Topic: kernel panic

[Kees Vonk]

We recently had a planned power outage, but as I was on holiday they forgot to switch of the server (SME 5.6), ever since it occassionally crashes with the following screen on the console:



Process swapper (pid: 0, stackpage=c02e1000)
Stack: c01f9e61 00000000 c02ca18c 00000020 00000020 c10d64b0 00000000 00000020
       c013256d 00000020 c012e62d 00000003 0000003a c2159980 4202010a ff02010a
       4202010a c10d64d8 00000246 c10d64b0 00000020 c012e860 c10d64b0 00000020
Call Trace: [] udp_queue_rcv_skb [kernel] 0x1b1
[] __get_free_pages [kernel] 0xd
[] kmem_cache_grow [kernel] 0xbd
[] kmem_cache_alloc [kernel] 0xc0
[] alloc_skb [kernel] 0xc8
[] ei_receive [8390] 0x1a1
[] update_wall_time [kernel] 0xb
[] ei_interrupt_Rb93d8fa5 [8390] 0xee
[] handle_IRQ_event [kernel] 0x3a
[] do_IRQ [kernel] 0x6d
[] stext [kernel] 0x0
[] default_idle [kernel] 0x0
[] stext [kernel] 0x0
[] default_idle [kernel] 0x24
[] cpu_idle [kernel] 0x20


Code: 8b 8b e0 00 00 00 85 c9 75 13 68 77 01 00 00 68 e9 82 22 e0
 <0>Kernel panic: Aiee, killing interrupt handler!
In interrupt handler - not syncing



Then when I reset the machine (using the reset button) it starts but ends up hanging with the following message:



esi: c10c710c   edi: c3d498d4   ebp: 00000008   esp: c3d75f10
de: 0018   es: 0018   ss: 0018
Process insmod (pid: 11, stackpage=c3d75000)
Stack: 48485600 00000001 00000000 00000000 c3d49820 c3d74000 c3d75f30 00000002
       00030002 00000000 c118e2a0 ffffffea 0000b010 c0129efe c118e2a0 c118e2c0
       c3d75f5c c0129dd0 00000000 00007fc0 00003050 4001cfc4 00000000 00000000
Call Trace: [] generic_file_read [kernel] 0x7e
[] file_read_actor [kernel] 0x0
[] sys_read [kernel] 0x96
[] sys_lseek [kernel] 0x6e
[] system_call [kernel] 0x33


Code e9 ea fe ff ff 8b 74 24 10 8b 46 44 8b 56 48 0f ac d0 39
 ERROR: /bin/insmod exited abnormally!
Mounting /proc filesystem
Creating root device
Mouting root filesystem
<3>kmod: failed to exec /sbin/modprobe -s -k ext3, errno = 2
kmod: failed to exec /sbin/modprobe -s -k freeext3, errno = 2
mount: error 19 mounting ext3
pivotroot: pivot_root(/sysroot,/sysroot/initrd) failed: 2
Freeing unused kernel memory: 292k freed
Kernel panic: No init found. Try passing init= option to kernel.



However when I switch the machine off and back on it works (until the next time it crashes). The crashes seem to be unpredictable and the server can be running anything from half an hour to more than a week. Can anyone tell me where to start looking for the fault? I suspect it will be hardware, as this started after a power failure.

Any help / pointers are appreciated,


Kees

[Scott]

When I reinstalled sme5.6 I had the same problem.
The kernel panic was due to a difference in the
amount of ram that was available, and the amount
that the default instalation was claiming existed.

Real strange in my book. However, if you
do a search on the site using kernal panic
as your search key, and insure that all
previouse dates are selected in the drop down
menu, you will find the post that tells you
how to eliminate the error when installing.

My guess is that the you may have some
faulty ram, or the file is corrupt in some way.

In any case, I would start to look for the file
that tells the os how much ram is available.
I'm no expert, real raw with sme. But it does
stand to reason that if upon installing the
server, one can get a kernal panic simply
because there is more memory than what
is specified in the configuration file, then
this would be the first place to look.

I can't tell you what files to look in, I'm way
to green and can only go based on my little
experience and some logic. And of course,
being logical doesn't necessarily mean I'm
going in the right direction. But it is a start.

Have you tried the reinstall disk? But then,
that may not be to smart. Like I said,
I'm to knew at Linux and SME to give you
a whole lot of direction.

[ChocoKiller]

Hi,

Sometimes, it's more simple than anything expected. I was having nearly the same error, but the only reason for this is weather. My poor server were too hot, and only this. After buying some extra case and hard disk fans, all seems to be good ... )

Hope this helps ...

[ryan]

After almost a year, I nailed down random infrequent kernel panics on my SME servers.  I have eliminated the kernel panics by removing old SMC 10 mb pci nic cards I was using for the internet nic ( I have lots of these lying around).  I think my problem had to do with IPSEC traffic through these old nic cards.  All old nics where replaced with new intel epro100 cards...been stable since.  Note that SME and IPCop server where experiencing kernel panic with these old nics.    

If your using old 10mb nic cards, maybe try newer 10/100 cards.

ryan