Digi Accelport XP causing Kernel Panics

[dahven]

I have a Dell PowerEdge 600sc server running RedHat 7.3.

Recently I installed a Digi Accelport XP 8 port internal adapter in the system for use with dumb terminals.

I installed 40002059_F.src.rpm driver according to the instructions given by the Digi techs.

Anytime more than one terminal was shut off, if a terminal was left off for awhile then turned on, or if a terminal was shut off multiple times we would get a Kernel Panic and the system would freeze requiring the power to be turned off which causes the database to crash.

After 2 months of phone tag, run around, and have to convince the office to not ever turn the terminals off.
I get a e-mail from the Digi techs saying that I don't have the most up-to-date driver installed.

So I uninstalled the old driver, installed the new driver

40002059_G.src.rpm

After the installation we tested the new driver, turning off one terminal at a time leaving it overnight and turning it off and on mulitiple times, then turning off multiple terminals overnight. Until we could turn them all off and have them come up to a login in the morning. For two weeks after testing everything seemed to be working fine.

Then two kernel panics in one day and one the next day all crashing the database on the server:

On the console:

code 8b 08 8b 58 04 89 c8 29 d8 8b 76 0c 89 75 f4 21 f0 39 c7 7e
[0] kernel panic;
aiee, killing interrupt handler
interrupt handler not syncing

(it was the same as what we had with the first driver)

Now the Digi techs are insisting that:

"Unfortunately, since we have not had any other reports of panics with this adapter in Red Hat 7.3, we will need to once again request the complete panic dump for analysis. There would be no other means for us to debug this panic without it."


Problem. The system is out of state and I don't have physical access to it and no one in that office is particularly technical.

I need a way of capturing the dump to disk/tape/file from within the system so I can take the file to the Digi techs.

Is there a tool that can do this for RedHat 7.3?

 

[Morsing]

I don't know what they're after but read /usr/src/linux/Documentation/oops-tracing.txt and the ksymoops man page.
I seriously doubt that they can use it for anything though??

Cheers Henrik Morsing
Certified AIX 4.3 Systems Administration
& p690 Technical Support

 

[dahven]

/usr/src/linux/Documentation/oops-tracing.txt is not on the system -- under the /usr/src is only the redhat directory.

and ksymoops needs the output of the oops to work. How do I capture that?

 

[dahven]

i recieved this from the redhat techs:

Try to go into recscue mode and check the logs in /var directory. You could give the messages, boot.log and dmesg files. These files are some enough to be copied in a floppy disk. See the URL below on how to go into the rescue mode.

http://www.redhat.com/docs/manuals/linux/RHL-8.0-Manual/custom-guide/ch-rescuemode.html

Once in the resuce mode run these commands

mount /dev/fd0 /mnt/floppy
cp /var/log/messages /mnt/floppy/
cp /var/log/boot.log /mnt/floppy/
cp /var/log/dmesg /mnt/floppy/

The you can now send those files to Digi techs.


Would this work? I though that when the system was rebooted after a kernel panic. The panic dump is lost.