K460 dual proc one proc is missing after reboot

[Thomas Gabel]

Hi,

my K460 is booted every night before backup via cron (shutdown -r). Sometimes it comes up with only one of the two processors active. First not very often, perhaps once every 3 month, now getting more and more often, at last 2 times a week. I can't find any error-messagees in the logfiles (Maybe i looked in the wrong ones).
To get the 2nd proc to work again, i have to switch off the machine. reboot only doesn't help.
Has anybody had the same problem? Or does anybody know a solution?

Thanks and best regards,

Thomas Gabel
Germany

[Pedro Sousa]

If you can, you should do a reboot and stop it in the Main Menu. Then you can see if it sees the 2 processors.

Another thing is, you should install OnlineDiagnostics to get some more feedback into the syslog, and to check it yourself running 'stm'.
good luck.

[Insu Kim]

I guess that cpu was disabled for some reasons like HPMC.
Try to see if HPMC was occurred.
Take a look at the file /var/tombstones/ts99 and identify that it has a meaningful values and timestamps in it such as 0x20b0.
If so, place a hardware support call to hp.

Best regards,

[paul courry]

A HPMC will stop the music right quick. Machine shuts down. Period. Only a LPMC can be recovered from.

It sounds more like a flaky CPU or backplane. Time to start documenting the problem and get ready to call HP.

To pull the HPMC info the following UTL and below that, the document Id go to the following site. How to pull the HPMC is at the end of the document.

http://us-support.external.hp.com/cki/bin/doc.pl/sid=7fc8cbf30756a5322c/screen=ckiDisplayDocument?docId=200000052636575


HP document Id: PAM39EF55E1

[Patrick Wessel]

I sneaked in after a long time just to look how the forum is doing?

Thomas, make sure your PDC version is at least 39.12 Earlier versions had a bug that influenced the driver of the runway bus to which the CPUs are connected.

If STM is installed on you system you can get the PDC version by:

echo "selclass qualifier cpu; info; wait; il" | cstm | grep Firmware

[Paula J Frazer-Campbell]

Hi Thomas

My point is why you should wish to reboot your server every night?

I run a k460 which runs a universe datbase and has been up for 14 months without any problems.

Unless you have problems with things like memory leaks your server should stay up very happily providing that you keep it tidy - clear log files etc.

These are threads on that subject.


http://forums.itrc.hp.com/cm/QuestionAnswer/1,1150,0xf679dfe5920fd5118fef0090279cd0f9,00.html

http://forums.itrc.hp.com/cm/QuestionAnswer/1,1150,0xefba6af52b04d5118fef0090279cd0f9,00.html



Paula

[Thomas Gabel]

hi Paula,

i have to reboot, cause some apps (and users) won't close their files correctly, so my backup would be corrupted.

Thanks for help,

Thomas

[Vincenzo Restuccia]

Change the processor.