Hi *,

i'm new to linux admin, and just had this problem on one of my machines. The oracle 9 DB was unreachable, so someone rebooted the server, after everything was fine. Server ran ok for a few years now, not seen this problem before. However , on checking the messagesfile, i saw the error:

sysctl: error: unknown error 22 setting key 'kernel.shmmax'

System is Red hat AS 4.

Question is how to diagnose the "kernel BUG at kernel/exit.c:840!".

Errormessage in messagesfile during the lost connection:

Dec 19 11:31:17 localhost kernel: ------------[ cut here ]------------
Dec 19 11:31:17 localhost kernel: kernel BUG at kernel/exit.c:840!
Dec 19 11:31:17 localhost kernel: invalid operand: 0000 [#1]
Dec 19 11:31:17 localhost kernel: SMP
Dec 19 11:31:17 localhost kernel: Modules linked in: parport_pc lp parport autofs4 sunrpc dm_mod button battery ac md5 ipv6 u
hci_hcd ehci_hcd hw_random tg3 sg st ext3 jbd cciss mptscsih mptbase sd_mod scsi_mod
Dec 19 11:31:17 localhost kernel: CPU: 5
Dec 19 11:31:17 localhost kernel: EIP: 0060:[<c0122c26>] Not tainted VLI
Dec 19 11:31:17 localhost kernel: EFLAGS: 00010246 (2.6.9-5.ELsmp)
Dec 19 11:31:17 localhost kernel: EIP is at do_exit+0x3b6/0x3c0
Dec 19 11:31:17 localhost kernel: eax: 00000000 ebx: e0c657f0 ecx: e0c65200 edx: c3c69d60
Dec 19 11:31:17 localhost kernel: esi: f7ffb680 edi: e0c652b0 ebp: 00000000 esp: e4565fdc

TIA,
H.

At this point, continue to gather information. Look at previous logs to see if there are any other occurrences of this or any similar problem. If the system has been running "for years," without change, it could well be somehow related to a hardware problem, such as memory or motherboard. In fact it is fairly unlikely to be anything else...

Ok, tnx for the info!