ASP.NET StateServer connection timeouts on IIS 6

[AV]

I have a web server in production running IIS 6 (Windows Server 2003 SP1) and hosting multiple virtual websites with each website hosting many ASP.NET applications in virtual directories.  Each website is running in it's own worker process and all the application use the out of process ASP.NET State Server running on the same webserver (eg. connecting to 127.0.0.1:42424).

 

Now recently the following errors have been showing up in the application event log on the server.  At first there would maybe be 20 in a row one day randomly spaced minutes apart and would only occur every couple weeks.  Recently these errors have started showing up everyday, 25+ spaced randomly minutes apart.  Another tidbit is that if I recall correctly I believe this started happening around the same time AJAX applications started being deployed to the server, I'm not sure if they're related.  I am unable to correlate if a specific application causes this error.

 

This is the error everytime:

>>>

Event Type: Error
Event Source: ASP.NET 1.1.4322.0
Event Category: None
Event ID: 1078
Date:  1/31/2006
Time:  10:41:38 AM
User:  N/A
Computer: WebServer

Description:
The state server has closed an expired TCP/IP connection. The IP address of the client is 127.0.0.1. The expired Read operation began at 01/31/2006 10:41:04.

<<<

 

This error does not seem to be causing any problems with session states being lost or dropped as there have been no reports from users of problems along those lines (potential users for the sites number in the 10's of thousands).

 

I have searched the net high and low and newsgroups and the only real piece of advice on this specific error comes from MS KB 308097 (http://support.microsoft.com/default.aspx?scid=kb;en-us;308097).  Seems to be describing the problem I'm having except that this webserver is not under high load or high CPU usage, I would describe the load as merely moderate and CPU usage as low.  I increased the timeout in the machine.config and the registry to the recommended 20 seconds anyway to see if that would help.  It appeared to actually help for a while, but steadily has gotten worse and worse.  I increased the timeout further to 30 seconds, and this change had no effect on it.

 

When this server was built the following recommendations were used as well for increasing scalability and performance as this server makes substantial use of web services on itself and other servers.

From KB 821268: http://support.microsoft.com/default.aspx?scid=kb;en-us;821268

•	Set the values of the maxWorkerThreads parameter and the maxIoThreads parameter to 100.
•	Set the value of the maxconnection parameter to 12*N (where N is the number of CPUs that you have).
•	Set the values of the minFreeThreads parameter to 88*N and the minLocalRequestFreeThreads parameter to76*N.
•	Set the value of minWorkerThreads to 50. Remember, minWorkerThreads is not in the configuration file by default. You must add it.

Can anyone add some insight into this issue?  I was hoping with the long winded question that people would be able to find the answer to this specific problem easier in the future once it is solved.

 

Thanks in advance for any help!

 

AV