2

Edit: Sep. 26 2012 - I am updating this question and its tags because I have now encountered this error again, on a Server 2k8R2 VM. Still, there seems to be no authoritative information out there on the net for this, so in case I ever solve it, this page could become a valuable resource for others having the same issue.

I've Googled the hell out of this and can't find an answer.

I have a Windows Server 2003 VM. I run the following command on the server:

typeperf "\Processor(*)\% Processor Time"

And here is the output:

"(PDH-CSV 4.0)","\\SERVER01\\Processor(0)\% Processor Time","\\SERVER01\Processor(_Total)\% Processor Time"
"03/30/2012 11:40:13.881","0.000000","0.000000"
"03/30/2012 11:40:14.795","22.374429","22.374429"
"03/30/2012 11:40:15.807","-1","-1"
"03/30/2012 11:40:16.707","0.000000","0.000000"
"03/30/2012 11:40:17.607","18.402778","18.402778"
"03/30/2012 11:40:18.507","0.000000","0.000000"
"03/30/2012 11:40:19.407","-1","-1"
"03/30/2012 11:40:20.307","-1","-1"
"03/30/2012 11:40:21.207","-1","-1"
"03/30/2012 11:40:22.107","-1","-1"
"03/30/2012 11:40:23.007","-1","-1"
"03/30/2012 11:40:23.907","0.000000","0.000000"
"03/30/2012 11:40:24.807","-1","-1"
"03/30/2012 11:40:25.707","-1","-1"
"03/30/2012 11:40:26.607","-1","-1"
"03/30/2012 11:40:27.507","-1","-1"
"03/30/2012 11:40:28.407","0.000000","0.000000"
"03/30/2012 11:40:29.307","-1","-1"
"03/30/2012 11:40:30.207","-1","-1"
"03/30/2012 11:40:31.107","-1","-1"
"03/30/2012 11:40:32.007","-1","-1"
"03/30/2012 11:40:32.907","0.000000","0.000000"

Error: A counter with a negative denominator value was detected.

Alright, so why the intermittent negative ones? I have a thousand other systems on which this behavior does not occur, but I can't figure out why it's happening here. Any help would be appreciated.

2

After six months of (off and on) searching, I have finally found the solution to this problem.

I had seen that KB mentioned by brain99 before as well. It vaguely alludes to a "hardware timing issue," but doesn't give any specifics. Since I had really only seen this on VMs, the idea of a "hardware timing issue" seemed strange, but it did make me think about how VMs communicate with their host.

These were VMware virtual machines, and in the VM Tools, there is a "Time synchronization between the virtual machine and the ESX server" setting that was checked. Unchecking that solves the problem.

All this time I never fiddled with the VMware tools and the solution was so simple. :P

1
+50

According to this Microsoft KB article, the error has something to do with timing issues but does not necessarily indicate a problem. It doesn't explicitly mention the negative values, but it stands to reason that the two are connected since the article specifically mentions a problem reading the Processor Utilization counter.

Unfortunately, no specific solution is provided in the article.

  • I figured that's about as good of an answer as I was going to get. It's not a big deal - I can rebuild the counters database or even rebuild the entire operating system... I mainly just wanted to know for academic purposes how exactly it happened. – Ryan Ries Aug 30 '12 at 15:10
  • I'm sorry I had to un-check you... I found the definitive answer. :) – Ryan Ries Oct 2 '12 at 19:38

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.