- Old (>4 hour) directories are deleted at the beginning of the
execution, so the disk full problem should be out!
- There are 2 types of errors:
- Controlled errors: The backup process finish although something
has been wrong.
- Uncontrolled errors: The backup process dies (memory fault, apache
problem, electric problem...:-)).
- Now scheduled backup is able to detect previous uncontrolled errors to
avoid repeating the same course always. It'll be deferred to the next
execution.
- Every time that a backup cycle has finished, admins will receive an email
with the summary of the execution.
- The 'ERROR' text will be present in email subjects if something was wrong.
With this, all should be ok (I hope).
Merged from MOODLE_14_STABLE
Currently it shows laststatus only.
Detailed view comming this afternoon.
====
Note: As you have released 1.2beta and this is a new little functionality,
please, let me know if I must "delete" it from CVS before 1.2 (although
I think that we are on time to add this...)
