SQL jobs are partially successful

[Stanleyj]

I have a job setup for sql that is doing a nightly full backup of the entire instance on several sql servers. All of the sql parent jobs are yellow with a status = 1 and all of the actual streams are 100% successful. I have checked the logs and nothing really indicates where the issue is. All the detailed status gives me is "The job was successfully completed, but some files may have been busy or unaccessible. See the problems report or the client's logs for more details." every sql backup i have setup is doing this? Am i creating the bch file wrong? I do is choose full backup method select the all radio button and then save it.

[Marianne]

Check dbclient log on the client: <install-path>\veritas\netbackup\logs\dbclient. Create the dir if it does not exist and check the log after the next backup.

[Stumpr2]

If you get a status 1 on SQL backups then there is at least one database listed from the batch files that is not getting backed up on the first attempt. I have seen too many netbackup admins passover a status 1 and get burned when a restore request comes for a database that you do not have a backup.

After you determine which database was not backed up you may want to backup just that database. You don't have to rerun the entire backup of all the databases on the server. You only need to backup the one that was missed.

1. You can do an immediate backup from the client of a specific database

2. You can create a new batch file and run that from the master server

3. You can tempoarily modify the current batch file to only backup the missing database and then replace it with the original after you get a successfull  backup.

4. setup auto retry. This will automatically retry a missed database however the parent job will result in a status 1 even though eventually ALL of the databases were successfully backed up

http://www.symantec.com/business/support/index?page=content&id=TECH66111

The method to use is determined by how you want to monitor for future occurrences and how to followup after the issue is resolved. If you rely on method 4 which sounds nice then you may or may not have all the databases covered. You can use the activity monitor to check for the retry after the failure.

I have used all of the above methods and I lean to doing #4 first.

A few years ago I would just rerun the entire backup but when we switched over to charging the customer per GB backed up and coupled with the dwindling tape resources available in the backup window I started using method 1. Also I like the auto retry that came out in Netbackup 6.x

[Stanleyj · Accepted solution]

I opened a ticket with symantec and come to find out I was setting up my policy wrong for the sql jobs. I had all sql servers in one polcy setup to do fulls everynight. In the backup selection tab I had a bch file named uniquely for each server i was attempting to backup. This will cause a status = 1. Because each parent job now searches for those bch files on each server and because only one actually existed on each server it would give a partially backup status. I changed all of my bch files to just say SQL_Full.bch instead of _SQL_Full.bch and that fixed the problem