Troubleshooting › Diagnose Java Task Problems

Diagnose Java Task Problems

This section describes problems that you may encounter with the Java task.

Does Not Start, JCL Error

Symptoms

Started task does not become active.

SYSLOG or JOBLOG contain JCL error.

Possible Causes

Incomplete ReportCenter implementation.

REPORTCENTER Customizer parameter group has not been actioned, or failed when it was last actioned.

HFS problems including filesystem not defined, full, not mounted, or insufficient file permissions.

Action

• Check the REPORTCENTER Customizer Ilog. This Customizer parameter group generates the JCL, batch, and parameter files.

  Ensure that this group is actioned before you start the Java Task. For more information about actioning this parameter group, see Starting the ReportCenter Control Region in Chapter 10.

• Ensure that you copied the generated JCL to your MVS PROCLIB data set.
• Check REPTEST Filesystem-related output (messages WRSTIN*, WRSTCO*, WRSTHF*) for HFS details.

Does Start, No JCL Error

Symptoms

Started task does not become active.

No JCL error message on the SYSLOG or JOBLOG.

No Java Task log is written.

Possible Causes

Security problems such as started task not defined to RACF/ACF2, OMVS segment setup incorrect, started task not allowed to execute BPXBATCH.

Java product implementation errors, USS environment errors.

Action

• Check the SYSLOG at the time the Java Task started and stopped.
• Check the Java Task JOBLOG.
• Check REPTEST task-related messages (messages WRSTJF*). This produces errors about the inactive task, but should still display environment variable values, STDERR, and STDOUT text.

  If it does not, browse the STDERR and STDOUT files manually.

  Specific Causes

If the STDERR file contains an error similar to the following:

…/adaptor/Adaptor.bat 1: FSUM7351 not found

the z/OS Java product is not installed or is installed but cannot be found.

From OMVS, navigate to the Java /bin directory and enter the following command to verify whether a z/OS Java product is installed:

java -version

You should see output similar to the following:

java version "1.4.2"  
Java(TM) 2 Runtime Environment, Standard Edition (build 1.4.2) 
Classic VM (build 1.4.2, J2RE 1.4.2 IBM z/OS Persistent Reusable VM build cm142ifx-20070813 (SR9a) (JIT enabled: jitc))  

If Java is correctly installed, review the Java path you entered in the REPORTCENTER Customizer parameter group.

Starts, but Comes Down After a Short Time

Symptoms

Java Task becomes active, a Java Task log is written, but the task always fails shortly afterwards.

Possible Causes

The Timeframe and Prompt Generator services run immediately at task startup, and problems discovered at the first database access most frequently cause this situation.

DB2 subsystem not active

Incorrect DB2 security setup.

JDBC implementation incorrect.

Note: This condition can happen to a working system if JDBC maintenance is applied, and the corequisite JDBC profile regenerations and rebinds are forgotten.

RRS security setup incorrect.

DB2 sections of the /REPORTCENTER parameter group incorrectly completed.

CA Datacom/AD

Datacom MUF address space not active.

Datacom Server (JDBC) address space not active.

Datacom product installation incorrect.

Datacom Common Services prerequisites incorrect.

ReportCenter database setup incorrect.

Datacom sections of the /REPORTCENTER parameter group incorrectly completed.

Java or USS Environment problems.

Incorrectly generated CLASSPATH.

Action

• Review the Java Task log. Usually this situation produces error messages such as SQL error codes indicating security or database problems.
• For more information about common DB2 and Datacom errors, see the DB2 and CA Datacom/AD sections.
• Review the SYSLOG at the time task started and terminated.
• Review the Java Task JOBLOG.
• Ensure that the JDBC DBRMs have been rebound after DB2 maintenance.
• Review the REPTEST task-related messages (messages WRSTJF*).

  These messages show environment variable values, and STDERR and STDOUT text.

Specific Causes

If the following message appears in the ReportCenter Java Task log:

Cannot establish database connection. Message: DB2SQLJConnection error in native method: constructor:  RRS "IDENTIFY" failed using DB2 system:D…

Proceed as follows, according to the REASON code:

• If the REASON code is 00f30012, verify that the DB2 subsystem has started.
• If the REASON code is 00f30049, verify that the statements in the supplied CC2DSAMP(WRBD2GRA) member have run.
• If the REASON code is 00f30091, verify that RRS/MVS has started.

  Note: If RRS is recycled, you must recycle DB2; therefore, you must recycle the ReportCenter Java Task when your DB2 subsystem is recycled.

Was Running, but Suddenly Terminates

Symptoms

Started task is active and processing, but unexpectedly terminates.

Possible Causes

Deliberate

If the Java Task detects an external condition that is difficult to recover from, it terminates. Such conditions include problems with database access, security, database resource shortages, and so on.

If these problems are present when the Java Task starts, it terminates immediately and this situation is indistinguishable from the one above. However, the task can also process correctly for a long time, before an unrecoverable condition appears. The database or DBMS being shut down is a frequent cause of such a condition.

Accidental

The task abends on the USS side - running out of Java storage, uncaught Java exceptions, JVM heap damage, and so on.

The task abends on the MVS side - 0CX abends in DB2 or LE modules.

Action

• Check the Java Task log. If the Java Task deliberately terminates, it always writes a terminating … message to its log, with the contributing error details.

  For more information about common DB2 and Datacom errors, see the DB2 and Datacom sections.

• If there are no errors on the log and it stops in the middle of a task, this indicates an involuntary abend. Check the SDSF JOBLOG for the Java Task and the SYSLOG at the time of the STC termination for an MVS abend messages
• If there are no JOBLOG or SYSLOG errors, the REPTEST task-related messages (messages WRSTJF*) should show STDERR and STDOUT text. If there has been an abend or dump on the USS side, there is usually a message in either of these.