OSHMON_INIT_RETRY

When a parallel job is starting up the DSD.OshMonitor process is launched which connects to the JobMonApp process to request metadata and to monitor progress. JobMonApp only responds successfully to these requests when the parallel job lead osh process for the job has started and connected to JobMonApp. By default the DSD.OshMonitor waits for 300 seconds for a successful response.

In a GRID environment the startup of the lead osh process might be delayed waiting for resources to become available. If the job startup is delayed until after the DSD.OshMonitor startup retry period has expired, metadata and progress data are not available for that job run.

You can set the DSD.OshMonitor startup retry period to be longer, by setting the environment variable OSHMON_INIT_RETRY to an integer value, in seconds, that is greater than 300.