Latest HP Agents and mysterious reboots...

[waaronb]

I had a weird issue today with one of my DL380p Gen8 systems.

It's running Windows Server 2012 and had all the latest HP software from the 2014.02 SPP.

So here's the weird part... over the past week or two since I installed that SPP, this server has mysteriously rebooted a few times.  Nothing is logged in the IML, the Windows event log shows nothing except the fact that it rebooted for some unknown reason, etc.

In other words, there is absolutely nothing to indicate why this happened.

Then today it got even worse... the server rebooted and it would come back up, then reboot again after about 1 minute.  I could sometimes login but then it would reboot again before too long.  At some point, Server 2012 recognizes that it's in a reboot cycle and goes into it's recovery screen, which is where I found it when I first got wind that it was down.

I booted into safe mode with networking and everything was good that way, so I checked a few things and did a normal bootup... once again, it just reboots a minute or two after coming up.

So, back to safe mode.  I set the HP agents to manual start instead of automatic, did a reboot, and what do you know, it was working great after that.  So that was the insight agents, nic agents, etc... the whole set.  Basically any service that has "HP", I kept it from auto-starting.

This server also had the agentless service installed... I don't know why I have both, but I do on my gen8 servers and it hasn't caused problems on anything else.

I would have worked on it more but I only have so much time in a working day to deal with issues like this, and we were due to start upgrading servers to 2012 R2, so I backed up the Hyper-V virtual disks this server hosts and just blew everything away, did a fresh install.

It's working fine now, once again with all of the HP stuff installed, but rather than doing the usual HP Agents *and* the agentless thing, I've only got the agentless stuff installed.  Which basically narrows it down to the NIC, Server, Foundation, or Storage agents that come with the old agent install.

This server had been working just fine for the past 4 months prior, and besides the normal monthly Microsoft updates (which I only bother with every few months anyway), these new HP agents were the only other change.  And I do have other Gen8 servers that are just like this one but they're fine.

Has anyone else experienced anything quite like that?  Odd, unexplained reboots or anything that you thought might be related to the HP agents?  Now I'm worried about my similarly configured servers and whether they might be in danger.

[waaronb]

Oh, it may be worth mentioning... this server is also one that I had to replace the motherboard shortly after we got it because one of the DIMM sockets was bad.

And it also has something wrong with it's drive backplane because with both the old and new motherboard, one of the genuine HP drives (in bay 8) always says it's not able to be authenticated as genuine during POST, so it throws an error. Even with different hard drives that work great in other systems, that bay 8 always gives that error.

Next time I'm on site, I'll reopen the case I had with HP for the motherboard issue and replace that cage backplane as well, but for now, it's not hurting anything except giving me an annoying warning.

So who knows, maybe it's something like the HP Storage Agent freaking out because of that, or some other related issue.