Ramdom Reboot Windows Server 2008 R2

Hi,

A reasonably simple case of one random reboot on one of our Windows Servers (2008 R2).
The server runs Hyper-V with 4 virtual machines and is a HP Proliant DL380 G7.
Everything is fully patched and the server runs NO other software apart from the Hyper-V role and the HP management web application. Around 3 days ago the server randomly rebooted with the following errors (attached jpeg). I've done tons and tons of reading and it seems that it a hotfix might be needed but has anyone experienced such an issue. It has only happened once but the Windows Server 2008 R2 has recently been installed (2 weeks ago) The articles I have read are located

here:
http://social.technet.microsoft.com/Forums/en-US/windowsserver2008r2general/thread/281d43e9-5f42-45c5-9c90-692de8ba27a4

and here:
http://support.microsoft.com/default.aspx?scid=kb;en-us;976586&sd=rss&spid=14481

Error.jpg
dqnetAsked:
Who is Participating?

Improve company productivity with a Business Account.Sign Up

x
 
Joshua1909Connect With a Mentor Commented:
Is the server a domain member?
Are there any other errors or warnings in your log--even if they do not correspond to the time of  the reboot?

Do you have any peripherals attached?

You don't have to use Windows' memory utility, download http://www.memtest.org/ and burn ISO, although you're right it's not likely to be a memory error.
0
 
pvlierCommented:
Could be a faulty memory module. Did you run a memtest?
0
 
dqnetAuthor Commented:
I doubt it is faulty memory as many others whilst doing a google search seem to have experienced a similar case.
Memory Test doesnt exist in the server version of Windows, only Windows vista and 7.
0
Creating Active Directory Users from a Text File

If your organization has a need to mass-create AD user accounts, watch this video to see how its done without the need for scripting or other unnecessary complexities.

 
dqnetAuthor Commented:
The server hasnt yet been joined to the domain but uses DNS servers that are part of the domain.
It will soon be joined to the domain once all the VM's have been completely configured..

i've looked at all things before and after.. nothing strange.. services all started up correctly after a reboot and a message on the VM's said this os shut down unexpedtly... so the VM's were all up and the parent os crashed.

I really dont think its a memory problem. I mean the only thing that might cause it to be 'something of a memory' problem is the westmere xeon cpu's like to talk to a 3 dim configuration each side.. our server has 9 dims each side and we have populated each side with 4 giving us a total of 16gb... so its not the ideal configuration for the proliant dl 380 g7.. 2 gb * 4 gives 8 gb per side... the ideal configuration would be 2 * 3 each side giving a total of 12gb or 2gb * 6 each side giving 24gb.
0
 
cshepfamCommented:
I have a setup similar to yours and that happened to me as well.  The problem was that I didn't disable Windows Update and that was what caused my server to reboot.

You may want to check to make sure it's disabled.
0
 
OriNetworksConnect With a Mentor Commented:
I actually had a similar setup and problem as well except with a g4 server. The problem actually went away on its own after occuring 4 or 5 times over the course of a few months. I think a windows update patched the problem because itsnow been more than a year without a problem. Were all windows updates installed on the server before the problem?
0
 
pvlierConnect With a Mentor Commented:
I've had a faulty memory module in a dl360g4 server that caused the server to crash and then it was restarted by Automated System Recovery. In your case the faulty module may cause hyperV to crash which in turn crashes the server. I would definately run a memtest.
If that doesn't resolve the problem you can use the HP manamagent agents to see if there are alerts on any sensors or the ASR log to see if the server was rebooted because of an error (overheating?).  Have you installed the agents (smartstart cd or afterwards)?
The proliant support pack and Insight Manager can be found here at the bottom:
http://h20000.www2.hp.com/bizsupport/TechSupport/SoftwareIndex.jsp?lang=en&cc=us&prodNameId=4091432&prodTypeId=15351&prodSeriesId=4091412&swLang=13&taskId=135&swEnvOID=4064
 
0
 
dqnetAuthor Commented:
OK, the last few days seem to be fine and the server stable.
I' havent had the chance to look at the ASR log but yes, we install all servers with smartstart then install the webm extensions for all sensors. So, i'll be sure to check this tomorrow.

I'll also run a memory test to see if that brings up any issues. Like I said, those 4 dims per side is a little worrying and I think that '....might....' have something to do with it.. although its not a supported configuration it shouldnt cause issues. (as far as I know).

regarading windows update, yes it was set to automatic. However that wasnt an automatic update restart because of two reasons. 1) no service was stopped gracefully and 2)when the server came back to life all the VM's had the usual 'Windows did not shutdown correctly' box... windows normally suspends the VM's to reboot for any reason including windows update as far as I am aware?

Any clues as to the best memory test application?


0
 
kevinhsiehConnect With a Mentor Commented:
There are many hotfixes for Hyper-V R2 that are hard to find and don't show up in Windows Update.  You can find a list of them here.

http://blogs.msdn.com/b/virtual_pc_guy/archive/2010/08/09/important-hotfixes-for-hyper-v-on-windows-server-2008-r2-reminder.aspx
0
 
pvlierCommented:
For a memtester: I always use Memtest86+: http://www.memtest.org/, an advanced version of memtest86.
0
 
dqnetAuthor Commented:
right, a small advancement.. i've had a look at the hp management and there were a couple of post 207 error's... they all were from a long time ago and some are saying check dimm 1 and the operating system may not be able to see all the ram in this machine. The os is definitely seeing all the ram no problem. Let me run that memory test tomorrow and update you with my results... the dates on those memory problems were long ago and on the night of the windows crash there has been no logs whatsoever. generally speaking the logs will just overwrite the older ones if it gets full right? I dont want to be missing new events because of this?
0
 
pvlierCommented:
You can set te maxsize of the logs higher or export them and clear afterwards
0
 
Joshua1909Commented:
But yes they will just overwrite themselves when they reach maximum size--starting with oldest entries first.
0
 
dqnetAuthor Commented:
Right - the test has started and I hope to have an answer for you all today (if its a memory issue)
0
 
pvlierCommented:
Hi dqnet, which program did you use?
0
 
dqnetAuthor Commented:
I used the memory test application at www.memtest.org as per earlier posts.
It run for a whole night without a single error. I did it in ECC mode too...

Any other advice?
I'm begining to think it was just a one off??
0
 
Joshua1909Commented:
It could be any number of things. No way to tell if it's once of if we don't narrow down the cause.
Really the best bet is to read all event logs around that time--informational ones included. And not just in the immediate time period.
Once-of things can be the hardest to narrow down!
0
 
dqnetAuthor Commented:
Ok guys, i've been watching the server recently and it seems rock solid.
I will re-open the question should things change.
I dont really know where to assign the points so I guess I am going to split them.
Thanks to all who inputted essential valuable info.
0
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

All Courses

From novice to tech pro — start learning today.