HPE 9000 and HPE e3000 Servers
cancel
Showing results for 
Search instead for 
Did you mean: 

Chasis Logs Filling Up

 
SOLVED
Go to solution
Scott Frye_1
Super Advisor

Chasis Logs Filling Up

Has anyone ran into a problem with their chasis logs filling up. We received an EMS message telling us "The Chassis Code output FIFO is full". Basically, it is saying either the GSP is not able to read them fast enough or the log is missing.

One thought was there maybe there is a parameter we can set so this log doesn't fill up. The other is maybe we just go and clear the chasis logs.

Any thoughts?

Scott
3 REPLIES 3
Steven E. Protter
Exalted Contributor

Re: Chasis Logs Filling Up

I'd go through the GSP setup (Ctrl-b on console) and run through the logs. SL is the command I think.

There may be information in there that you want to deal with.

You may have a GSP card thats getting ready to go or another problem revealed by the actual entries. There is a procedure to clear those logs.

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
Andrew Merritt_2
Honored Contributor
Solution

Re: Chasis Logs Filling Up

Hi Scott,
What type of system are you seeing this with?

There is a known problem with the 1.x Firmware for the rp8420 which has the same symptoms. If your system is an rp8420 (or possibly also a similar system), in the MP logs you would expect to see entries like this:

IPMI event hex: f6800ad500e00000 0
Time Stamp: Wed Jul 7 07:57:37 2004
Event keyword: WATCHDOG_RESET_PARTITION

If that's the case, then the answer is to upgrade to Version 2.0 of the Firmware. There is a further known problem where the chassis logs get full with minor events, such as OS heartbeats; that will be fixed in Version 3.x.

Andrew
Scott Frye_1
Super Advisor

Re: Chasis Logs Filling Up

This is on an rp8420. Thank you so much! We will try the code upgrade to see if that fixes our problem.