I have a database with thousands of jobs, two of which on occasion go BROKEN. They fail a couple of times before BROKEN, but what I have failed to identify is why they go BROKEN so early. They are the only two jobs that suffer this problem.

The failures are typically due to database connectivity issues, but this isn't the issue, the issue is that they become BROKEN.

As I understand BROKEN is expected is a job continually fails. I believe something like 15 times in a row. These 2 go BROKEN after a couple of failures.

Can someone point me at something that I may be missing please.