Troubleshooting: Jobs do not appear in the job queue

Issue

Under extreme circumstances (for example, when experiencing an unusually high load or network problems) signals between processes can be lost and jobs may not appear in the job queue.

Details

To ensure that all signals are accounted for, the SCU queries the database every 240 seconds to check for missed jobs (lost signals). If the query determines that jobs were missed, it refreshes the job queue list.

If the refresh takes longer than 600 seconds to complete, to ensure that jobs are completed in a timely manner, the SCU temporarily stops querying for missing jobs and logs a warning to the log file. The SCU then continues processing jobs already in the queue. After an additional 600 seconds, the SCU attempts to query the database again for missed jobs.

Solution

The QUEUE_POLL_INTERVAL value in the map_ini table determines how frequently the SCU queries the database for missed jobs. The default value is 240 seconds. The QUEUE_MAX_BUDGET_DISPATCH_TIME value in the map_ini table determines how long the SCU waits for the job queue to be refreshed and, if the refresh is unsuccessful, how long the SCU waits before it queries the database again for missed jobs. The default value is 600 seconds.