reinstated May 5 04:57:53 disarm multipathd: mpath3: remaining active paths: 2 May 5 04:57:53 disarm multipathd: dm-2: add map (uevent) May 5 04:57:53 disarm multipathd: dm-2: devmap already registered May 5 04:57:55 disarm kernel: sd 1:0:0:2: SCSI error: return code = 0x00020000 May 5 04:57:55 disarm kernel: end_request: I/O error, dev sdg, sector 11721728 May 5 04:57:55 disarm kernel: device-mapper: multipath: Failing path 8:96. If I switch the multi-pathing policy to failover (ACTIVE/PASSIVE), everything appears to work correctly. It's only when I use multibus (ACTIVE/ACTIVE) that things go south. The stmftrace data is attached, and I'm curious to get your thoughts on what might be happening here. Thanks, - Ryan -- http://prefetch.net
_______________________________________________ storage-discuss mailing list [email protected] http://mail.opensolaris.org/mailman/listinfo/storage-discuss
