One reason could be that IIS restarts your application due to e.g. reached a memory limit. You can configure when IIS should restart your site through IIS manager. Another option is that IIS restarts the apllication due to a unhandled exception in a background thread, if that is the case you should found an entry written in event log.
You should also ensure that the idle time-out on the application pool does not occur by setting it to 0
Thank you for your suggestions, i will try the above mentioned IIS configurations and see if that is the problem.
I’m having a problem running a custom scheduled job. The job is a content importing job, creating and updating pages through the Data Factory programmatically and creating and updating VPP files - all based on the content of an xml document and some files on the filesystem. The job has an average execution time of 3-5 minutes.
The main problem is that the job is unreliable. About half of the time the job runs to the end, but the rest of the time the job is being aborted with the message in the scheduled job log “Thread was being aborted”. Using log4net I am able to follow the execution of the job, and the job simply dies at random positions without exceptions. If the job is executed a couple of times with the exact same input it sometimes reaches the end as it should. The issue has been experienced running the job manually (“Start Manually”) as well as scheduled.
We have also experienced issues when running the EasySearch Scheduled index. At times we experience the “Thread was being aborted” exception – at other times the job execution is simply omitted and it skips a day in an otherwise daily scheduling.
Any suggestions how to troubleshoot this issue?
Is it possible to get more debug information besides the “Thread was being aborted” message of a scheduled job execution?
Would it be possible to run the job outside an EPiServer context and still access and create pages through the Data Factory? If it has something to do with recycle of the application pool are there any settings that would always allow the job to be fully executed. Could it be that the job is interfered by other scheduled jobs?
Any help would be much appreciated.