-
Posted to
IIS7 - General
by
Mike426
on
08-11-2009, 3:00 PM
You generally won't get anything in the NT Event Viewer for a "hang" (other than notification that the w3wp.exe process isn't responding).
You might configure the application pool for multiple worker processes (and increase the recycle frequency for the worker processes) - it sounds like the hang is specific to the ...
-
Posted to
IIS7 - General
by
Mike426
on
08-10-2009, 10:15 PM
If it can hang twice in ten minutes in production, then there must be a way to reproduce it. I'm guessing it either has to do with the specific order in which users perform certain operations, or maybe concurrent access to something. When you stress tested it, did you do so with multiple simulated clients acting concurrently, or ...
-
Posted to
IIS7 - General
by
Mike426
on
08-07-2009, 10:10 AM
OK, great - if you control the server it should be easier to troubleshoot. If the same code works on one machine and not another then it's probably a configuration or permissions issue. Is there anything useful in the NT Event Viewer log? I would still go through the process of commenting out most of the code and then ...
-
Posted to
IIS7 - General
by
Mike426
on
08-06-2009, 10:06 AM
Can you go through the process of eliminating variables? For example, can you run your app on a local instance of IIS and see if you get the same behavior? If not, then you might need to talk to the hosting company. Alternatively, you can comment out most of the code and verify that you don't get the error on the hosting ...
-
Posted to
Troubleshooting
by
Mike426
on
08-05-2009, 10:08 PM
Maybe you just need to recycle your IIS worker processes more frequently. Can you reproduce this in a test environment, or does it only happen in production? Is there any sort of pattern to how much activity your web app gets before you get the crash? If you've got a memory leak but can't locate it in order to fix it, the ...
-
Posted to
Troubleshooting
by
Mike426
on
07-24-2009, 11:52 AM
It sounds like this happens consistently. Can you reproduce the behavior in a single-user scenario in a test environment? If so, then you ought to be able to step through the code while it's executing and see where the hanging starts. If you can't reproduce it this way, then can you add some application-level logging to ...
-
Posted to
IIS in SySWoW Mode
by
Mike426
on
07-21-2009, 3:36 PM
You could look into using "impersonation" for just the part of your code where you're working with the COM object. There are a few ways to do it, and a lot of good examples on the internet. The upside is that your application pool primarily runs under a less-privileged account. The downside is that you ...
-
Posted to
IIS in SySWoW Mode
by
Mike426
on
07-21-2009, 2:58 PM
What happens if you temporarily configure the IIS application pool to run under the same account you were using to run the console app? Is this an ISAPI extension or an ISAPI filter? Also, can you run a quick test of instantiating a COM object in your ISAPI dll that is written in unmanaged code (Msxml2.DOMDocument, etc.) instead ...
-
Posted to
IIS7 - Performance
by
Mike426
on
07-20-2009, 4:36 PM
Can you reproduce the issue in a test environment, or only in production? Likewise, does it happen in a single-user scenario, or only when there are multiple users? Ultimately, you would want to be able to make it happen at-will and step through your application code while it is happening. If you're not currently getting the ...
-
Posted to
IIS7 - General
by
Mike426
on
07-08-2009, 11:38 AM
It would be great to get to the point where you can reproduce the issue in a lab environment. If you could create a copy of your web app and set it to refresh itself every minute, etc. and leave it running in a test environment for a couple of days that might eventually reproduce the issue. Once you get to the point of reproducing it, ...