[Solved] "Prune" doesn't del content

sterzy said:
Did you set up a garbage collection job [1]?

[1]: https://pbs.proxmox.com/docs/backup-client.html#garbage-collection
Click to expand...
Hi, strezy. First prune, after garbage.

Code: 
2025-04-16T14:24:44+03:00: prune job 's-603fc6a5-733f'
2025-04-16T14:24:44+03:00: Starting datastore prune on datastore 'PBS_Backup_sql_day', root namespace, down to full depth
2025-04-16T14:24:44+03:00: No prune selection - keeping all files.
2025-04-16T14:24:44+03:00: Pruning group :"host/pbs01"
2025-04-16T14:24:44+03:00: remove host/pbs01/2025-04-16T10:52:58Z
2025-04-16T14:24:44+03:00: remove host/pbs01/2025-04-16T11:06:28Z
2025-04-16T14:24:44+03:00: TASK OK


Code: 
2025-04-16T14:25:06+03:00: starting garbage collection on store PBS_Backup_sql_day
2025-04-16T14:25:07+03:00: Access time update check successful, proceeding with GC.
2025-04-16T14:25:07+03:00: Using access time cutoff 1d 5m, minimum access time is 2025-04-15T11:20:06Z
2025-04-16T14:25:07+03:00: Start GC phase1 (mark used chunks)
2025-04-16T14:25:07+03:00: marked 25% (1 of 4 index files)
2025-04-16T14:25:07+03:00: marked 50% (2 of 4 index files)
2025-04-16T14:25:07+03:00: marked 75% (3 of 4 index files)
2025-04-16T14:25:07+03:00: marked 100% (4 of 4 index files)
2025-04-16T14:25:07+03:00: Start GC phase2 (sweep unused chunks)
2025-04-16T14:25:07+03:00: processed 1% (39 chunks)
2025-04-16T14:25:07+03:00: processed 2% (76 chunks)
2025-04-16T14:25:07+03:00: processed 3% (113 chunks)
2025-04-16T14:25:07+03:00: processed 4% (143 chunks)
2025-04-16T14:25:07+03:00: processed 5% (181 chunks)
2025-04-16T14:25:07+03:00: processed 6% (224 chunks)
2025-04-16T14:25:07+03:00: processed 7% (269 chunks)
2025-04-16T14:25:07+03:00: processed 8% (314 chunks)
2025-04-16T14:25:07+03:00: processed 9% (356 chunks)
2025-04-16T14:25:07+03:00: processed 10% (397 chunks)
2025-04-16T14:25:07+03:00: processed 11% (438 chunks)
2025-04-16T14:25:07+03:00: processed 12% (477 chunks)
2025-04-16T14:25:07+03:00: processed 13% (522 chunks)
2025-04-16T14:25:08+03:00: processed 14% (555 chunks)
2025-04-16T14:25:08+03:00: processed 15% (592 chunks)
2025-04-16T14:25:08+03:00: processed 16% (624 chunks)
2025-04-16T14:25:08+03:00: processed 17% (662 chunks)
2025-04-16T14:25:08+03:00: processed 18% (703 chunks)
2025-04-16T14:25:08+03:00: processed 19% (734 chunks)
2025-04-16T14:25:08+03:00: processed 20% (769 chunks)
2025-04-16T14:25:08+03:00: processed 21% (799 chunks)
2025-04-16T14:25:08+03:00: processed 22% (827 chunks)
2025-04-16T14:25:08+03:00: processed 23% (858 chunks)
2025-04-16T14:25:08+03:00: processed 24% (905 chunks)
2025-04-16T14:25:08+03:00: processed 25% (948 chunks)
2025-04-16T14:25:08+03:00: processed 26% (987 chunks)
2025-04-16T14:25:08+03:00: processed 27% (1030 chunks)
2025-04-16T14:25:08+03:00: processed 28% (1074 chunks)
2025-04-16T14:25:08+03:00: processed 29% (1101 chunks)
2025-04-16T14:25:08+03:00: processed 30% (1139 chunks)
2025-04-16T14:25:08+03:00: processed 31% (1185 chunks)
2025-04-16T14:25:09+03:00: processed 32% (1221 chunks)
2025-04-16T14:25:09+03:00: processed 33% (1258 chunks)
2025-04-16T14:25:09+03:00: processed 34% (1320 chunks)
2025-04-16T14:25:09+03:00: processed 35% (1357 chunks)
2025-04-16T14:25:09+03:00: processed 36% (1392 chunks)
2025-04-16T14:25:09+03:00: processed 37% (1433 chunks)
2025-04-16T14:25:09+03:00: processed 38% (1469 chunks)
2025-04-16T14:25:09+03:00: processed 39% (1505 chunks)
2025-04-16T14:25:09+03:00: processed 40% (1533 chunks)
2025-04-16T14:25:09+03:00: processed 41% (1576 chunks)
2025-04-16T14:25:09+03:00: processed 42% (1621 chunks)
