This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

Category does not exist. (Yes it does)

Running into an odd issue deploying some standard templates out to some new R2 64 bit boxes.

IIS and Windows components are reporting "Category does not exist." 

So I logged in as my service account to one of the servers in question, opened perfmon and sure enough it does exist. Double check my work by doing a runas on perfmon from my APM box using the same account and connect to the server in question. Sure enough, Web Service and System Categories don't exist...

Few articles and blogs out there saying its a 32 bit counter issue vs 64 bit but I don't think thats the case as these templates are working fine on other identical servers cloned at the same time.

Anyone run into this yet?

Parents
  • Are you seeing this error exclusively on Windows Performance Counters or does this apply to services as well? Do none of the component monitors in the template work against this node or only some?

Reply
  • Are you seeing this error exclusively on Windows Performance Counters or does this apply to services as well? Do none of the component monitors in the template work against this node or only some?

Children
  • Just on perfcounters at this point.

    Its clearly a windows issue of some sort as I can replicate it in perfmon which explains why WMI isnt seeing them.

     

    Just not sure why I have it on a subset of machines all created at the same time.

  • You can try and rebuild the performance counter by running "lodctr /R" on the machine with issues. I'd also verify it's not a permissions issue by trying again using the local administrators credentials and see if that works.

  • Thanks thats on my list of things to try.

    Interestingly it seems APM is handling this poorly, in that its causing APM to spend a lot of resources before it fails the component. I have a case open for failing map loads #267870 and I believe this component failure is causing that. If I unmanage the nodes that issue goes away, remanage them and within a few polling cycles it returns.

    I had similar slowdowns in the past with too many applications going into fallback mode and causing severe degredation of APM performance #253990. Dev helped me turn off the fallback modes and performance went through the roof.

    This smells very similar to that one in that APM seems to spend too many cycles trying a dead component at the expense of every other component in need of a check.

  • Here is the fix to this:

    Winmgmt /resyncperf 
    Net stop winmgmt /y 
    Net start winmgmt