β10-22-2011 03:43 PM
Is anyone experiencing SQL 2005/2008 DB backup failures with ERR - Error <DATABASENAME> is involved in another operation. I have been upgrading all the SQL Clients to 7.1.0.2 from 6.5.5/7.0 and then begin experiencing failures during incremental scheduled backups. Later on during the day when the failed backups are reconciled the back successfuly complete. I have tried changing the scheduling but it did not work. Anyone have any ideas on what logs to enable from NetBackup or SQL?
Solved! Go to Solution.
β10-25-2011 01:42 AM
I dont believe it has changed that much so i would say it is a coincidence
If theyt do transaction logs all day then as far as your NetBackup jobs go you may as well just do your nightly full backups
Hope this has resolved things for you
β10-23-2011 04:17 AM
NBU: dbclient (under ...\veritas\netbackup\logs) and progress logs in ...\veritas\netbackup\logs\user_ops\mssql\logs.
SQL: SQL error logs (ERRORLOG usually located in <install_path>\mssql\log), the VDI log (vdi.log usually in <install_path>\mssql\log), and the Application Event Viewer log.
β10-23-2011 04:55 AM
I see this when TX and DB backup runs in the same time.
I got this error when I had an old vnetd (maybe hung) process.
Restart Netbackup services and related processes and try again.
If you have SQL admin, ask them about some internal backups that runs in the same timeframe
β10-23-2011 02:51 PM
Check that your SQL admins are not running SQL Maintenace schedules and doing their own backups which can conflict.
Also, if there been any problems with backups, check that there are no any orphaned dbbacked.exe processes running when no backups are running.
β10-23-2011 04:48 PM
I need to check with the DBA's to make sure that they are not running those maintenance schedules. If they were orphaned dbbacked.exe processes my ad-hoc backup would also fail. So it leads me to believe that the DBA's are doing something. Since have changed scheduling for some policies and they no longer fail.
I will also look at the vdi and error logs like suggested.
β10-23-2011 09:11 PM
The SQL logs and Application log will report "other operations".
You can then go to SQL dba with the proof and negotiate schedules that will suite all parties.
β10-24-2011 01:31 AM
If your SQL Admins are running their own schedules then bear in mind that if you run transaction log backups during the day that for them to work the NetBackup Full backup must be the last one to have run.
If the Admins run a Full backup after NetBackup then your transaction log backups will fail as they do not have a valid Full backup to work from.
If you find VSS errors in the Windows event logs and they refer to VDI issues then check for VSS Rollup Packages from Microsoft as these exist for most systems and are not part of automatic Windows updates but do help on many occasions
β10-24-2011 08:55 PM
As everyone predicted I found a TLOG backup in the ERRORLOG almost at the same time as the differential backup was executed.
2011-10-22 05:30:01.11 spid98 Setting database option RECOVERY to FULL for database hdcvcl06042 .
2011-10-22 05:30:23.00 Backup Log was backed up. Database: hdcvcl06042, creation date(time): 2011/10/08(01:14:13), first LSN: 7450143:1704:1, last LSN: 7450186:28:1, number of dump devices: 1, device information: (FILE=1, TYPE=DISK: {'\\PATH\TRN_LOGSShared\HDCUNIPRDBSV026\TRN\hdcvcl06042_TRN_2011-10-22 05.30.00.497.bak'}). This is an informational message only. No user action is required.
What I do not understand is why only after upgrading to 7.1 from 6.5.5 has this been occuring. The DBA's have always perform the TLOG backups every 15 mins. Has the mechanism on how backup are queued changed from 655/71 or is it a coincidence?
β10-25-2011 01:42 AM
I dont believe it has changed that much so i would say it is a coincidence
If theyt do transaction logs all day then as far as your NetBackup jobs go you may as well just do your nightly full backups
Hope this has resolved things for you
β10-27-2011 05:06 PM
Talked to my DBA today and I mentioned the situation to him...He still does not understand why is it that a NetBackup SQL backup job does not wait for the TLOG backup to complete before its deamed a failure. He did tell me that he can stop/suspend TLOGs from running but only at a scheduled time or suspend if a backups is already running. Below is the log that I'm using to explain the situation...
1st - The DBA's have TLOG backups every 15 minutes
2nd - I perform fulls once a week and differentials daily except when a full runs. The reason is because we are constantly performing restore for RND purposes so it will not make sense to restore a full plus hundreds of TLOGs.
3rd - I still have quite a few SQL Clients that are running pre-7.0 Clients (which none of those backups fail)