We are running Networker V7.5.1 on Windows 2003 with a Dell PowerVault 136T library attached.
As you can see below, we have a number of tapes that do not seem to be automatically recycling :
volume expires location (%) 000131L2 expired PowerVault136T full 000132L2 expired PowerVault136T full 000133L2 expired PowerVault136T full 000134L2 expired PowerVault136T full 000106L2 expired PowerVault136T full 000107L2 expired PowerVault136T full 000101L2 expired PowerVault136T full 000103L2 expired PowerVault136T full DPN230L4 expired PowerVault136T full DPN231L4 expired PowerVault136T full DPN233L4 expired PowerVault136T full DPN232L4 expired PowerVault136T full 000129L2 26/01/2010 full 000105L2 4/02/2010 full 000130L2 7/02/2010 full 000108L2 10/02/2010 full 000109L2 13/02/2010 full DPN234L4 17/02/2010 full 000104L2 18/02/2010 full 000110L2 21/02/2010 full 000111L2 23/02/2010 full 000112L2 27/02/2010 full 000115L2 28/02/2010 full 000116L2 2/03/2010 full 000117L2 5/03/2010 full 000113L2 9/03/2010 full 000114L2 12/03/2010 full 000120L2 16/03/2010 full 000118L2 19/03/2010 full 000119L2 22/03/2010 full 000122L2 25/03/2010 full 000102L2 4/04/2010 full 000124L2 6/04/2010 full 000121L2 9/04/2010 full 000125L2 1/04/2010 full 000126L2 13/04/2010 full 000127L2 27/04/2010 full 000128L2 2/05/2010 PowerVault136T 100%
e.g. Volume '000129L2' has an expiration date of 26/01/2010, which indeed is the date of the most recent backed up saveset (26/12/2009) plus the configured retention / browse period (one month).
I think this volume should have gotten the 'recyclable' status by now (02/04/2010). Apparently, it did not.
All save sets on this volume are either windows (C:'s, D:'s, SYSTEM FILES:) or 'Networker Module for SQL sever' (MSSQL:)
You will need to check for dependant savesets. There may be savesets on these volumes on which other savesets (not past retention) are dependant. For each incremental backup on a volume, there needs to be a more recent valid full backup of that saveset for the saveset to be expired. Therefore, if for some reason a full backup performed after one of these incremental backups fails, then all of those incrementals (back to the last full) will not be recycled until such time as a more recent valid full backup is done. Hope this is clear.
Thanks for pointing that out. I had thought of that before, but since we are working with a one-week backup cycle (one full and 6 incremental backups) and none of the backups have failed for quite some time now, I don't think there are any dependencies left.
Coming to think of it, there is however one backup-client (the Networker Module for SQL) that runs several times a day in order to keep the log-files under control. This could very well be the culprit. I'll have to check for the last time a full backup of the database was performed.
In my effort to avoid multiple full SQL backups a day, I had specified the 'incr' level in the client overrides section of the backup group (tab 'Advaced'), thus resulting in no full backup taken any more. Mea culpa, mea culpa, mea maxima culpa... Will rectify immediately.
coganb
736 Posts
488
0
Posted April 2nd, 2010 03:00
You will need to check for dependant savesets. There may be savesets on these volumes on which other savesets (not past retention) are dependant. For each incremental backup on a volume, there needs to be a more recent valid full backup of that saveset for the saveset to be expired. Therefore, if for some reason a full backup performed after one of these incremental backups fails, then all of those incrementals (back to the last full) will not be recycled until such time as a more recent valid full backup is done. Hope this is clear.
-Bobby