WebLogic Server Threads are Reported as Stuck For Longer Than the Request Timeout When Accessing A Coherence Cluster Using Coherence*Extend (Doc ID 1465750.1)

Symptoms

WebLogic Server is running an application that accesses a Coherence cluster using Coherence*Extend.  The remote-cache configuration has its initiator configured with a 120 second request timeout, and the application is designed to failover to use a database if the requests to the cluster timeout.  In addition the proxy service had not been configured with a thread-pool and as a result was only capable of processing a single request from a single client at a time.  As can be seen in the error stack below the WebLogic Server instance is reporting the Extend related threads being stuck for longer than 120 seconds:

Cause

My Oracle Support provides customers with access to over a million knowledge articles and a vibrant support community of peers and Oracle experts.