Random Reboots - DL380 G3

[Nick_Beck]

Jason,

1. Memory not being identical in bank 3 & 4
2. Lack of memory to handle the jobs
3. C:\ drive has less than 1GB of free space available.

Point 1 is a possibility, we all know how manufacturers like to tell us memory should be identically matched.

Point 2 is unlikely, my 380 is currently running almost 40 people's mail boxes plus all the other stuff like McAfee GroupShield and AntiVirus etc on the backup machine of the RSO which only has a single GB memory, unlike the 4GB in the primary machine.

Point 3 is a possiblility - I had serious performance issues on this server when GroupShield suffered a file corruption on the detected items database and it almost completely filled the system drive. Reinstalled on a different partition and everything was fine again. However, I did get error messages citing lack of space when this was happening. Maybe the lack of space has caused it to frag up or something?

[Brado23]

As promised, I'm reporting back after setting a new uptime record on my server immediately after replacing the CPU's.

I'm not going to jinx myself and say the problem is fixed, but I'll state the facts and you can draw your own conclusions:

Hit 40 days uptime today which I had never been able to do in the 1 year 3 months I have had the server. Previous record was 38 days, 8 hrs which I achieved about a year ago. There was only 1 other time that I got over 30 days which was almost a year ago too. The most uptime I had in the last 10 months was about 24 days. In very recent times before the CPU replacement, I was lucky to get 2 weeks uptime without a reboot. I suspect the system was getting worse over time as the CPUs degraded.

That about sums it up. If things do go sour I'll be sure to report back but it's looking good at this stage.

[Brado23]

61 days uptime on my system today.

Jason, If you haven't done so already, I'd be trying to get CPU replacements from HP especially if the system is under warranty. Sounds like you have been through everything I have, and new CPU's was the fix for me.

[JoshIT]

hi, I have a DL 380 G3 with Windows 2003 SP2. It also is a backup exec 12d media server, connected to an HP robotic library. Recently it started the random reboot thing.

No pattern or anything, no blue screens and nothing useful from event viewer or the HP management log viewer.

After some research on this forum on the same topic, I went into the BIOS and disabled ASR.After that it has stopped rebooting so far (it's not been long though). For some reason I think it's something to do with backup exec because it all seems to have started after backup exec 12d live updates started happening (live updates were not working and only recently we had that fixed).

Any suggestions anyone can offer will be greatly appreciated. I didn't update the firmware or the PSP because no one has ever reported the problem getting resolved just by updating these.

Thanks very much, folks.

[Amaraa]

I had this problem before. As my remember i fixed so simple. My computer right click properties-Advanced-Start up Revocery Settings untick Automatically restart

[shawnn]

maybe the solution reformat the OS ?

[Aaron Cunningham]

It's not an OS problem. I've been chasing this issue for the last 6 months (driving me crazy!), and have taken much of the same corrective actions as everyone above (too much to list but, in short: new RAM, new drives, new power supply, extended ASR timeout value, disabled ASR, disabled hyperthreading, etc). No luck -seems to crash whenever there is a load on the system - no messages anywhere.

However, I've also seen the issue when I was in the BIOS utility - so I think we can rule out OS completely.

Server: Suse 10.3, DL380 G3 2 x Xeon 3.2gz, 4gb. Running Asterisk and VMware (Windows7 + Exchange)

There's a Dell 2650 for sale on Craigslist.. and my HP loyalty prevents me from buying it right now ....but the case is getting stronger.

-Aaron Cunningham

[Brado23]

I didn't see you list that you have replaced the CPU's. That fixed the problem for me. See my earlier posts. The only outages I have had on my box since replacing them have been my shutting the server down for maintenance. I have not had a single unexpected downtime since.

[Aaron Cunningham]

Thanks - I'll try replacing the CPUs, and will report back.


..wait a minute...

dl380:~ # dmesg | grep Xeon
CPU0: Intel P4/Xeon Extended MCE MSRs (12) available
CPU0: Intel(R) Xeon(TM) CPU 3.06GHz stepping 05
CPU1: Intel P4/Xeon Extended MCE MSRs (12) available
CPU1: Intel(R) Xeon(TM) CPU 3.20GHz stepping 05

..I never noticed that before, however, I'm certain that the BIOS reports the same speeds - because I've been watching it reboot for the last 6 months. I wonder if there is a way to set the CPU speed on the motherboard?

I'll report back after the CPU replacement.
-Aaron

[Aaron Cunningham]

I replaced the two small "power conditioner" boards and the system has been up for 4 days straight - a new record. I also have new CPU's (2.4's) and yet another power supply, in case this doesn't fix the problem, but so far so good.

The system still has the different speed CPU's (as mentioned above).

-Aaron