2025-04-16T14:25:09+03:00: processed 43% (1661 chunks)
2025-04-16T14:25:09+03:00: processed 44% (1707 chunks)
2025-04-16T14:25:09+03:00: processed 45% (1751 chunks)
2025-04-16T14:25:09+03:00: processed 46% (1788 chunks)
2025-04-16T14:25:09+03:00: processed 47% (1816 chunks)
2025-04-16T14:25:09+03:00: processed 48% (1870 chunks)
2025-04-16T14:25:09+03:00: processed 49% (1909 chunks)
2025-04-16T14:25:10+03:00: processed 50% (1947 chunks)
2025-04-16T14:25:10+03:00: processed 51% (1981 chunks)
2025-04-16T14:25:10+03:00: processed 52% (2028 chunks)
2025-04-16T14:25:10+03:00: processed 53% (2061 chunks)
2025-04-16T14:25:10+03:00: processed 54% (2098 chunks)
2025-04-16T14:25:10+03:00: processed 55% (2132 chunks)
2025-04-16T14:25:10+03:00: processed 56% (2174 chunks)
2025-04-16T14:25:10+03:00: processed 57% (2213 chunks)
2025-04-16T14:25:10+03:00: processed 58% (2253 chunks)
2025-04-16T14:25:10+03:00: processed 59% (2296 chunks)
2025-04-16T14:25:10+03:00: processed 60% (2331 chunks)
2025-04-16T14:25:10+03:00: processed 61% (2372 chunks)
2025-04-16T14:25:10+03:00: processed 62% (2417 chunks)
2025-04-16T14:25:10+03:00: processed 63% (2448 chunks)
2025-04-16T14:25:10+03:00: processed 64% (2489 chunks)
2025-04-16T14:25:10+03:00: processed 65% (2523 chunks)
2025-04-16T14:25:10+03:00: processed 66% (2558 chunks)
2025-04-16T14:25:11+03:00: processed 67% (2602 chunks)
2025-04-16T14:25:11+03:00: processed 68% (2633 chunks)
2025-04-16T14:25:11+03:00: processed 69% (2676 chunks)
2025-04-16T14:25:11+03:00: processed 70% (2715 chunks)
2025-04-16T14:25:11+03:00: processed 71% (2759 chunks)
2025-04-16T14:25:11+03:00: processed 72% (2785 chunks)
2025-04-16T14:25:11+03:00: processed 73% (2816 chunks)
2025-04-16T14:25:11+03:00: processed 74% (2862 chunks)
2025-04-16T14:25:11+03:00: processed 75% (2896 chunks)
2025-04-16T14:25:11+03:00: processed 76% (2938 chunks)
2025-04-16T14:25:11+03:00: processed 77% (2976 chunks)
2025-04-16T14:25:11+03:00: processed 78% (3015 chunks)
2025-04-16T14:25:11+03:00: processed 79% (3061 chunks)
2025-04-16T14:25:11+03:00: processed 80% (3094 chunks)
2025-04-16T14:25:11+03:00: processed 81% (3136 chunks)
2025-04-16T14:25:11+03:00: processed 82% (3171 chunks)
2025-04-16T14:25:11+03:00: processed 83% (3211 chunks)
2025-04-16T14:25:12+03:00: processed 84% (3249 chunks)
2025-04-16T14:25:12+03:00: processed 85% (3272 chunks)
2025-04-16T14:25:12+03:00: processed 86% (3314 chunks)
2025-04-16T14:25:12+03:00: processed 87% (3357 chunks)
2025-04-16T14:25:12+03:00: processed 88% (3383 chunks)
2025-04-16T14:25:12+03:00: processed 89% (3428 chunks)
2025-04-16T14:25:12+03:00: processed 90% (3466 chunks)
2025-04-16T14:25:12+03:00: processed 91% (3509 chunks)
2025-04-16T14:25:12+03:00: processed 92% (3557 chunks)
2025-04-16T14:25:12+03:00: processed 93% (3603 chunks)
2025-04-16T14:25:12+03:00: processed 94% (3645 chunks)
2025-04-16T14:25:12+03:00: processed 95% (3692 chunks)
2025-04-16T14:25:12+03:00: processed 96% (3722 chunks)
2025-04-16T14:25:12+03:00: processed 97% (3758 chunks)
2025-04-16T14:25:12+03:00: processed 98% (3798 chunks)
2025-04-16T14:25:12+03:00: processed 99% (3832 chunks)
2025-04-16T14:25:12+03:00: Removed garbage: 0 B
2025-04-16T14:25:12+03:00: Removed chunks: 0
2025-04-16T14:25:12+03:00: Pending removals: 7.244 GiB (in 3775 chunks)
2025-04-16T14:25:12+03:00: Original data usage: 302.251 MiB
2025-04-16T14:25:12+03:00: On-Disk usage: 302.253 MiB (100.00%)
2025-04-16T14:25:12+03:00: On-Disk chunks: 95
2025-04-16T14:25:12+03:00: Deduplication factor: 1.00
2025-04-16T14:25:12+03:00: Average chunk size: 3.182 MiB
2025-04-16T14:25:12+03:00: TASK OK

