We run numerous scheduled jobs using MS SQL 2005 job scheduler.
Over the past few weeks, we've been witnessing that a running job will simply STOP and will not record any history in the Log File Viewer.
What appears to be happening is that the server will start a job and record these steps into msdb. At a random step during a job, the server will stop running the job and it records no error. This silent failure prevents a "Step 0" record from being created and we won't see anything in the Log File Viewer.
Anyone else experience this?
Try this method to get some more details on what the job is doing:
In SQL Server Management Studio:
Never experienced this behavior.
Is this any running job under SQL Agent or a particular one? Anything remarkable in the SQL log or application or system event logs? You can see the recorded steps in msdb? What about starting Profiler, starting your job and examining the Profiler output?
I always always always (get the idea :) ) ...always direct the output of each step of every job to a text file somewhere on the server. This is ideal for these kind of scenarios.
step 1 I set for the output file to overwrite and every subsequent step I step to append to that same file.
...and I ensure this is done on each and every job to a standard output location. There is stuff that is outputted that sometimes doesnt usually make it into job history but will into a text file.