[pnfs] New GPFS for CITI ibmcl cluster and crash info for current image
Marc Eshel
eshel at almaden.ibm.com
Wed Jan 23 14:16:11 EST 2008
I found my problem. When a file is deleted my fs was calling
sync_layout_recall which was added to test layout recall. In
sync_layout_recall() it gets the error message ...has no callback path and
proceeds to call put_layoutrecall() which will call .... that will call
put_nfs4_file() which will free the nfs4_file.
Marc.
Benny Halevy <bhalevy at panasas.com> wrote on 01/23/2008 08:11:51 AM:
> On Jan. 23, 2008, 18:03 +0200, Marc Eshel <eshel at almaden.ibm.com> wrote:
> > I will build and send GPFS later today so we can compare results, but
you
> > will have to use a branch that includes Dean's changes to export ops.
It
> > looks like Benny might be hitting the same problem with nfsd4_close.
>
> Hmm, the trace below is not the same as the one I fixed today.
> (see 22feaa118f497260b2dec5cdf83009915343a4f9)
>
> Benny
>
> > Marc.
> >
> >
> >
> >
> > "William A. (Andy) Adamson" <andros at citi.umich.edu>
> > Sent by: androsadamson at gmail.com
> > 01/23/2008 07:53 AM
> >
> > To
> > "Marc Eshel" <eshel at almaden.ibm.com>
> > cc
> > pnfs at linux-nfs.org
> > Subject
> > Re: New GPFS for CITI ibmcl cluster and crash info for current image
> >
> >
> >
> >
> >
> >
> > Hi Marc
> >
> > I'm running a 2.6.24-rc4 kernel based on a pre-benny ricardo
> > linux-2.6-latest git tree.
> > Quite old.
> >
> > I built the kernel on
ibmcl1:/usr/local/src/kernel/linux-pnfs-2.6-latest
> >
> > Which connectathon lock test failed?
> >
> > I want to install the latest 2.6.24-rc8 on the cluster and debug it.
The
> > question is, do I need a new GPFS image?
> >
> > -->Andy
> >
> > On Jan 22, 2008 10:41 PM, Marc Eshel < eshel at almaden.ibm.com> wrote:
> > Hi Andy,
> > Can you please remind me version of Linux you are running on the
server
> > and which branch of the tree are you using. I tried to run just the
lock
> > test from the connectathon suite and got the following results. I am
using
> > filelayout-draft-13 branch with Linux 2.6.24-rc8-pnfs. I will try to
debug
> >
> > some more tomorrow.
> >
> > Marc.
> >
> >
> > Jan 22 15:20:45 bear107 kernel: xxx server proc 4 CLOSE
> > Jan 22 15:20:45 bear107 kernel: NFSD: nfsd4_close on file lockfile8090
> > Jan 22 15:20:45 bear107 kernel: NFSD: preprocess_seqid_op: seqid=0
stateid
> >
> > = (47967970/00000020/000000
> > 0c/00000000)
> > Jan 22 15:20:45 bear107 kernel: NFSD: find_stateid flags 0x45
> > Jan 22 15:20:45 bear107 kernel: renewing client (clientid
> > 47967970/00000001)
> > 000000100108 RIP:
> > Jan 22 15:20:45 bear107 kernel: [<ffffffff883ad38a>]
> > :nfsd:free_nfs4_file+0x39/0x85
> > Jan 22 15:20:45 bear107 kernel: PGD 22a96a067 PUD 22b50f067 PMD 0
> > Jan 22 15:20:45 bear107 kernel: Oops: 0002 [1] SMP
> > Jan 22 15:20:45 bear107 kernel: CPU 0
> > Jan 22 15:20:45 bear107 kernel: Modules linked in: nfsd auth_rpcgss
> > exportfs mmfs mmfslinux tracedev a
> > utofs4 nfs lockd sunrpc dm_mirror dm_mod ehci_hcd ohci_hcd usbcore
bnx2
> > qla2xxx ext3 jbd mptsas scsi_t
> > ransport_sas mptspi scsi_transport_spi mptfc scsi_transport_fc
mptscsih
> > mptbase sd_mod
> > Jan 22 15:20:45 bear107 kernel: Pid: 10145, comm: nfsd Not tainted
> > 2.6.24-rc8-pnfs #1
> > Jan 22 15:20:45 bear107 kernel: RIP: 0010:[<ffffffff883ad38a>]
> > [<ffffffff883ad38a>] :nfsd:free_nfs4_f
