by Richard van Iwaarden.
![]()
Does automated hit the same course and fail? That's one thing I've always hated about autobackups ... no way to exclude a course or courses having issues. Auto hits the course having issues, then doesn't recover well ... stops. Builds up each night ... until????
The problem is, the automated backup does not fail nor gives it any errors. The Email clearly says 'Error: 0'. It just stops. I have set the task processing up like this, dont know if this has anything to do with it:
You might have to resort to finding out which courses the largest and then creating a bash shell script to handle just those in one run and then another bash shell script to handle all other others.
When pending adhoc builds up you might never catch up.
Yes, that's my fear. But with only 10 backups per night, we will never catch up. I need to make sure this will run total. Perhaps I need to run the script multiple times, f.e. every 30 minutes at night. I think Moodle will backup around 10 courses every time, and not the same ones.
When it gets that bad, one might need to truncate the tables that track auto. Moodle has no record of anytthing auto then thus it would be like the first time it ran ... but ... you hit those same courses having issues, and back to square one.
Inspection of failed backup directories in moodledata/temp/backups/ might give some clues ... minimally the course ID's having issues and thus could exclude those ID's in the bash shell script that loops through courses and runs moosh to back them up.
moodledata/temp/backups/ only holds 46 files of 0 bytes. Nothing else. I think that 's good no?
18140 courses .. skipped 16014 ... so what's the status of those that were skipped? Are they still being used? There is an option in auto to skip courses that have not changed.
For your other solution: yes, thought about something similar. But I need another summer holiday to make sure no one is working. But if I look at the logs, our Moodle is being used 24/7.