Strange failure in IIS6RSS

6 replies

Last post Jul 13, 2006 06:13 PM by ericdeily

  • Strange failure in IIS6

    Jun 15, 2006 08:53 AM|Wallym|LINK

    Has anyone had a situation where an anti-virus program has caused IIS to
    stop working?  We are using Norton/Symantec Anti-Virus.  The latest theory in our
    ongoing problem is that the anti-virus is locking a file/directory and this
    then blocks IIS from using a necessary file and then IIS fails.  Has anyone
    seen anything like this?  We are looking at excluding IIS from scanning a
    couple of directories, but before going through this, we are looking for
    other people's experience with this type of issue.  I am not 100% convinced that the anti-virus is actually causing the problem, I am just working with the info that the MS Support Engineer gave me at this time.  Any thoughts are appreciated?

    Other pieces of information:

    • We are running on IIS6/Windows 2003.  We have opened up a support call with MS Support.  There is a bug in COM+ in W2k3 SP1 that is causing us a problem getting the IIS intrinsics.  That specific problem has been solved by chaning the COM+ objects to operate as a library application as oppossed to a server application.
    • This is not an ASP.NET application.  It is a classic asp app that does 99.9% of its processing within a set of COM+ objects.  The ASP pages load the COM+ objects.  The COM+ objects do almost all of the processing.
    • We are not using Exchange on the same server, and I didn't write this application.

     Wally

  • Re: Strange failure in IIS6

    Jun 16, 2006 11:39 AM|mvolo|LINK

    Can you describe the problem your are having in more detail - "stops working" is a very general condition :)
    Mike Volodarsky
    CTO at LeanSentry
    Former IIS/ASP.NET PM

  • Re: Strange failure in IIS6

    Jun 17, 2006 02:54 AM|qbernard|LINK

    Typically, in my previous asp setup, we exclude the .asa, .inc, etc from the virus scan, as this will trigger the app to recycled. Other than that, nothing much. Oh ya. script blocking is the other setting that may cause potential problem as well.

    Cheers,
    Bernard Cheah
  • Re: Strange failure in IIS6

    Jun 19, 2006 09:35 AM|Wallym|LINK

    Literally, the web server stops responding to requests.  Sometimes we get a message in the event log, sometimes not.  The web client will eventually timeout waiting on its requests.  Unfortunately, there is not a lot more information that I can provide.  The server admins will not allow us access to the web server's console.
  • Re: Strange failure in IIS6

    Jun 19, 2006 10:57 AM|chrisad|LINK

    Hey Wally,

    What is the event id and\or event desc when it does happened?  Is this Service Pack 1 by chance?  I know you gave details in the first thread but the reason I ask I was wondering if you have thought about using tracing?  Although, it might be hard as it doesn't sound like you have a lot to go on -

    Also, can you set LogEventOnRecycle for the application pool in question -

    cscript adsutil.vbs set w3svc/LogEventOnRecycle 255

    This will ensure that if the apppool for some reason recycles (or fails) then a log event will be recorded in the system event viewer...

    Thanks,

    ~Chris
    Program Manager
    IIS Product Unit (Microsoft)
  • Re: Strange failure in IIS6

    Jun 19, 2006 10:56 PM|Wallym|LINK

    Here is the the event description that we have in the windows logs:

    A process serving application pool 'DefaultAppPool' suffered a fatal communication error with the World Wide Web Publishing Service. The process id was '352'. The data field contains the error number.

    Is there anyway to reset the app pool when this happens?  Unfortunately, I don't know enough about the IIS6 reliability features to know what is and is not possible.

    The next question is "what is the data field?" as stated in the error message?  What does that term refer to?

  • Re: Strange failure in IIS6

    Jul 13, 2006 06:13 PM|ericdeily|LINK

    Wally - if this happens frequently (i.e. 5 times in 5 minutes), you'd see another event saying that DefaultAppPool has suffered a flurry of failures and is being put into Rapid Fail Protection.  This means you'd be getting 503's back for requests to that pool and they'd actually be logged in the HTTPERR log file. 

    Now, as to why the WP is failing...I'd recommend you check out DebugDiag and create a crash rule.  This event you indicate means the worker process crashed, and DebugDiag can help you figure out why.  Check the numerious postings here on DebugDiag -> it's a great tool. 

    -ericdeily
    Program Manager - IIS
    Release/Proj Mgmt & Support/Health/Instr/Tracing/Logging