Recently I discovered that maintenance was not running due to incorrect configuration, thus blob count/sizes where disproportionately large compared to the source data.
Only one host sends data to the repo, and so I decided to run kopia maintenance run --full --safety=none in an effort to recover space quickly, this achieved the anticipated outcome (without error).
Since that time I noticed that quick maintenance was also disabled, unsure why, but it is now re-enabled. Perfect! Except when I invoke kopia maintenance run the following error is thrown.
Running full maintenance...
Looking for active contents...
Processed 4042 contents, discovered 8623...
Processed 15034 contents, discovered 30880...
Processed 31081 contents, discovered 44058...
Processed 49351 contents, discovered 55592...
Processed 64548 contents, discovered 66015...
Processed 67214 contents, discovered 67214...
Looking for unreferenced contents...
Rewriting contents from short packs...
Found safe time to drop indexes: 2021-10-23 13:05:33.8730227 +0100 BST
Dropping contents deleted before 2021-10-23 13:05:33.8730227 +0100 BST
Finished full maintenance.
ERROR error dropping deleted contents: error performing compaction: error advancing deletion watermark: deletion watermark time cannot move backwards
Here is the output of kopia maintenance info
Owner: kopia@euw-iws065
Quick Cycle:
scheduled: true
interval: 1h0m0s
next run: 2021-10-23 20:58:46 BST (in 49m13s)
Full Cycle:
scheduled: true
interval: 24h0m0s
next run: 2021-10-24 19:58:46 BST (in 23h49m13s)
Log Retention:
max count: 10000
max age of logs: 720h0m0s
max total size: 1 GiB
Recent Maintenance Runs:
full-rewrite-contents:
2021-10-23 19:58:52 BST (15s) SUCCESS
2021-10-23 14:05:46 BST (2m23s) SUCCESS
2021-10-23 13:47:56 BST (4m33s) SUCCESS
2021-10-22 17:16:21 BST (3s) SUCCESS
2021-10-22 15:35:38 BST (2m9s) SUCCESS
snapshot-gc:
2021-10-23 19:58:46 BST (5s) SUCCESS
2021-10-23 14:05:33 BST (12s) SUCCESS
2021-10-23 13:47:49 BST (7s) SUCCESS
2021-10-22 19:12:50 BST (5s) SUCCESS
2021-10-22 17:16:15 BST (5s) SUCCESS
cleanup-logs:
2021-10-23 14:08:17 BST (0s) SUCCESS
2021-10-23 13:52:30 BST (0s) SUCCESS
2021-10-22 19:12:57 BST (0s) SUCCESS
2021-10-22 17:16:24 BST (0s) SUCCESS
2021-10-22 17:12:49 BST (0s) SUCCESS
full-delete-blobs:
2021-10-23 14:08:10 BST (7s) SUCCESS
2021-10-22 19:12:56 BST (0s) SUCCESS
2021-10-22 17:12:45 BST (3s) SUCCESS
2021-10-22 15:29:55 BST (1s) SUCCESS
full-drop-deleted-content:
2021-10-23 19:59:07 BST (0s) ERROR: error performing compaction: error advancing deletion watermark: deletion watermark time cannot move backwards
2021-10-23 14:08:09 BST (0s) SUCCESS
2021-10-23 13:52:30 BST (0s) SUCCESS
2021-10-22 17:16:24 BST (0s) SUCCESS
2021-10-22 17:12:45 BST (0s) SUCCESS
Do I need to worry about this, or will the error self-rectify given enough time?