Fixed Issues - 11.3.6 SP4

This article details the issues fixed in CA Workload Automation AE Release 11.3.6 SP4.
wcc114sp5p
This article details the issues fixed in CA Workload Automation AE Release 11.3.6 SP4.
UPDATE_MACHINE LOOKS EXEC SUPERUSER UNDER NATIVE SECURITY
Problem Number: 4149
When CA Workload Automation AE is running in native security mode, you could not update a machine using the update_machine subcommand if you are not an EXEC superuser. You can insert and delete a machine, but cannot update it.
JOBS STUCK IN STARTING/RUNNING AFTER FAIL-OVER/FAIL-BACK
Problem Number: 4390
Some jobs were stuck in the STARTING or RUNNING state after the scheduler fails over or fails back. This happened because the scheduler did not contact all active agents to send updated scheduler communication attributes. This happened for agents whose port is changed after the last successful communication.
SCHED HANGS WHEN PROCESSING MACH_ONLINE EVENT
Problem Number: 4393
When the scheduler processes the MACH_ONLINE event, it expects to evaluate all jobs in the PEND_MACH state. When this expectation is not met and the scheduler skips a job start evaluation, the scheduler did not evaluate the start of other PEND_MACH jobs. Also, during shutdown, the scheduler blocked indefinitely.
SCHED LEAVES EVENTS BEHIND WHEN SP FAILS DUE TO BIND ERRORS
Problem Number: 4415
The scheduler performs the following actions to execute a database query: parse the SQL, bind the input and output parameters to memory, execute the SQL, and fetch each row from the result set in a loop. Normally, when the execute operation fails due to a recoverable error, the scheduler retries the query. The exception is the ujo_batch_get_event stored procedure as it is executed at regular intervals. If, during one interval, the ujo_batch_get_event stored procedure fails to return a batch of events, those events are marked so that they remain behind out of the reach of the scheduler. For this reason, the scheduler does not retry the query so that it can recover and fetch the events as part of the next event polling iteration.
However, when the database operation to fetch events failed because of errors specific to the bind action, the scheduler retried the query and left a batch of events in the ujo_event table.
SCHED DOES NOT UPDATE JOB STATUS AFTER FAILOVER
Problem Number: 4417
When the primary scheduler detects that the shadow scheduler has taken over, it shuts itself down. Due to a timing issue, the primary scheduler may not send events for a job that has already completed as it shuts down. The shadow scheduler is not aware of the primary scheduler's final job activities and ignored the job. The job remained stuck in the RUNNING state until the primary scheduler is restarted.
COMM TABLE RECONCILE SCRIPT DOESNT WORK DUE TO SQL ERROR
Problem Number: 4419
Prior to upgrading the product binaries, it is recommended that you execute a SQL script that repairs table data that is used to store information about all agents. You may be required to execute the SQL script several times especially if your networked environment uses multiple levels of virtual node name associations. For example, you may be running in a networked environment where a virtual node name resolves to one of the many virtual IP addresses, each of which resolves to numerous physical IP addresses. In this case, the output of the SQL script may prompt you to execute one or more additional times to repair each level of virtual node name association that is detected in the table data. However, in some cases, the script may experience a database error and leave the table data unrepaired regardless of the number of executions.
COMM TABLE