Logging level for errors

Hi, I started to monitor the Server log using PRTG and their file content sensor to pickup on any failing jobs based on the logging level being ERROR but I found that some of the jobs were failing but had INFO as the logging level.

Perhaps a suggestion maybe to identify these as ERROR if possible.

12/09/2017 09:20:00 AM 003398 INFO      Job 'AAAExp' (83e5d9e2-8c6a-4bd2-a262-0d674fb76467) failed with 1 error(s).
12/09/2017 09:20:00 AM 003399 INFO      Job 'BBBExp' (066e0272-dcd0-4fb7-b4e7-a5f169ae44b3) failed with 1 error(s).
12/09/2017 09:20:00 AM 003401 INFO      Job 'CCCExp' (5f0dd6a9-c1ef-4e07-9d50-c02500a1c4ae) failed with 1 error(s).


This is surely a bug. Thank you for reporting it.

I took a better look (after a coffee) and noticed that these messages were in the EasyMorphServer.log the job log files had more detail and identified them as ERRORs

This is helpful, thanks.

Can you please monitor the task logs for now. The bugfix will be available in version 1.2 planned for release before the end of September.

No problems at all. May I make a suggestion in regard to the Job Logs? They seem to get a new name each time the server is restarted and/or when they reach a certain size. Understandable. But it would make it a world easier for me if they rolled instead and kept the same name for the current log.

For example EasyMorphServer-Job.log (as an example) and then either roll on a restart, or a size or a period (eg. at midnight) and date-stamped the rolled log (eg. YYYYMMDD-hhmmss-EM-Job.log (as an example).

That way monitoring can be against a fixed file rather than have it try to figure out which file to read when looking for errors.

Thanks again

My understanding was that a job keeps writing into the same log file unless “Reset log” is clicked (which would add a timestamp to the current log file name and create and new one with the same job ID).

Log file name is hardcoded to be the same as the job ID. Each job is just an XML file named <job-id>.xml. The name doesn’t change even if the job was edited, or the server restarted. If a new job is created (even for the same project) then it will get a new ID. But for existing jobs the IDs shouldn’t change no matter what.

Can you please confirm that a log file name changes on a server restart? If yes, it would be a bug.

Ok. Sorry. maybe I have assumed too much. I will take better notice.