Problem:

• Product Name: Borland Enterprise Server, AppServer Edition
• Product Version: 6.0.1 Patch bes601_p13, bes601_p20, bes601_p25
• Product Component: VisiConnect
• Platform/OS version: Solaris 8

The following exceptions are observed in the partition log.

1. 2009-07-21 01:00:39,777 ERROR - Warning: the Connection Pool seems to have a problem: null

   java.lang.Error: Warning: the Connection Pool seems to have a problem: null
   at com.borland.enterprise.visiconnect.Global.log(Global.java:580)
   at com.borland.enterprise.visiconnect.resource.ConnectionPool$PoolUpdater.doWork(ConnectionPool.java:2684)
   at com.inprise.j2ee.scheduler.WorkerThread.run(WorkerThread.java:66)
   2009-07-21 01:00:39,777 ERROR - STACK TRACE
   java.lang.NullPointerException
   at com.borland.enterprise.visiconnect.resource.ConnectionPool.maintainUseAverage(ConnectionPool.java:1365)
   at com.borland.enterprise.visiconnect.resource.ConnectionPool$PoolUpdater.doWork(ConnectionPool.java:2659)
   at com.inprise.j2ee.scheduler.WorkerThread.run(WorkerThread.java:66)2009-07-17 15:58:26,004 ERROR - Warning: the Connection Pool seems to have a problem: / by zero

2. java.lang.Error: Warning: the Connection Pool seems to have a problem: / by zero

   at com.borland.enterprise.visiconnect.Global.log(Global.java:606)
   at com.borland.enterprise.visiconnect.resource.ConnectionPool$PoolUpdater.doWork(ConnectionPool.java:2809)
   at com.inprise.j2ee.scheduler.WorkerThread.run(WorkerThread.java:66)
   2009-07-17 15:58:26,004 ERROR - Warning: the Connection Pool seems to have a problem: / by zero
   java.lang.Error: Warning: the Connection Pool seems to have a problem: / by zero
   at com.borland.enterprise.visiconnect.Global.log(Global.java:606)
   at com.borland.enterprise.visiconnect.resource.ConnectionPool$PoolUpdater.doWork(ConnectionPool.java:2809)
   at com.inprise.j2ee.scheduler.WorkerThread.run(WorkerThread.java:66)

3. 2009-07-21 01:07:50,135 ERROR - STACK TRACE

   java.lang.NullPointerException
   at com.borland.enterprise.visiconnect.resource.ConnectionPool.maintainUseAverage(ConnectionPool.java:1365)
   at com.borland.enterprise.visiconnect.resource.ConnectionPool$PoolUpdater.doWork(ConnectionPool.java:2659)
   at com.inprise.j2ee.scheduler.WorkerThread.run(WorkerThread.java:66)

What is the root cause for the above logs? Is there any impact to the management of managed connections?

Resolution:

The exceptions occur because the following XML parameters are not defined in the ra-borland.xml. The workaround is to ensure these 2 elements are defined within the element in the ra-borland.xml.

true

600

These 2 elements are designed as optional. A defect CR 10069 has been raised. Currently, BES 6.0.1 is already de-supported. There is no definite timeline when the bug will be fixed.

The cleanup-delta determines the frequency with which the AppServer calculates the need for connection pool size reduction. It is used for controlling the resource adapter resource usage and maintaining the size of the connection pool to an efficient level.

Eg. The max capacity is defined as 10 managed connections. When the application peak load occurs, all the 10 managed connections are utilized. Subsequently, the load is low and the application utilize only 3 managed connections. The cleanup-delta will reduce the connection pool from 10 to 3 over a period of time.

When the above exceptions occur, the AppServer will continue to maintain 10 managed connections in the pool instead of shrinking to its optimum size. It will not impact the application in getting a managed connection from the AppServer

References

1. Borland Resource Adapter 1.0 DTD

2. Controlling System Resource Usage, Page 303 of the Borland Enterprise Server Developer guide