Server (ML350 g5) reboots every weekend at random time

[dahund2]

Suddenly after 2 years of running perfect our server reboots at random. The system is (ESXI 4.0) which logs this, what i dont get is that it looks like it only happens in the weekend:

Apr  1 02:43:37 Hostd: [2012-04-01 02:43:37.027 2ADD1B90 verbose 'vm:/vmfs/volumes/4ce92ed1-ccd802da-a138-002655de9d35/Server 2008_drift/Server 2008_drift.vmx'] Actual VM overhead: 210427904 bytes
Apr  1 02:43:37 Hostd: [2012-04-01 02:43:37.028 2ADD1B90 verbose 'vm:/vmfs/volumes/4ce92ed1-ccd802da-a138-002655de9d35/VMASSBS01-ny/VMASSBS01-ny.vmx'] Actual VM overhead: 658477056 bytes
Apr  1 02:43:37 Hostd: [2012-04-01 02:43:37.029 2ADD1B90 verbose 'Vmsvc'] RefreshVms updated overhead for 2 VMs
Apr 1 02:46:06 vmklogger: Successfully daemonized.
Apr  1 02:46:06 vmkernel: TSC: 0 cpu0:0)Init: 293: cpu 0: early measured tsc speed is 2000070975 Hz
Apr  1 02:46:06 vmkernel: TSC: 18504 cpu0:0)Cpu: 346: id1.version 1067a
Apr  1 02:46:06 vmkernel: TSC: 34398 cpu0:0)CPUIntel: 361: C1E enabled by the BIOS
Apr  1 02:46:06 vmkernel: TSC: 39978 cpu0:0)Cpu: 408: APIC ID mask: 0xff000000

  In ILO2 i have this :

Informational

iLO 2

04/15/2012 01:19

04/15/2012 01:19

1

Server power restored.

Informational

iLO 2

04/15/2012 01:19

04/15/2012 01:19

1

Server power removed.

ILO dosnt show anything else, first we suspect a heat problem and did some cleaning, but that didnt help! second we disabled all C-states an set the server to run ” HP Static High Performance mode” but that didn't help.

Is this a PSU problem ?? can anyone help, we are getting some database issue now because of all the reboots :-) and we dont want anymore random reboots :-)

/regards A

[Dennis Handly]

Do you have any other servers that lose power at the same time?

Do you have a UPS?

[VBO]

Hello,

Same problem into 1 of us 3 ML350 G5 From 2009.

Contacted HP, and HP send me a PowerSupply. But i think this is not the power supply, why we add redundant power supply in 2010 and changing nothing.

The reboot come into the night (Backup - High HDD Load).

No error can be see into the IML.

[dahund2]

We have a UPS! but no redondant PSU :-( it happens at random, even in ours of none activity.

[dahund2]

SRY, we dont have ekstra server attached. But now that when you say it we have a managed switch (maybe that one have a log)

[dahund2]

Switch have been running for 285 days :-) so i dont expect it to be the UPS.

[dahund2]

@VBO wrote:

Hello,

Same problem into 1 of us 3 ML350 G5 From 2009.

Contacted HP, and HP send me a PowerSupply. But i think this is not the power supply, why we add redundant power supply in 2010 and changing nothing.

The reboot come into the night (Backup - High HDD Load).

No error can be see into the IML.

---

So did you ever fix it ?

[VBO]

Actually, no. But having a open case into HP Support. I give you feedback when i have a solution.

For information, if i run HPReport, the server reboot (reset). I go try to re-run it after the PSU "defective" (for me not defective) and i say you if the problem come from the PSU.

Best Regards
Vincent

[dahund2]

I read that someone had good experience with disabling ASR in the bios, did any of you try that ?

[dahund2]

WTF!! i read these comments from other thread: HP forum 

" the Technical support person at the other end informed me that the P/S's on the ML-350 sever's are very sensative and are not compatible with APC UPS and that I need to purchase an HP UPS to fix this problem."

"The APC 1500 UPS is well known for causing rebooting and power down issues in several G5 models. We have fixed tons of issues that even after replacing parts nothing seemed to work. Everything was resolved once the customers decided to stop using the ACP1500 UPS"

We have a APC 1500 UPS, can this really be the problem??? and if thats the case what kind of PSU's did HP put in the ML350 G5 since it can only run with a HP UPS?? Does anyone else have this problem?

(by the way checked the BIOS and firmware yesterday all of them is the newest version, but it is still rebooting every weekend!!

Only thing left is the UPS and PSU......