Replication Director - expiring backups under catalog, snapshots not getting deleted in storage

We have a replication director setup which is a new implementation in our environment.

Configuration is completed successfully backups are going through fine. We want to clear some old snapshots which are done as part of testing and we expired all the backups in catalog. however we are not seeing this getting reflected under NetApp Storage. Its been a week that the catalog is cleaned however the NetApp storage catalogs still reflect.

Along with catalog, do we need to clean anything specific for clean-up ? Any thoughts are highly appreciated.

4 Replies
Highlighted

I belive you use the bpfis

I belive you use the bpfis comman for that

http://www.symantec.com/docs/DOC6489

Page 148-149

 

 

Highlighted

The nbdelete -allvolumes

The nbdelete -allvolumes -force command found in the admincmd runs a clean up of the disk. Expiring images from the catalog technically doesnt remove it from the disk yet. A clean has to be ran. This "should" be automatically run every 12 hours or so.

Give that a shot and let us know

Highlighted

Don't think that will work

Don't think that will work for array based mirrors snap-off. 

The nbdelete trick used netbackup storage unit clean ups, but snapshot are not storage unit in a traditional Netbackup sense.  

Highlighted

If these are replication

If these are replication director snapshots that are not getting deleted automatically even after they are expired from NetBackup then tou would need to check with the NBUPlugin logs on the DFM server with trace enabled.. This should provide you more information why the snapshot isn't getting deleted from there.. you may need to raise a case with NetApp as well since this issue is from the filer side.. Once the image is expired from the NetBackup catalog, nbu will inform the dfm server to remove the snapahot... if this snapshot is not removed it means there is either a communication issue between the DFM server and the Filer or it could also be due to some mis-configuration on the DFM server.