Restore jobs stay Queued for a long period of time before going Active

When a restore job is started, NetBackup must search the backup image(s), to build a list of the files to be restored.  The restore job will remain queued while the file list is being built.

If a very large number of files are to be restored, the restore job may stay queued for a long time - a restore job for a few million files will stay in Queued state for several hours before it goes to Active state.

 

If the files are being restored to an alternate location, NetBackup must also build a set of directives, to tell the restore job where the files must be restored - this will further increase the time before the job goes active.

 

The time taken before a restore job goes active depends mainly on the number of files that are being restored.  The larger the number of files to be restored, the longer the job will remain in a Queued state.

 

The time that a restore job stays queued also depends on the specification of the master server (CPU, memory, and the disk holding the image catalog), and the amount of those resources available to the restore job while it is building the list of files to be restored.  If the master server is swapping heavily and/or there are many other jobs running on the master server, this will significantly increase the amount of the time that the job stays in Queued state.

 

For additional information about slow restore performance, and how to identify the backup image(s) required for a restore, please see the documents in the Related Articles section below.