result:

browser_T9mPN1ykcI.png
 
macromoon said:
2025-04-16T14:24:44+03:00: No prune selection - keeping all files.
Click to expand...
Well you did not define any keep options. In that case all backups will be kept, as a prune job that removes everything would be dangerous.

macromoon said:
2025-04-16T14:25:12+03:00: Pending removals: 7.244 GiB (in 3775 chunks)
Click to expand...
This indicates that some backups are going to be removed after the grace period passes [1]. So they have not been marked as unused long enough to actually be removed.

[1]: https://pbs.proxmox.com/docs/maintenance.html#garbage-collection
 
sterzy said:
Well you did not define any keep options. In that case all backups will be kept, as a prune job that removes everything would be dangerous.


This indicates that some backups are going to be removed after the grace period passes [1]. So they have not been marked as unused long enough to actually be removed.

[1]: https://pbs.proxmox.com/docs/maintenance.html#garbage-collection
Click to expand...
sterzy, thank's for fast response. I bkp daily data at night > at day this datastore bkp to tape > at evening del all data from datastore. This bkp as temp folder\datastore for bkp data to tape. I didn't pay attention to the setting 'Keep Hourly' - set 1. Content deleted. How i can change grace period for del metadata (garbage)?
 
sterzy said:
Well you did not define any keep options. In that case all backups will be kept, as a prune job that removes everything would be dangerous.


This indicates that some backups are going to be removed after the grace period passes [1]. So they have not been marked as unused long enough to actually be removed.

[1]: https://pbs.proxmox.com/docs/maintenance.html#garbage-collection
Click to expand...
@sterzy, is possible edit grace period? Yesterday backup doesn't del today after prune 'Keep Hourly' = 1 not solve task delete all bkp.

all Keep is empty:

Code: 
2025-04-17T09:07:20+03:00: Starting datastore prune on datastore 'PBS_Backup_sql_day', root namespace, non-recursive
2025-04-17T09:07:20+03:00: No prune selection - keeping all files.
2025-04-17T09:07:20+03:00: Pruning group :"host/pbs01"
2025-04-17T09:07:20+03:00: remove host/pbs01/2025-04-16T11:06:28Z
2025-04-17T09:07:20+03:00: TASK OK

must del all and in log deleted: 2025-04-17T09:07:20+03:00: remove host/pbs01/2025-04-16T11:06:28Z, but really not.

'Keep Hourly' = 1:

Code: 
2025-04-17T09:09:18+03:00: Starting datastore prune on datastore 'PBS_Backup_sql_day', root namespace, non-recursive
2025-04-17T09:09:18+03:00: retention options: --max-depth 0 --keep-hourly 1
2025-04-17T09:09:18+03:00: Pruning group :"host/pbs01"
2025-04-17T09:09:18+03:00: keep host/pbs01/2025-04-16T11:06:28Z
2025-04-17T09:09:18+03:00: TASK OK
 
Last edited:
1. having an old backup around will greatly speed up the new backup, unless there is no overlap in the backed up data
2. you can remove the whole group if you really want to keep no snapshots
3. since PBS 3.4 , you can change the grace period in the datastore settings (tuning->gc-atime-cutoff)
 
  • Like
Reactions: macromoon
fabian said:
1. having an old backup around will greatly speed up the new backup, unless there is no overlap in the backed up data
2. you can remove the whole group if you really want to keep no snapshots
3. since PBS 3.4 , you can change the grace period in the datastore settings (tuning->gc-atime-cutoff)
Click to expand...
Hi, Fabian! Thank's for response! 1 - ok, 2 - should i del bkp\group via bash only, ant use Prune job? 3 - ok, thank's!
 
you can use a manual prune - if you prune/forget all snapshots, the group will also be removed in recent PBS versions. or you can forget the whole group (not available on the UI, but for example via proxmox-backup-client). but again - I would not recommend removing all snapshots if you will create another snapshot based on it, instead set pruning to --keep-last 1 to benefit from incremental backups.
 
  • Like
Reactions: macromoon
You must log in or register to reply here.
Top Bottom