Skip to contentSkip to conversation

kjournald tries to access beyond raid device

DISCUSSION Jose Carlos Garcia Sogo 1 participant 1 message Dec 19, 2001

lore
Share & tools
lore
#0 lore raw

Hello!

I'm suffering some problems last two days with (seems ext3) It happens both for 2.4.16 and 2.4.15 (I have not checked yet what
happens with 2.4.17).

The log I get is attached below, but seems that kjournald is trying
to access beyond the device limits. This makes all access to that device impossible (even trying to umount it) and I have to reboot the system. After that, the journal is played and all seems fine. But after a while the error appears again. It seems that it has nothing to do with an concrete event, but some process trying to write something to disk (usually exim/procmail)

The device that fails is a software RAID 5 device (/dev/md3) with
three SCSI hard disks attached to an adaptec UW controller 19160.

I have another 3 raid devices (RAID 5 and RAID 0 levels), which have
not suffered any problem at the moment. I have also other RAID 0 device with reiserfs.

Anyway, currently I have returned to ext2 in all my devices for
safety (this is a production server)

The modules I have loaded are:
Module Size Used by Not tainted
ipt_MASQUERADE 1248 1 (autoclean)
iptable_nat 12884 0 (autoclean) [ipt_MASQUERADE]
ip_conntrack 13324 1 (autoclean) [ipt_MASQUERADE
iptable_nat]
ip_tables 10560 4 [ipt_MASQUERADE iptable_nat]
nfsd 66624 8 (autoclean)
lockd 48288 1 (autoclean) [nfsd]
sunrpc 59604 1 (autoclean) [nfsd lockd]
reiserfs 148448 1 (autoclean)
ne2k-pci 5056 1
8390 6016 0 [ne2k-pci]
3c59x 24968 1

I also can send you my configure file if you need it. Also, if you need any other info, please e-mail me. I have the SysRQ+T and the SysRQ+P log just after the failure.

Thank you

José Carlos García Sogo
[email protected]

Dec 19 16:03:54 matrix kernel: attempt to access beyond end of device
Dec 19 16:03:54 matrix kernel: 09:03: rw=1, want=1625616132, limit=10249216
Dec 19 16:03:54 matrix kernel: attempt to access beyond end of device
Dec 19 16:03:54 matrix kernel: 09:03: rw=1, want=134217732, limit=10249216
Dec 19 16:03:54 matrix kernel: Assertion failure in __journal_remove_journal_hea
d() at journal.c:1668: "buffer_jbd(bh)"
Dec 19 16:03:54 matrix kernel: invalid operand: 0000
Dec 19 16:03:54 matrix kernel: CPU:    0
Dec 19 16:03:54 matrix kernel: EIP:    0010:[__journal_remove_journal_head+125/2
24]    Not tainted
Dec 19 16:03:54 matrix kernel: EFLAGS: 00010282
Dec 19 16:03:54 matrix kernel: eax: 0000005c   ebx: d0acbc90   ecx: dea9a000   e
dx: 00000001
Dec 19 16:03:54 matrix kernel: esi: ce1c49c0   edi: ddd51540   ebp: 0000000a   e
sp: df18be64
Dec 19 16:03:54 matrix kernel: ds: 0018   es: 0018   ss: 0018
Dec 19 16:03:54 matrix kernel: Process kjournald (pid: 122, stackpage=df18b000)
Dec 19 16:03:54 matrix kernel: Stack: c0222b20 c02232ef c0222af1 00000684 c02233
0d ce1c49c0 d0acbc90 c015cb5c
Dec 19 16:03:54 matrix kernel:        ce1c49c0 ce1c49c0 c0159515 d0acbc90 de9942
50 de994200 de994200 00000000
Dec 19 16:03:54 matrix kernel:        de994294 de994250 00000000 00000fdc ce78d0
24 d5db7930 00000000 00000000
Dec 19 16:03:54 matrix kernel: Call Trace: [journal_unlock_journal_head+76/96] [journal_commit_transaction+1877/3408] [schedule+704/752] [kjournald+267/416] [co
mmit_timeout+0/16]
Dec 19 16:03:54 matrix kernel:    [kernel_thread+40/64]
Dec 19 16:03:54 matrix kernel:
Dec 19 16:03:54 matrix kernel: Code: 0f 0b 83 c4 14 39 33 74 2a 68 1c 33 22 c0 6

8 85 06 00 00 68