Take the 2-minute tour ×
Server Fault is a question and answer site for professional system and network administrators. It's 100% free, no registration required.

I have a server that has several application pools (basically one per CPU core). Each application pool: 1) has several sites, 2) runs .NET Framework v4.0.30319, and 3) is running in Managed pipeline mode: 'Integrated'.

All the sites in each of the application pools point to the exact same codebase (MVC.NET 3 site) and have 2 virtual directories (one vd is the same for all sites and one vd is unique to each site).

All sites are all running fine except for one. It hangs consistently. The only major symptom of the problem is its Worker Process' Requests queue seems to grow out of control. Requests to all the other sites (and all other sites on the same App Pool) are handled just fine. The long running requests are only for the hanging site. I accessed the request queue as follows:

[Server Node] > 'Worker Processes' > [double-click worker process that hanging site lives on]

...when I recycle the worker process, these requests vanish and the site runs fine for a bit. After a little time (30min - 1hr), the requests have begun hanging again (time elapsed just continues increasing indefinitely).

What could be causing this issue? I have tried:

  • Switching the site from one app pool to another
  • Switching the site from one web server to another (on a different IP)
  • Completely rebuilding the site from scratch

Any ideas?

Server Config

  • OS: Windows Server 2008 R2 Standard SP1
  • IIS: 7.5.7600.16385
  • CPUs: 2 x Intel Xeon E5430 2.66GHz
  • RAM: 32GB
share|improve this question
    
Does the troubled web site perform any external http(s) calls? –  Mathias R. Jessen Jan 28 at 19:32
    
@MathiasR.Jessen - all the sites rely on a common image server, yui.yahooapis.com, googleapis.com, and maps.gstatic.com. –  fordareh Jan 28 at 22:39
    
Is it using some ODBC connection perhaps? Have you tried adding this to it's own app pool w/o sharing? –  Schrute Feb 8 at 21:36
    
@Schrute - we recently did move it to its own app pool. the problem occurs less frequently, but we did finally have to setup the app pool to recycle every hour. –  fordareh Feb 8 at 23:20
    
I haven't had much time managing windows servers mostly because I don't care about them anyways and do not want to meddle billy's os. But looks like slowloris DoS attack. Set up req timeouts. –  Marcel Feb 10 at 18:52

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Browse other questions tagged or ask your own question.