We want to run a long job to re-index large amount of pages using external search engine. The job always ends half-done with message FAILED Thread was being aborted. on the History tab.That's sad.
Is it caused by short httpRuntime/executionTimeoutsetting? Is it programmatically controllable?
Is right decision to register a job for this kind of 'manual' maintenance tasks or workflow task would have some advantages?
How long execution time are we talking about here?
There is no maximum execution time for a scheduled job, however as EPiServer scheduled jobs runs within the context of the webapplication, and if it is recycled, currently running jobs could/would eventually be aborted.
There are a number of reasons for the application to recycle, you may want to check that up.
There is also the difference between running the scheduled job manually or letting the scheduler start it.
If you click the "Start manually" button the job will run in the request to the server and you might get a timeout. But if you set a date and let the scheduler start the job, then it will run in a different thread and will not be subject of request timeouts.
Thank you, I guess you right. Because it is not a periodical task we always run it with 'Start Manually' button. Now I am going to schedule and see what happen. In theory it should take about couple of hours to complete. BTW I think one additional button 'Run once ASAP' will be good UI enchancemend.