IIS 5 & IIS 6
IIS Thread Pool restarts unexpected/Sporadic.
Last post Sep 12, 2006 08:55 AM by tomkmvp
Sep 11, 2006 02:34 AM|Alsvha|LINK
I hope this is the correct part of the forum, it was the one which made most sense. Anyways...
I'm investigating why a website seemingly restarts ever so often. The website runs on .NET 1.1 (although .NET 2 is also installed) and on IIS 6.0.
I found this support article from Microsoft: http://support.microsoft.com/kb/918041/en-us?spid=8940&sid=global which states that recycling of worker threads in IIS 6.0 and ASP.NET 2 can cause the app.pool to restart.
Now, this I find suspecious, because the website in question seemingly makes (extensive) use of Response.Redirect and Response.End, which throws the System.Threading.ThreadAbortException (which are caught in the code)
Now, I'm not terrible server-minded, but to me this could indicate that the threadpool of .NET gets recycled often and which in turn could trigger the aforementioned bug from the support article.
My "only" trouble with this is that it is - as mentioned - ASP.NET 1.1 application, or so it says in properties in the IIS, but as said - running IIS 6.0 and .NET 2 framework is installed on the server.
Does anybody have any experience with if this bug can also affect ASP.NET 1.1 application run on IIS 6.0??
Also - as far as I can tell, at least step one of the article have been performed on the server by the hosting company, however it could be that the fix wouldn't work for ASP.NET 1.1 applications and something else needs to be performed?
I hope somebody can help, because I'm running out of options.
Thanks in advance.
Sep 11, 2006 01:47 PM|ApurvaJoshi|LINK
The article that you are refering to is specific to .NET 2.0. The article is refering to worker process and not worker threads. I dont think this article has anything to do with your problem.
Can you please provide more information on what exactly are you trying to troubleshoot?
Sep 12, 2006 02:50 AM|Alsvha|LINK
Thanks for your feedback (it is about the only one I've been able to get all over the web :)).
My problem is that I'm not a server-guy, but a programmer, so I rarely worry about stuff such as the IIS and similar, because that is what our hosting companies are for, but they seem to come up with a loss on this one, and then I've been assigned to try
and look into it.
But before I start walking through the code for possibilities, I was directed to the linked article and found it fitted nicely into the equation.
We have a website which runs in its own application pool (because it chrashes the pool/the pool restarts) on a server. This websites pool - as mentioned - restarts sporadic and sometimes quite often, thus loosing all session and similar. The hosting company
automatically sends us an e-mail each time the pool goes up and down. Just yesteday the pool went down 7 times (which I don't know if is much), but it has at times done it 20 times a day (which does sounds as much). The website has been moved from one server
to another server, just to see if it was some other site/software which played into the equation. But this is the only site which goes down that often of all the websites we had on the two servers.
The problem is that the clients doesn't see exceptions occuring on the page, and the exceptions which are caught are 99% of the time System.Threading.ThreadAborted exceptions (due to response.redirect/response.end), hence why it would fit nicely into the
But because I'm not a server-guy, my server-knowlegde is limited, and I have little idea what to look for - also when the hosting company (which should be the server-capable people :D) seems to be at a loss, and because quite frankly - I can't see this to
be neither a websitecode or database connective issue (which have been checked, and seems to behave as expected).
So I'm at a loss here, feeling like a lamb being brought to the slautherhouse - luckily it isn't my website, I was just assigned to help. :)
But thanks for the feedback.
Sep 12, 2006 08:55 AM|tomkmvp|LINK