server (almost) dies

[Joris Dobbelsteen]

>-----Original Message-----
>From: ubuntu-users-bounces at lists.ubuntu.com 
>[mailto:ubuntu-users-bounces at lists.ubuntu.com] On Behalf Of 
>Luca Ferrari
>Sent: donderdag 2 augustus 2007 17:19
>To: ubuntu-users at lists.ubuntu.com
>Subject: server (almost) dies
>
>Hi all,
>ubuntu server 6.1, sometimes it seems to die, that is it 
>repond to pings, but no way to enter either via ssh or thru 
>the console. In the log I found the following messages:
>
[reformatted]
>Aug  2 11:18:41 backup kernel: [43102179.570000]  
><c0147e3c> softlockup_tick+0x9c/0xf0  
><c012aff1> update_process_times+0x31/0x80
>Aug  2 11:18:41 backup kernel: [43102179.570000]  
><c02998f0> tcp_ioctl+0x0/0x180  
><c0113cd5> smp_apic_timer_interrupt+0x55/0x60
>Aug  2 11:18:41 backup kernel: [43102179.570000]  
><c0103acc> apic_timer_interrupt+0x1c/0x24  
><c02998f0> tcp_ioctl+0x0/0x180 
>Aug  2 11:18:41 backup kernel: [43102179.570000]  
><c02998f3> tcp_ioctl+0x3/0x180  
><d05c6a82> server_sock+0x32/0x70 [smbfs] 
>Aug  2 11:18:41 backup kernel: [43102179.570000]  
><c02b4c7a> inet_ioctl+0x3a/0xb0  
><d05c7062> smb_recv_available+0x32/0x50 
>[smbfs] Aug  2 11:18:41 backup kernel: [43102179.570000]  
><d05c8fe7> smbiod+0x1a7/0x220 [smbfs]  
><c0135020> autoremove_wake_function+0x0/0x50
>Aug  2 11:18:41 backup kernel: [43102179.570000]  
><d05c8e40> smbiod+0x0/0x220 [smbfs]  
><c0101005> kernel_thread_helper+0x5/0x10
>
>
>Is this related to the problem? Any idea about?

If you did happen to manually reboot the system (reset button or power
cycle) its not very unlikely. I'm no expert, but I don't wanna see any
kernel related things in the fashion as above (looks like stack traces).
It mostly seems to have something to do with the network subsystem and
the smb driver (samba/windows file sharing), but I'm not sure about it.

In order to get on track:

* Are there any messages above or below these that seem related? This
looks like a small part of a dump of something, probably the kernel.

* Perform a memory test, let memtest86 run for a while. Generally its
sufficient to let is end a few tests, I don't think you need to complete
all of them. If it doesn't give any errors its very likely you memory
works as expected. If not, this could be a problem.

* How old is you hardware? Did you have problems before?
Old hardware might begin to wear out, brand new hardware might have some
problems that are not discovered after production.

* Perhaps there are some problems or imcompatibilities with Ubuntu 6.1.
Perhaps you can use 6.06 LTS for the server, which was very stable and
will be supported for a longer time. However I'm not sure if this will
change anything.


- Joris