I use Calendar schedules for my full backups and have checkpoints set along with "retries allowed after runday" in the Calendar schedule. In NetBackup 6.0 / 6.5 the Calendar backups would recover from a problem and continue from the checkpoint (second attempt of the job) even after the runday (note backup window of up to 50 hours). When I upgraded to 7.0 the backups would fail and start a new job from the beginning.
I reported this to Symantec several months ago but they are not convinced that I'm seeing a bug as the NetBackup documentation is ambiguous regarding the way that checkpoint / "retreis allowed after runday" work. Has anyone seen this problem ?
Check for Number ofattempts and wait time intervel to move job to Done state parameters on the Master Global Attributes as in below screenshot and adjust them accordingly. If a backup job move to Done state then backup will start from beginning.
Unfortunately we've tried all combinations of these (and other) settings and Symantec have reproduced my problem. But because the documentation is ambiguous they are considering whether the way that jobs resumed in NetBackup 6.0 / 6.5 was incorrect rather than making NetBackup 7.x work the way that NetBackup 6.0 / 6.5 did.
There is a key section in the Admin guide about checkpoints that states "NetBackup starts a new job in the following situations:
* For calendar scheduling, if another run day has arrived."
- and for NetBackup 6.0 / 6.5 this should say 'unless the option "Retries allowed after run day" is set' or at least that was how it was acting.
I'd like to hear if anyone else is affected by the issue in NetBackup 7.x wit hCalendar schedules as that would strengthen the case to provide a fix.