My rentention policy is the following:
restic -r $RESTIC_REPOSITORY forget --keep-within 1m --prune
yet, the output of restic snapshots is the following:
Applying Policy: keep all snapshots within 1m of the newest
keep 1 snapshots:
ID Time Host Tags Reasons Paths
----------------------------------------------------------------------------------------------
458d9d2a 2024-01-27 17:13:23 host within 1m /home/.snapshots/11
----------------------------------------------------------------------------------------------
1 snapshots
keep 1 snapshots:
ID Time Host Tags Reasons Paths
-----------------------------------------------------------------------------------------------
7b671e1c 2024-03-08 12:35:55 host within 1m /home/.snapshots/346
-----------------------------------------------------------------------------------------------
1 snapshots
keep 1 snapshots:
ID Time Host Tags Reasons Paths
-----------------------------------------------------------------------------------------------
49d182c7 2024-02-29 19:26:01 host within 1m /home/.snapshots/281
-----------------------------------------------------------------------------------------------
1 snapshots
keep 1 snapshots:
ID Time Host Tags Reasons Paths
-----------------------------------------------------------------------------------------------
def595d3 2024-02-25 12:21:34 host within 1m /home/.snapshots/225
-----------------------------------------------------------------------------------------------
1 snapshots
keep 1 snapshots:
ID Time Host Tags Reasons Paths
----------------------------------------------------------------------------------------------
56972a3a 2024-01-31 20:16:16 host within 1m /home/.snapshots/49
----------------------------------------------------------------------------------------------
1 snapshots
keep 1 snapshots:
ID Time Host Tags Reasons Paths
-----------------------------------------------------------------------------------------------
5b3b4ec5 2024-02-09 12:00:01 host within 1m /home/.snapshots/100
-----------------------------------------------------------------------------------------------
1 snapshots
And we are march 8th 2024. Clearly, snapshot 458d9d2a should be deleted. Yet it keeps it with the reason that it is within 1m.
I read the doc, it should be working, so I don’t get it.
By default restic groups your snapshots by host,paths to apply forget policy to each group independently. All your snapshots have different paths, so forget policy applies to each of them separately, leaving alive at least 1 latest snapshot in the each group. You can change this specifying "--group-by host", so all your snapshots will be considereder together. But you should understand the consequences.