> > ile+0x39/0x85
> > Jan 22 15:20:45 bear107 kernel: RSP: 0018:ffff81022370fd80 EFLAGS:
> > 00010286
> > Jan 22 15:20:45 bear107 kernel: RAX: 0000000000100100 RBX:
> > ffff810221575070 RCX: ffff810221575078
> > Jan 22 15:20:45 bear107 kernel: RDX: 0000000000200200 RSI:
> > 0000000000000046 RDI: ffffffff883bb0fc
> > Jan 22 15:20:45 bear107 kernel: RBP: ffffffff883ad351 R08:
> > 0000000000000000 R09: 0000000000000000
> > Jan 22 15:20:45 bear107 kernel: R10: 0000000000000000 R11:
> > 0000000000000000 R12: ffff81022e51aa80
> > Jan 22 15:20:45 bear107 kernel: R13: 0000000000000000 R14:
> > ffff810222a4b000 R15: 0000000000000000
> > Jan 22 15:20:45 bear107 kernel: FS: 00002b3f01091b00(0000)
> > GS:ffffffff8056f000(0000) knlGS:0000000000
> > 000000
> > Jan 22 15:20:45 bear107 kernel: CS: 0010 DS: 0000 ES: 0000 CR0:
> > 000000008005003b
> > Jan 22 15:20:45 bear107 kernel: CR2: 0000000000100108 CR3:
> > 000000022e61f000 CR4: 00000000000006e0
> > Jan 22 15:20:45 bear107 kernel: DR0: 0000000000000000 DR1:
> > 0000000000000000 DR2: 0000000000000000
> > Jan 22 15:20:45 bear107 kernel: DR3: 0000000000000000 DR6:
> > 00000000ffff0ff0 DR7: 0000000000000400
> > Jan 22 15:20:45 bear107 kernel: Process nfsd (pid: 10145, threadinfo
> > ffff81022370e000, task ffff81022d
> > b6c860)
> > Jan 22 15:20:45 bear107 kernel: Stack: ffff810221575070
ffffffff802c8e32
> > ffff810008a45210 ffff81022fc
> > ea2e0
> > Jan 22 15:20:45 bear107 kernel: ffff810221575070 ffffffff883ae41a
> > ffff810222dd5dd8 0000000000000004
> > Jan 22 15:20:45 bear107 kernel: ffff81022e51aa80 ffffffff883ae4e9
> > ffff81022fcea250 ffffffff883ae37f
> > Jan 22 15:20:45 bear107 kernel: Call Trace:
> > Jan 22 15:20:45 bear107 kernel: [<ffffffff802c8e32>]
kref_put+0x74/0x82
> > Jan 22 15:20:45 bear107 kernel: [<ffffffff883ae41a>]
> > :nfsd:release_stateid+0x162/0x175
> > Jan 22 15:20:45 bear107 kernel: [<ffffffff883ae4e9>]
> > :nfsd:release_stateowner+0xbc/0xf1
> > Jan 22 15:20:45 bear107 kernel: [<ffffffff883ae37f>]
> > :nfsd:release_stateid+0xc7/0x175
> > Jan 22 15:20:45 bear107 kernel: [<ffffffff883b0899>]
> > :nfsd:nfsd4_close+0x96/0x124
> > Jan 22 15:20:45 bear107 kernel: [<ffffffff883a5255>]
> > :nfsd:nfsd4_proc_compound+0x29d/0x440
> > Jan 22 15:20:45 bear107 kernel: [<ffffffff88396878>]
> > :nfsd:nfsd_dispatch+0xe2/0x1d6
> > Jan 22 15:20:45 bear107 kernel: [<ffffffff8812f70c>]
> > :sunrpc:svc_process+0x408/0x6de
> > Jan 22 15:20:46 bear107 kernel: [<ffffffff88396678>]
> > :nfsd:nfsd+0x1a4/0x2c2
> > Jan 22 15:20:46 bear107 kernel: [<ffffffff8020c3f8>]
child_rip+0xa/0x12
> > Jan 22 15:20:46 bear107 kernel: [<ffffffff883964d4>]
:nfsd:nfsd+0x0/0x2c2
> > Jan 22 15:20:46 bear107 kernel: [<ffffffff8020c3ee>]
child_rip+0x0/0x12
> > Jan 22 15:20:46 bear107 kernel:
> > Jan 22 15:20:46 bear107 kernel:
> > Jan 22 15:20:46 bear107 kernel: Code: 48 89 50 08 48 89 02 31 c0 48 c7
41
> > 08 00 02 20 00 48 8b 73
> > Jan 22 15:20:46 bear107 kernel: RIP [<ffffffff883ad38a>]
> > :nfsd:free_nfs4_file+0x39/0x85
> > Jan 22 15:20:46 bear107 kernel: RSP <ffff81022370fd80>
> > Jan 22 15:20:46 bear107 kernel: CR2: 0000000000100108
> > Jan 22 15:20:46 bear107 kernel: ---[ end trace a9871b2cb4f37ac4 ]---
> >
> >
> >
> > _______________________________________________
> > pNFS mailing list
> > pNFS at linux-nfs.org
> > http://linux-nfs.org/cgi-bin/mailman/listinfo/pnfs
>
More information about the pNFS
mailing list