really annoying problem: hpacucli hangs on debian servers

[TanelKokk]

we have bunch of servers HP DL381-G1 (ca 30), all running debian sarge 3.1, kernel 2.6.18.8-smp, hpacucli-7.70, disk-controller SA-6400 with EM (fw version: 2.68).

hpacucli have worked fine approx. half a year on these servers. now, when uptime is about 169-174 days), hpacucli don't work anymore (at first on three servers). hpacucli (well, actually /opt/compaq/hpacucli/bld/.hpacucli) will hang with status "uninterruptable sleep" and anykind of disk monitoring is impossible. restarting hpasmd won't get any positive results. Only restarting servers solves problem and anything is fine again.

now there are again 4 servers, where hpacucli hangs on! What is going on? Is there some counter/buffer or something, what might overflow? I'm afraid, that I have to restart all my 30 servers during one or two week :( And this is nightmare!

Does anybody have experienced something similar?

Tanel

P.S. This thread has been moved from disk to ProLiant Servers (ML,DL,SL). -HP Forum Moderator

[James ~ Happy Dude]

Hello Tanel,

Indeed a very strange issue;
Y don't you try using the Latest HPACUCLI ??
http://h18023.www1.hp.com/support/files/server/us/download/27573.html

Let us know if this helped.

Regards,
James.

[TanelKokk]

I tried to use 7.85-18, but it didn't help. It hangs in the same way as 7.70 does.

couple of notes:
- hpasmd is nothing to do with newer hpacucli anymore (as I remember older ones (7.30 etc.) required hpasmd).
- I was able to bring about the same hanging situation by my own: I initialized 8 crontab processes, each running at '* * * * *' (once every minute). There were everything ok during 2 days. After 2 days hpacuclies hung with "uninterruptable sleep"