I took a look at the core file from this panic and I don't think we've
seen this issue before. I've asked Mike to file a bug to track this.
The fix looks straightforward -- we seem to by trying to get a hold on
the session in one thread while tearing it down in another thread. We
should be checking the session state before attempting to get the hold.
I haven't had a chance to look into the fdisk issues yet -- we have run
with RHEL before on both sockets and iSER.
-Peter
Michael F. Biniasz wrote:
i am working with a customer that is testing/validating OpenSolaris dev build
snv_111 iSCSI target
functionality. we are seeing intermittent panics, the console information
follows at the end of
this email.
- configs:
iSCSI target (HP460c server), OpenSolaris snv_111, ZFS storage, all mirrored
pairs, ZIL on
4 x 30gb SAS SSDs, the cache is on 3 SSDs. the Sybase volumes are all zvols
mapped as raw
devices on the client.
iSCSI initiator (HP385 G2), RHEL 4.5.2, Sybase 12.5.4 (raw devices, writes 2k,
logs 16k)
we are seeing issues with the RHEL iSCSI initiator not being able to correctly
map the zvols
as fdisk sometimes hangs and Sybase complains when initializing the raw partition. this may
very well be an issue on the client side but i wanted to verify whether other people have seen
similar issues. i will be working with the customer to further investigate. i
have the full core dump.
Apr 27 16:49:49 proxstore2 last message repeated 2 times
Apr 28 09:32:08 proxstore2 unix: [ID 836849 kern.notice]
Apr 28 09:32:08 proxstore2 ^Mpanic[cpu14]/thread=ffffff007ae90c60:
Apr 28 09:32:08 proxstore2 genunix: [ID 103648 kern.notice] mutex_destroy: not
owner, lp=ffffff117da3f030 owner=ffffff007b5bbc60 thread=ffffff007ae90c60
Apr 28 09:32:08 proxstore2 unix: [ID 100000 kern.notice]
Apr 28 09:32:08 proxstore2 genunix: [ID 655072 kern.notice] ffffff007ae90af0
unix:mutex_panic+73 ()
Apr 28 09:32:08 proxstore2 genunix: [ID 655072 kern.notice] ffffff007ae90b20
unix:mutex_destroy+ce ()
Apr 28 09:32:08 proxstore2 genunix: [ID 655072 kern.notice] ffffff007ae90b40
idm:idm_refcnt_destroy+24 ()
Apr 28 09:32:08 proxstore2 genunix: [ID 655072 kern.notice] ffffff007ae90b60
iscsit:iscsit_sess_destroy+1a ()
Apr 28 09:32:08 proxstore2 genunix: [ID 655072 kern.notice] ffffff007ae90b90
iscsit:iscsit_sess_unref+6e ()
Apr 28 09:32:08 proxstore2 genunix: [ID 655072 kern.notice] ffffff007ae90bc0
idm:idm_refcnt_unref_task+5e ()
Apr 28 09:32:08 proxstore2 genunix: [ID 655072 kern.notice] ffffff007ae90c40
genunix:taskq_thread+193 ()
Apr 28 09:32:08 proxstore2 genunix: [ID 655072 kern.notice] ffffff007ae90c50
unix:thread_start+8 ()
Apr 28 09:32:08 proxstore2 unix: [ID 100000 kern.notice]
Apr 28 09:32:08 proxstore2 genunix: [ID 672855 kern.notice] syncing file
systems...
Apr 28 09:32:09 proxstore2 genunix: [ID 904073 kern.notice] done
Apr 28 09:32:10 proxstore2 genunix: [ID 111219 kern.notice] dumping to
/dev/zvol/dsk/rpool/dump, offset 65536, content: kernel
Apr 28 09:34:37 proxstore2 genunix: [ID 409368 kern.notice] ^M100% done:
1205915 pages dumped, compression ratio 4.82,
Apr 28 09:34:37 proxstore2 genunix: [ID 851671 kern.notice] dump succeeded
Apr 28 09:37:55 proxstore2 genunix: [ID 540533 kern.notice] ^MSunOS Release
5.11 Version snv_111 64-bit
Apr 28 09:37:55 proxstore2 genunix: [ID 943908 kern.notice] Copyright 1983-2009
Sun Microsystems, Inc. All rights reserved.
thank you,
mike b.
_______________________________________________
storage-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/storage-discuss
