condensing down a dataset to only the most recent files - question from non-IT professional

[philx509]

I have to start this question by saying that I am not an IT professional who has in-depth knowledge and training on Retrospect.  I'm an "enthusiast" user with reasonable technical knowledge in general who uses Retrospect because it works very well to back up my home LAN, so I'm wiling to spend the money to buy Retrospect Professional.  I need someone to point me in the right direction and then I will gladly read the manual to work out all the details.

 

I use multiple datasets to back up various kinds of data, and I start a new dataset of each type either monthly, quarterly, or yearly.  I have some datasets going back 5+ years now, and I would like to condense them so that I have just one copy of each file in each dataset, the newest version of that file.  How do I go about doing that?  Is this grooming, or it is something else? 

 

Whatever this process, can it be scripted?  Can I select a dataset and then apply this script, without having to edit the script the way I normally do for backup processes?

 

Again, once I am pointed in the right direction, then I will read the manual to work out the details. 

 

Thanks in advance.

 

x509

[Scillonian]

Are you wanting to keep the most recent version of the files that still exist on the volumes that are in the backup set or do you also want to keep the most recent version of files that are no longer existing on the volumes?

[philx509]

Are you wanting to keep the most recent version of the files that still exist on the volumes that are in the backup set or do you also want to keep the most recent version of files that are no longer existing on the volumes?

Scillonian,

 

Good question! If I have a choice, I would like to also keep the most recent version of files that no longer exist on the volume, as long as I can distinguish between the two types of files.  I would be OK with some process that resulted in two "children" files, one each for the files that still exist and no longer exist on the volume.

 

I should have said in my initial post that I'm trying to recover a lot of space devoted to backup of files that aren't the latest version in that dataset.

[Scillonian]

If you only wanted to have the most recent version of files that still exist on the volume(s) remain in the backup set then grooming would work.

 

However if you want to keep specific files that are no longer on the volume(s) this will be more complicated. When grooming a backup set, Retrospect bases whether a file is kept in a backup set based on its presence in a snapshot and provides no facility to mark individual files for retention or deletion from a backup set. (Delete the snapshot containing and the file will be physically removed from the backup set at the next grooming.)

[philx509]

If you only wanted to have the most recent version of files that still exist on the volume(s) remain in the backup set then grooming would work.

 

However if you want to keep specific files that are no longer on the volume(s) this will be more complicated. When grooming a backup set, Retrospect bases whether a file is kept in a backup set based on its presence in a snapshot and provides no facility to mark individual files for retention or deletion from a backup set. (Delete the snapshot containing and the file will be physically removed from the backup set at the next grooming.)

The more that I think about my work habits, if I deleted a file, chances are it was a short-term or temp use, OR the file got renamed to better ID the file.  Which means I would be happy with just the most recent set of files that still existed at the end of the backup period for that dataset.

 

So, tell me if this is what I need to to so that I keep only the most recent version of all files in a backup set:

 

1. I select Configure.

2. I select Backup Sets

3. I select the backup set I want to groom.

4. I click on the Options tab

5. I click the "Groom to remove backups" box and I enter "1" for "older than."

6. I sit back while Retrospect runs the grooming process on that dataset.

7. I repeat this process for each dataset I want to groom.

8. I'm very happy, so I have a few beers.

 

x509

[Scillonian]

From personal experience I would recommend that you have at least 20 GB free on the volume that contains the catalog file and the Retrospect application as the catalog file will temporarily increase dramatically in size during the grooming. (I've seen a 1 GB catalog file grow to 14 GB during grooming!)

 

If the grooming fails a rebuild of the backup set catalog from the backup set media will often allow a second attempt at grooming to succeed.