IIS 7.5 crashes after a few requests (with Django + PyISAPIe)

I managed to run Django using IIS as webserver (using PyISAPIe) and everything goes well in my test server, mounting Windows 2008 Server R2 64bit.

Then I installed the application on another server with the same configuration and it works fine for the first request. Then when I reload the page, I get a "Service not working" page.

On the event log I see an Application error saying that python26.dll had some problems:

Faulting application name: w3wp.exe
Faulting module name: python26.dll
Exception code: 0x40000015
Faulting application path: C:\Windows\SysWOW64\inetsrv\w3wp.exe
Faulting module path开发者_开发技巧: C:\Windows\system32\python26.dll

Can you give me some hint on how to solve that problem?

UPDATE: "Rapid-Fail Protection" in the Advanced Settings of the Application Pool was set to 5 failures; disabling it, all worked well.

So, now the question is: how can I detect what caused the failures?

UPDATE: I discovered that IIS crashes when there are multiple requests (img, css, js). PyISAPIe is called for each of them, passing them to static server once recognized. No idea why this happens...

PyISAPIe is not a good choice to run Django on Windows 2008. In this article you can find better solution: Running Django on Windows (with performance tests)

Check the eventlog it should be in there.

You may also find some more detail in the httperror log (C:\Windows\System32\LogFiles\HTTPERR).

I discovered that IIS crashes when there are multiple requests (img, css, js). PyISAPIe is called for each of them, passing them to static server once recognized. No idea why this happens...

Do multiple request cause the error on both machines? When there are multiple requests in an ISAPI application, each request runs in its own thread. The Python multi-threading model is GLOBAL - all threads running under that Python process are co-mingled and sharing all global resources, so you must serialize all multi-threaded code running in all applications and processes using your Python engine. This is a serious downside in Python multi-threaded processing and may be the source of your problems. See http://docs.python.org/library/multiprocessing.html and other sources.

But even this only occurs on one machine and not the other, that may still be the cause - it could also depend on many other environmental variables - number of requests, resources of the machine, processors, etc.

Check memory usage on machine (total physical).