1748262 Members
3564 Online
108760 Solutions
New Discussion юеВ

Re: kernel panic

 
Alvin_14
Advisor

kernel panic

our servers keep hanging due to kernel panic.

from the /var/log/messages we get:

Jun 25 08:28:06 hostname1 kernel: st1: Error with sense data: Current st1: sense key Not Ready
Jun 25 08:28:06 hostname1 kernel: Additional sense: Medium not present
Jun 25 09:26:27 hostname1 kernel: st1: Block limits 1 - 16777215 bytes.
4 REPLIES 4
Matti_Kurkela
Honored Contributor

Re: kernel panic

At 08:28, someone tried to use a tape drive with no tape inserted. It's an error, but it should never be a cause for kernel panics.

At 09:26, the drive notifies the system for acceptable block size limits (for the current drive + media combination?). This message is not an error, just information.
Maybe someone has inserted a tape into the drive and is now starting a system backup?

Nothing here indicates any cause for a kernel panic.

A wild guess: does the kernel panic happen whenever you're taking backups?
If so, are you backing up the /proc filesystem?

/proc is a virtual filesystem which contains process information and kernel tunables. All the information in it is generated dynamically, so there is no reason to back it up. With some kernel versions, trying to back up /proc is known to cause kernel panics.

You should always exclude /proc from your backups. If you're using a 2.6 series kernel, you should exclude /sys too: it's another virtual filesystem.

MK
MK
Steven E. Protter
Exalted Contributor

Re: kernel panic

Shalom,

This kernel panic indicates a hardware problem of some sort.

Due to the tape drive being the trigger, check the scsi card, connectors cables and scsi ID setup for problems.

Test the hardware, perhaps with the vendor. If the tape drive is bad it could trigger this kind of problem.

SEP
Steven E Protter
Owner of ISN Corporation
http://isnamerica.com
http://hpuxconsulting.com
Sponsor: http://hpux.ws
Twitter: http://twitter.com/hpuxlinux
Founder http://newdatacloud.com
Ragu_3
Trusted Contributor

Re: kernel panic

The tape drive is going bad, may already have!

To escape from a kernel panic and get your machine to reboot automagically, add this into /etc/sysctl.conf

kernel.panic = 7 (after a 7 sec delay, kernel reboots)

Make sysctl re-read this by sysctl -p /etc/sysctl.conf
Debian GNU/Linux for the Enterprise! Ask HP ...
Gokul Chandola
Trusted Contributor

Re: kernel panic

Hi,
This is related to Tape drive, yes it may be hardware connection error.
Please check the stape status in Kernal, that shoud be static (Current value and Next Value).

Regards,
Gokul Chandola
There is always some scope for improvment.