tags:

views:

3131

answers:

5

Hey guys,

I've recently wiped and reinstalled/configured all the components of my web and DB servers. I'm running IIS 6, .NET 3.5, SQL Server 2005. The two servers are separate VM's in the same domain. My web app functions perfectly... 90% of the time. But every now and then I get this error or a generic runtime error:

System.Data.SqlClient.SqlException: A transport-level error has occurred when receiving results from the server. (provider: TCP Provider, error: 0 - The semaphore timeout period has expired.)

Can you guys give me some leads on where to start troubleshooting? I really think I've missed something in IIS.

Thanks in advance!

A: 

Can you double check the load on the db server? We do get them once in a while in our dev env, but never in the prod env.

renegadeMind
The load should be fine, this error was produced while only one person was testing it.
erick.brower
Load here doesn't necessarily mean the number of people using the application, but the number of processes running on the db server and the amount od cpu/network the db instance is being allocated.
renegadeMind
A: 

Where to start troubleshooting? Maybe here ;)

It could be that there are too many open connections -- are you sure that the connections are being closed properly ?

You are getting a SQL exception, so the first place I'd look is the database...

markt
+1  A: 

Well I would start by ensuring I'm running the latest version of my virtual machine's network drivers.

Sounds like comms is failing between your 2 VMS. Try opening a TCP socket between the two machines and see if it dies after a certain amount of time.

Sam Saffron
A: 

A few pointers are here .Basically the error is being thrown by the network layer and SQL server is just reporting it.

Hope that helps.

Learning
+2  A: 

After troubleshooting this for hours and sitting on the phone with my hosting group they discovered that there was a problem with their networking configuration. The solution was made clearer when during my testing one of the VM's suddenly couldn't find the domain, and a simple 'ping' to the IP of each box from the other would occasionally time out. This ruled out DNS entirely. After the hosting group applied the proper configuration on their end the app has been stable and FAST!

Thanks for everyone's help!

erick.brower