Resolving Bluescreens in Windows 2008 R2 on Nehalem Processors

Hotfix Solves Clock Interrupt Time Interval Error

Current Job Listings

I was having nothing but problems with a new server, the system wouldn’t last more than 2 minutes without bluescreening, a phenomenon I hadn't seen since the old Windows NT days, I was having very bad flashbacks....  I did the typical problem solving by flashing the system bios with the latest bios the hardware vendor had on their site, unloaded anything slightly squirrelly I had loaded up on the system, stripped it down to just Windows 2008 R2, but still, within a couple minutes while my HyperV guest sessions were loading up, the system would bluescreen.  Got the point where I was pretty certain it wasn’t my configuration…

I stumbled across a Microsoft knowledgebase KB article that wasn’t exactly what I was having the problem with, but pretty close.  Basically there is a “known bug” with Windows 2008R2 with Nehalem processors (which the system I was working on had an Intel Nehalem i780 processor).  The KB article is http://support.microsoft.com/kb/975530 which I downloaded the hotfix (had to do it in Safe Mode with Networking as I couldn’t even keep the server going long enough to download the hotfix without bluescreening) finally got the hotfix loaded, thing has run like a champ ever since…

While the KB article refers to a slightly different bluescreen error, the error I was getting was:

A problem has been detected and Windows has been shut down to prevent damage to your computer.

A clock interrupt was not received on a secondary processor within the allocated time interval

(etc, etc)

*** STOP: 0x00000101

I've setup hundreds of Windows 2008 R2 HyperV hosts, but this was the first server of the new Nehalem processor that I had, and while I had some intermittent blue screen problems on another similarly configured system, this system wouldn't run for more than a couple minutes, long enough for me to take a photo of the screen (which in the past went so quickly that I couldn't read the error to properly diagnose and fix the thing).

With the installation of the hotfix, ALL bluescreen problems/issues have completed disappeared.

************

Note: Before I was able to capture the bluescreen message, after the system had rebooted after the crashes, it provided an error log message that had the following, but of course this information was of little use...

Problem signature:

  Problem Event Name: BlueScreen

  OS Version: 6.1.7600.2.0.0.274.10

  Locale ID: 1033

Additional information about the problem:

  BCCode: 101

  BCP1: 0000000000000019

  BCP2: 0000000000000000

  BCP3: FFFFF88001F3F180

  BCP4: 0000000000000004

  OS Version: 6_1_7600

  Service Pack: 0_0

  Product: 274_3

Files that help describe the problem:

  C:\Windows\Minidump\052810-20358-01.dmp

  C:\Users\Administrator\AppData\Local\Temp\1\WER-131009-0.sysdata.xml

 Updated 4/27/2011:  As I was reviewing this post for a client of ours, I found another (similar) KB article that refers to a similar issue and a separate hotfix (http://support.microsoft.com/kb/981791)  This other hotfix specifically notes that this does NOT replace the previous hotfix, so it would seem there are cases where you might need both hotfixes.  I have found that the original hotfix referenced at the start of my blog post here has solved ALL previous issues of blue screens in HyperV with Nehalem processors, but guessing that if that doesn't solve your problem, take a look at this other hotfix (when comparing the file changes, this other hotfix here replaces 4 files on the HyperV server with newer updates than the other hotfix I reference, however the original hotfix that I reference actually replaces additional files that this new one doesn't.  So again, if you are having bluescreen probs, I'd start with reviewing the hotfix I reference at the start of this blog post, and then look at this other one if the first one doesn't solve the problem...

Join the Network World communities on Facebook and LinkedIn to comment on topics that are top of mind.
Related:
Now read: Getting grounded in IoT