Do a real cleanup of unused files


Active Member
Sep 14, 2012
Vitória da Conquista, Brazil
Guys, I have a PBS on a client that has 94% of datastore usage and to free up space I went to some old backups that I kept only by history and used the "Permanently forget group" option.

I did that and after doing that, I had PBS run Prune and also the Garbege Collector, however the storage continues with exactly the same space used.
I would like to know from friends if I did something wrong or what I could do to make sure that everything is right and how to proceed to clean this space, because I am afraid of running out of space to run jobs during the week.

To make sure I didn't do anything wrong, I went there and had the number of versions of each backup reduced, I ran Prune and Garbege Collector again and the GC result:

2023-04-02T23:46:09-03:00: Removed garbage: 0 B
2023-04-02T23:46:09-03:00: Removed chunks: 0
2023-04-02T23:46:09-03:00: Pending removals: 1.335 TiB (in 774545 chunks)
2023-04-02T23:46:09-03:00: Original data usage: 19.205 TiB
2023-04-02T23:46:09-03:00: On-Disk usage: 2.093 TiB (10.90%)
2023-04-02T23:46:09-03:00: On-Disk chunks: 1232176
2023-04-02T23:46:09-03:00: Deduplication factor: 9.18
2023-04-02T23:46:09-03:00: Average chunk size: 1.781 MiB
2023-04-02T23:46:09-03:00: TASK OK

From what I understand it has Pending removals: 1,335 TiB (in 774545 chunks), but they weren't done! I wanted to know how to force this removal!

To be sure if something was removed or not, I used the command du -hs /mnt/pbs01 and gave the same space used, either before and after running Prune and GC.
search the forum u will find the answer
Searching better I found the information that cleaning is only really performed after 24 hours and 5 minutes of prune apparently.

But in this case I would like to see a way to try to do it before. There was less than 50GB released on PBS, I ended up deleting an extra datastore that we used completely to gain more space and not stop the backups, today we have 200GB, but I don't know if it's enough. =(

You have to wait at least 24h and 5m and then run the GC again:
I had found this information later (here it was already 2 am), but I needed to find a way to reduce this time. I ended up deleting a not so important datastore that I used to free up more space for PBS, if I don't find a way to free up that unused space by 8 pm here in Brazil, I'll have to pray that the 200GB I freed will take care of the backup (I believe so, but you never know what users have added to a fileserver. hehehehe)
of course, there is always a risk left. so u have to extend your datastore pool as fast as possible. good luck!
also u should include your pbs storage in your monitoring system.
Last edited:
of course, there is always a risk left. so u have to extend your datastore pool as fast as possible. good luck!
also u should include your pbs storage in your monitoring system.
Luckily the day after the problem, it only added 60GB to the 200GB free and I was able to wait for the GC to work.

In case my pool meets my needs, I was syncing information from one namespace to another and the process took up most of the PBS space. But after removing these backups, I managed to make the used space drop to 70% which is within my reality.

However, a great option would be to see the possibility of reducing this time for times of need like this.


The Proxmox community has been around for many years and offers help and support for Proxmox VE, Proxmox Backup Server, and Proxmox Mail Gateway.
We think our community is one of the best thanks to people like you!

Get your subscription!

The Proxmox team works very hard to make sure you are running the best software and getting stable updates and security enhancements, as well as quick enterprise support. Tens of thousands of happy customers have a Proxmox subscription. Get your own in 60 seconds.

Buy now!