No space for GC & Prune

K

kalluzerr

Member
Nov 25, 2019
11
0
6
26
Hello,
My backups are not working since the day before yesterday, I just realized that my storage space was full.

1659945568042.png

The problem is that I can't make room, I deleted some backups and some files in /var/log but I am still full and I can't run my garbage collect.

Can you help me please?
 
Hi,
unfortunately deleting backups just frees a small amount of space, because only the metadata is removed (until garbage collection). Removing package files in /var/cache/apt/archives is often a good option if you really need space and have nothing else to delete. Best to make sure the instance is not used by any clients, before the garbage collection can free up space afterwards.
 
Hello Fiona, thank you for your answer, since yesterday I finally managed to make some space, but I have a problem again.
Since I have space, I can't start a garbage collect, the error message is: Error: unable to start garbage collection job on datastore "datastore" - EACCES: Permission denied

Do you know what could be the cause of this?
 
Can you post the full log for the garbage collection task as well as the output of ls -al /path/to/your/datastore? What filesystem do you use?
 
The owner/group should be backup/backup for all files in the datastore.
 
I ran the following command: chown backup:backup -R /pathofmydatastore, it's a bit long, but I already have this:
1660033797992.png
 
Do you still get the permission error? Can you post the full task log when you try running GC?
 
I find it strange that it is a right problem because I type the following command chmod 777 /pathofmydatastore and I always have the same right problem, it is very strange
 
As far as I understand, taking a quick look at the code, the error could also come from the backup-proxy not being able to write the UPID file in the first place, which would also explain the subsequent messages. Could you please check the permissions for your tasks folder or if there is something else wrong with it? ls -l /var/log/proxmox-backup

The output of the full tasklog might still be helpful here, you can find it under Administration > Tasks in the webinterface.

One more thing concerning your storage:
kalluzerr said:
1659945568042.png
Click to expand...
I am assuming the 37TB volume is an ext4 filesystem? As you can see from "Used" and "Avail", only 35TB are available for the datastore, as ext4 by default reserves 5% of the drive for root use only. This space, 1,85TB in your case, is completely unusable for your datastore, as the backup user is unprivileged.
You can change the reserved size with the tune2fs tool.
You can find a good overview and some further links regarding ext4 filesystems, reserved blocks and how to change them here.
 
Last edited:
  • Like
Reactions: fiona
You must log in or register to reply here.
Top Bottom