I have a DB server running ext3 under fairly extreme load. The
system is a Dell 2550 (Dual P-III 1Ghz with a gig of RAM) running RedHat 7.2
with the stock 2.4.7-10 kernel. Things have been swell for 95 days, but I
just got bit:
Linux 2.4.7-10smp #1 SMP Thu Sep 6 17:09:31 EDT 2001 i686 unknown
[root@ root]# cat /root/oops.20020817
kernel BUG at journal.c:373!
invalid operand: 0000
CPU: 1
EIP: 0010:[<f88454c4>]
EFLAGS: 00010282
eax: 0000001d ebx: 00000000 ecx: c026059c edx: 00007d98
esi: d9a0da60 edi: dfee7f10 ebp: c960dc00 esp: f7de9e44
ds: 0018 es: 0018 ss: 0018
Process kjournald (pid: 152, stackpage=f7de9000)
Stack: f884a2e1 00000175 c01537e4 f71a6af8 00000000 00000000 dc13d310
d9a0da60
dfee7f10 c960dc00 f8842c1e d9a0da60 dc13d310 f7de9e9c 0000190c
c960dce4
00000001 00000ff4 c8d6700c 00000001 d9a0da60 ed9e64e0 dfee7bb0
c26af0a0
Call Trace: [<f884a2e1>] [<c01537e4>] [<f8842c1e>] [<c01c8afc>] [<c01c8afc>]
[<c01c8c89>] [<c01ccdc2>] [<c0105b2d>] [<c01180fc>] [<f8845306>]
[<f88451a0>]
[<c0105836>] [<f88451c0>]
Code: 0f 0b 5e 5f 8b 7c 24 28 8b 4f 0c 85 c9 74 2d c7 44 24 0c 01
Is this a known bug? I'm not finding anything in the errata. I have
a different machine running a stock 2.4.18 kernel, and I known about the
RedHat bugs in 2.4.18-3. Does the stock 2.4.19 kernel represent the most
uptodate and stable version of ext3?
Many thanks,
-poul
