by Richard van Iwaarden.
Just a follow up: our automated backups are now fully running, error free.
So what was happening? We use a webserver (Cpanel) and if you look at php settings, there's a local and global value.
The local value for max_execution_time was set to 3600 seconds. Moodle respected that setting. But we are running the cron to perform the automated course backup procedure, and the cron appeared to use the global value of 480 seconds.
So what was happening? We use a webserver (Cpanel) and if you look at php settings, there's a local and global value.
The local value for max_execution_time was set to 3600 seconds. Moodle respected that setting. But we are running the cron to perform the automated course backup procedure, and the cron appeared to use the global value of 480 seconds.
Because many course backups took longer than 480 seconds, they stopped. moodledata/temp/backup started filling up with data. This is because Moodle uses that temporary directory to make the course backup, and when the course backup is successful (status 1000) Moodle deletes the files from this temp directory.
However, since many course backups timed out, moodledata/temp/backup started filling up with more and more data, running our Moodle out of disk space at some point.
This then lead to the error above (although it could also be https://tracker.moodle.org/browse/MDL-82959)
So I'm marking this as fixed and thank you all for your wise input ![smile smile]()