Backup Job does not run on schedule.

[ioniancat21]

I had a weird occurance this weekend. I recently built a freshly installed BE2010 server. Just to give details, this is a standalone server with 7 windows clients deployed. Anyway, I have a full backup scheduled to run on Friday Evening at 8PM. When I arrived this morning, I noticed that I didn't receive an e-mail from the server with the logs of the completed job so I remoted into the server and low and behold, the backup job never ran at the scheduled time. There were no events showing a failure and I'm wondering if there have been any issues with schedules not running at the designated times with BE2010. Also, I did run the job manually 2 days before it was scheduled and BE2010 did confirm that the job would also run at the scheduled time as well when I launched that job manually.

[tech29]

Hi,

did u checked the job logs? check if there is any   'missed job' by any chance..  Unfortunately Backup exec will not send any notification for missed job so it might be possible that the job indeed missed and you did not get notified for it.

If this is not the case, then check under alert tabs if there are any alerts for media insert or anything related to that job.

if you really don't see any errors.. then try to schedule the job in quetion  to run after say 5 mins. and check if it really starts in next  5 mins are not.. if it starts running after 5 mins.. then u can again schedule it back to the time when u would like this job to run... however, if the job doesn't start after 5 mins.. then the job might have got corrupted.. in this case, click on job setup..  right click the job in question, click on copy this will create a duplicate job of the orignal job with identical settings.. now put the orignal job on hold and schedule the new duplicate job to run after say 5 min.. and that should run fine for you.. if this works then delete the old job and schedule the duplicate job..


Hope this resolves your issue.

let me know if have any queries.

Regards,
tech29

[ioniancat21]

So...............I did as you said and rescheduled the job for 5 minutes in the future and the job started successfully. Still puzzles me how the server could have missed the scheduled backup without reporting an error though...........

I'll add to this message if it becomes a recurring issue and thanks for the assistance...............

[tech29]

Ok.. so did you find a "job missed" log under job log then? there are number of reasons why a job gets missed..

Regards,
tech29

[SystemsAdminAS]

Ionian,

Some more thoughts....

(1) check server date/time - is it accurate?

(2) Is it possible your server was down / restarted during the time slot it was scheduled to run? (Check Windows event logs)

(3) Have you updated BE and applicable clients since installing?  I had intermittent issues with this sort of thing with BE 12.5 in the past

(4) When you set up the job, are you sure the first starting Friday for the job was on the date you thought, or might it have scheduled to start on the following Friday?

(5) Were BE services continually running during your backup window?  Some folks have had issues where BE services intermittently stop and do not restart because of lack of updating or other issues...

(6) Set up more notifications for your jobs, both on success and failure just to help while troubleshooting....

[Frances_Clarke]

I have a problem with jobs missed. I am running 12.5 and have jobs set up through policies. Last weekend, one of the jobs was waiting for an overwritable tape, however, all the other jobs showed as missed because one job was waiting for a tape. I have a Dell ML6030 with 6 tape drives. Also, I tried to erase the tape that was showing as overwritable, but the erase was unable to run while the initial backup job was waiting for a tape.
When the jobs run OK, several jobs are able to run at the same time and use the 6 tape drives available.
Do you have any idea as to why one job waiting for a tape will prevent the other jobs from running ?

[Colin_Weaver]

if the other jobs also used the tape library then they would not run if one is waiting a tape as the control of the whole library is held by the first job. this hold will happen even if you partitioned the slots in the library because the 1st job has a reservation on the library arm as well.

Also although the tape was overwriteable when you checked it after the weekend, it probably was not overwritable when the job actually started - bear in mind that the overwrite proection starts at the end of the last job to use the media

So if (as an example) your backup jobs is a weekly one and starts at 20:00 on a Friday and runs for 5 hours (till 01:00 Saturday) and your overwrite protection is 7 days then the tape can ONLY be used after 01:00 on the following Saturday. The following weekly job still tries to start at 20:00 on the Friday so can't find a tape and asks for one to be made available. When you then check the server on Monday morning it is puzzling as by then the tape will show as overwritable. The way to fix this is if you want a weeks protection from overwrite set the protection to either 6 days (or 6.5 days but defined in hours)


Against the original posters question - in the past we have seen differences between the timings in the selection list availability and the Job schedule can cause problems where a job won't even try and run - so no missed job errors or any other indication. The one that is most confusing is if the day specified in the selectioon list as the "First Available Date" differs from the day specified in the "Effective Date" in the Job Schedule and the job itself does nor run every day then strange schedule issues can be seen.

[Frances_Clarke]

if the other jobs also used the tape library then they would not run if one is waiting a tape as the control of the whole library is held by the first job. this hold will happen even if you partitioned the slots in the library because the 1st job has a reservation on the library arm as well.


The other jobs do use the tape library. We had been using Backup Exec 9.1 and only the job waiting for a tape would be held up, other jobs would run OK.

Also, even the job waiting for a tape has to be cancelled, as the library cannot do an erase or import while one job is waiting for a tape. Again, this was not the case with Backup Exec 9.1 and seems a backward step.

1.  Is there any way to override the reservation on the library arm ?

2.  Also, the job waiting for a tape did not give up, spent the whole weekend waiting for a tape. Is there anyway that the job could give up waiting for a tape after a certain period of time ?

[Colin_Weaver]

1) Not that I am aware of - but then I had no idea that the same scenario did work in 9.1  as from my experience the library lock would occur in various versions of BE (but I haven't really used 9.x)

2) You can set a selection list cancellation based on a time window - test this first though as although I believe we have fixed this one, at least one version of backup exec crashed a process if a job autocancelled whilst a media change request had not been acknowledged against the job being stopped.