If a backup job is running when the scheduled time for another instance of the same job comes around, the second instance starts but it usually screws up either (a) what it backs up or (b) the restore chain. An example: a job backs up a large filesystem. It runs in "full" mode once a month, and runs for 36 hours. It is scheduled to run in "incr" mode every day. Consequently on the day after the
"full" is started, it is still running when the "incr" starts. The result is that either the "incr" backs up the wrong list of files (it runs as if it were a "full", or a partial "full"), or the restore chain ends up being wrong (the "full" disappears from it), or something else is messed up.
It is not clear that there is even a well-defined answer as to what should happen in the case of overlapping instances of the same job. Therefore, I would like to see a mechanism built in whereby if a job is still running when a second instance of it wants to start up, that in this situation DP would simply suppress the second instance and not run it at all.