24x7 Windows Service Became Zombie

Using 24x7 Schedule (v. 3.4.26) as a Windows Service. A job started, but did not complete execution this morning. It launches a .NET console application, which has its own logging mechanism. The 24x7 logs (on the host) indicated that the job started, but there was no indication that the job ended. The application logs had no entries indicating that it executed.

I tried re-running the job remotely using 24x7 Remote Control to connect to the Windows Service instance. Then I received a Remote Job Execution error. Exit Code: -1.

At this point, I suspected that 24x7 Windows Service was in a zombie state--jobs were not executing automatically.

Restarting the Windows Service resolved the issue, but I'm interested to know if there are more effective ways to detect when the 24x7 Scheduler Windows Service is not properly running. Or, if there are any best practices to keep the Service instance in a stable state.

Thank you.

You may want to create a heartbeat job, to run every minute or so, and report some activity.

In response to jobs not running, in case if all jobs are assigned to the same job queue, and an synchronous job runs away, all following jobs will sit and wait in the queue until the run-away job is done (ended somehow). In this state it may look like nothing is running, while indeed it is just a congestion effect. Jobs for different projects/applications should be assigned to different queues to potentially avoid getting into this situation.

If a job leaks some resources and doesn't terminate cleanly after an error or success, it may cause to hung the process forever. The scheduler will wait forever for the job to return control back. It is a good idea to set job timeout parameter so that the scheduler is allowed to attempt to terminate a run-away job on timeout.