Sun Java System Application Server Enterprise Edition 8.1 2005Q2 Update 2 Release Notes

Connectors

This section describes known J2EE connector architecture issues and associated solutions.

Bug ID 

Summary 

6188343 

After restarting a DAS instance, undeploying the connector module fails when cascade is set to false. 

In this scenario, a standalone or embedded connector module is deployed in DAS and connector connection pools, and resources are created for the deployed module. After restarting the DAS instance, undeploying the connector module fails when cascade is set to false with the following exception: 

[#|2004-10-31T19:52:23.049-0800|INFO|sun-appserver-ee8.1|javax.enterprise.system .core|_ThreadID=14;|CORE5023: Error while unloading application [foo]|#]

Solution

Use cascaded undeploy (set the cascade option to true) for undeploying standalone and embedded connectors after restart of the DAS instance. 

6412132 

Using JDBC getConnection() for pooling connections.

If you get a Connection from an Application Server JDBC connection pool, create a Statement object, and then use the Statement.getConnection method, the statement returns the physical connection instead of the wrapped connection. When you close this physical connection, you break the connection pool logic. To avoid this problem, set the following JVM property in the domain.xml file:


<jvm-options>-Dcom.sun.appserv.jdbc.wrapJdbcObjects=truejvm-options>-Dcom.sun.appserv.jdbc.wrapJdbcObjects=true>

6412148 

ORBCommunicationsRetryTimeout error.

The ORBCommunicationsRetryTimeout property specifies the number of seconds the ORB client will try to establish a connection to an unreachable ORB backend. The default value is 60 seconds. With this default setting, you may see a large number of CORBA exceptions in the logs, as well as high network usage if the ORB backend is not reachable.

Solution

Set the ORBCommunicationsRetryTimeout to a lower value.