IPMI reboots DL360

[Joshua Pott]

Ok, you've nailed my setup. I am running a G5 with a P400i scsi controller, and to answer your question I have not opened a ticket.

But as I mentioned before I have two DL360 G5's. Both are running the same driver. Sorry to sound like a noob. But are you saying we have bad hardware?

[Blazhev_1]

Hi,

can't tell you for sure, but I don't think it is a hardware error, can be a bug in firmware or something like this.

If you have the actual firmware, system rom & latest PSP and so many people see the same error 0x0000002D in IML then seems either like a revision problem or firmware problem...

If the problem is only that the servers reboot without this 0x2d in IML, then the G5's have known issues with the PSUs and this is maybe the cause(but not the DL360 G5).

You can boot smartstart and try an offline hardware diagnosic to check if something will come up, bu I doubt.

If we assume it is storage related issue, storport plays role here :

http://support.microsoft.com/kb/941276

As for stopping the reboot, you can disable ASR from RBSU->Server Availabiity->ASR Status.
This can stop the server from rebooting or will show bluescreen pointing to the problem(for windows)

I don't think 0x2d is logged in IML by Linux/ESX users, so there the agents 8.0 and IPMI deinstall/ASR stopping will solve the problem.

Pac

[Blazhev_1]

About SP1 and SP2 thing, the main "problem related" difference is storport again...

http://support.microsoft.com/kb/946448/en-us

here is a newer version, you can give it a try it solves a lot of problems with hangs and bluescreens.

[Blazhev_1]

Sorry, me again :)

is there a dump file saved in c:\windows\minidumps or if you search c:\windows for *.dmp?

Someone analyzed the minidump file or if he can attach it if it is less than 1MB?

[Joshua Pott]

I appreciate your help with this. However it appears that it has writen a dump file. Is there something I need to enable on the ILO to get it to spit that out?

[Blazhev_1]

so there is or there is no dump?

the function should be enabled by default.
cotrol panel->system->Advanced options->Startup and recovery-settings.
under write debiggung information should be chosen minidump f.e.
from this screen you can disable the server restart too...

Can be that the server crashes without dump, but there should be a dump

[Joshua Pott]

Ok, had to change it from Kernel dump to small memory dump. We'll see what we get.

[rafaelito]

Some more log information:

May 8 13:01:30 farallon hpasmxld[4446]: OsKcsExecCmd: IPMI NetFN 0x36 CMD: 0x2 has timed out!
May 8 13:01:40 farallon hpasmxld[4446]: OsKcsExecCmd: IPMI NetFN 0x36 CMD: 0x2 has timed out!
May 8 13:01:50 farallon hpasmxld[4446]: OsKcsExecCmd: IPMI NetFN 0x36 CMD: 0x2 has timed out!
May 8 13:02:00 farallon hpasmxld[4446]: OsKcsExecCmd: IPMI NetFN 0x36 CMD: 0x2 has timed out!
May 8 13:02:00 farallon hpasmxld[4446]: iLO 2 Communications Error - Attempting synchronization!
May 8 13:02:45 farallon hpasmxld[4446]: iLO 2 has responded to reset request . . .
May 8 13:02:45 farallon hpasmxld[4446]: Stopping the Watchdog Timer . . .
May 8 13:02:45 farallon hpasmxld[4446]: Resetting Internal Data structures . . .
May 8 13:02:45 farallon hpasmxld[4446]: Initializing Internal Data structures from iLO 2. . .
May 8 13:02:45 farallon hpasmxld[4446]: The iLO 2 reset / synchronization has completed successfully
May 8 13:02:45 farallon kernel: hpasmxld[4446]: segfault at 0000000000000031 rip 0000000000000031 rsp 00007fff1b307168 error 4

HP Debian sowtware:
hpasm-7.8.0-91.etch26.amd64.deb
hp-OpenIPMI-7.8.0-108.etch26.amd64.deb
hprsm-7.8.0-104.etch26.amd64.deb
cmanic_7.9.0-5b.etch_amd64.deb
hpsmh-2.1.7-167.debian.amd64.deb

[Fabio_S]

@ Pac: thanks fot the tips. I just filled the form needed to request the patch to Microsoft. As soon I have it I'll install and check if the issue is solved.
Indeed my situation reflects the one described in the link you posted (WS2003 x64, 6GB RAM, P400I storage controller).

Regarding the ASR disabling, would it only disable the reboot, but system crash would occur anyway?

Last thing, how do I launch RBSU? system start, before Windows loading?

thanks

[Blazhev_1]

Hi,

the ASR disable, will not stop the system to crash, it will maybe point you to a the problems if it shows a bluescreen.

You can access RBSU when you start the server and press F9 when prompted.
Same thing you can do withot downtime from windows : cotrol panel->system->Advanced options->Startup and recovery-settings - UNCHECK the Automatically reboot checkbox.

@rafaelito :

Check :

http://h20000.www2.hp.com/bizsupport/TechSupport/Document.jsp?lang=en&cc=us&objectID=c01330219

although you don't have RHEL5...

For the IPMI reset, Server BIOS/iLO2 Firmware and E200/P400 controller firmware are important...

Pac