I am using Proxmox Backup Serve2.3.1. In the past, it was possible to delete VM snapshots in the directory where the data store is mounted, and then enable garbage collection on the web interface to free storage space. However, one day, I was unable to clear the data store space with the same operation. How should I solve this problem
I am using Proxmox Backup Serve2.3.1. In the past, it was possible to delete VM snapshots in the directory where the data store is mounted, and then e
- Thread starter tangshuai
- Start date
If you remove the VM snapshot, you only remove a directory in the datastore containing some index files. The actual data is stored in chunks, stored on the
Simply re-run the GC job on the next day, then the data should be cleared up.
.chunks directory within the datastore. A garbage collection job goes through all stored index files, marks all referenced chunks finally removes chunks that where not referenced by any snapshot. However, the GC will only remove chunks that were unused for at least 24 hours.Simply re-run the GC job on the next day, then the data should be cleared up.
I can give it a tryIf you remove the VM snapshot, you only remove a directory in the datastore containing some index files. The actual data is stored in chunks, stored on the.chunksdirectory within the datastore. A garbage collection job goes through all stored index files, marks all referenced chunks finally removes chunks that where not referenced by any snapshot. However, the GC will only remove chunks that were unused for at least 24 hours.
Simply re-run the GC job on the next day, then the data should be cleared up.
Indeed, it is possible to clear the occupied data storage through the garbage collection function the next day, but I am not sure if there is a better way to solve this problem. Due to cost issues, the PBS storage I have built happens to back up a cluster once. How can I quickly recycle it once a week and then back it up
Do I understand you correctly, you need to remove the exisiting snapshots in order to take a new backup because of a lack of free space?
Sim, porque eu vou copiar o conteúdo deste backup para outro armazenamento e não preciso mais dele, eu vou excluí-lo para liberar espaço e, em seguida, prosseguir para o próximo backup, com um ciclo semanal. Isso resultará em um novo backup semanal da VM. Existe alguma boa maneira de
I've put that through Google Translate now to understand you, but please only write in English in the future.
Okay, so you copy the contents of the datastore somewhere else, I assume as another copy in case of a disaster?
If you really want to force the deletion of chunks earlier than the 24 hours, take a look at this thread. However, read through it carefully, there are good reasons why we do it this way.
https://forum.proxmox.com/threads/force-delete-of-pending-removals.112316/
If in any way possible, I'd recommend to expand the storage of the Backup Server. Remember, the datastore is deduplicated and compressed, to the storage requirements of having multiple snapshots at the same time might not be as big as you expect. It depends on how much data changes between backups.
Okay, so you copy the contents of the datastore somewhere else, I assume as another copy in case of a disaster?
If you really want to force the deletion of chunks earlier than the 24 hours, take a look at this thread. However, read through it carefully, there are good reasons why we do it this way.
https://forum.proxmox.com/threads/force-delete-of-pending-removals.112316/
If in any way possible, I'd recommend to expand the storage of the Backup Server. Remember, the datastore is deduplicated and compressed, to the storage requirements of having multiple snapshots at the same time might not be as big as you expect. It depends on how much data changes between backups.