SQL server version: MS 2005 9.00.3050.00 Std edition
Development version is on a stand alone server. Production is on an active/active cluster.
Full error message below. There is an explanation and solution as noted below. However we do not see this issue in all our environments. Can anyone tell me why we see this issue in our production environment but not on our development environment, especially
as the issue seems to arise following low usage times.
Explanation / Solution
If you are using the v1.2 driver, and you can positively rule out transient network connectivity issues and multithreaded usage, the next most common cause I am aware of is where a connection pool is being used and the pooling manager (e.g. DBCP or C3PO)
does not test pooled connections for validity before handing them out to the app. What happens in this case is connections that sit idle in the pool for a very long time get closed by the TCP/IP stack, routing hardware, etc. Then they're already dead when
the pooling manager hands them out. The solution is to configure the pooling manager to validate connections before handing them out, typically using a validation query like "SELECT 1".
INFO | jvm 1 | 2010/09/07 07:52:11 | [09.07 07:52:11.396] LOG ERROR BROKO-EY-TBF-REDUCED-ENV-DP>
INFO | jvm 1 | 2010/09/07 07:52:11 | - broko.quark.QueueMonitorThread.getNextCommand(QueueM
View Complete Post