Duplicate jobs and failed/missing backup jobs

[Frogbit]

In BEX10D if a policy created backup job fails in certain ways, a subsequent duplicate job relying on that backup job will fail forever unless manually fixed. For the past few years we have dealt with this situation by removing the affected selection lists used by the policy and adding them back to the policy again. This effectively "resets" the duplicate jobs and things are "normal" again. This method doesn't just reset the failing duplicate job, it resets all the duplicate jobs created by the policy.

So we are moving to BEX2010R3. We also want to use synthetic backups which means one policy will have to replace two policies we currently use. This will mean that if we continue to reset failing duplicate jobs in the way described above we are going to be faced with one humongous mess!!!

The question is:

Is there a way to get a failing duplicate job to forget all the preceding backup jobs it is trying to duplicate without the crude and painful "delete jobs created by policy" and "add jobs created by policy" technique? More importantly can it be done without impacting other duplicate jobs created by the same policy?

Thanks

[pkh · Accepted solution]

What is wrong with using the method that you described to solve the problem?  When you re-generate the jobs, including the duplicate jobs, you are getting back what you have in the first place.  Alternatively, see Colin Weaver's reply to this discussion

https://www-secure.symantec.com/connect/forums/duplicate-catalog-error#comment-5605301

[Frogbit]

Hi PKH,

What's wrong with this method is all the jobs from the selection list are reset. There are daily and weekly jobs that may be waiting thier turn for duplication to tape. If a daily fails and the jobs are reset there is the risk of losing a weekly tape. Currently the daily and weekly jobs are seperate policies and don't interfere with each other, but implementing synthetic backups will mean one policy not two.

Also what happens to those synthetic backups? My guess is that resetting the jobs results in the baseline backup running again?

The link looks interesting: If this allows me to just remove the specific problem and leave everything else intact I will have my answer!

Thanks

[pkh]

 If a daily fails and the jobs are reset there is the risk of losing a weekly tape

Why is this so?  If your weekly job is set to run on Sat and you recreate the jobs on Tue, the next run for your weekly job would still be Sat.

Likewise, with your synthetic backup, if the next scheduled run is Sat, when you recreate the jobs, the next run would still be Sat.

There is no change in the job schedule unless you change the policy.