Liang Zheng
Liang Zheng
similar issues: https://github.com/goharbor/harbor-helm/issues/174 https://github.com/goharbor/harbor/issues/12135
you can refer the ceph document https://docs.ceph.com/en/latest/cephfs/eviction/#automatic-client-eviction > On an active MDS daemon, if a client has not communicated with the MDS for over session_autoclose (a file system variable) seconds...
hi @palthisailohith, You can check through `dmesg` to see if it contains any `medium error` or `unrecovered read error`.
Are you checking dmesg in the osd host?
Maybe you can try to run pg deep scrub > Then there's deep scrubbing, which is designed to catch bit-rot (aka, the correctness of the PG's), it's naturally a data...
Seems hardware is functioning normally.
重启还不如直接执行 umount 操作
> There should be a way to test for this in https://github.com/distribution/distribution/blob/main/registry/storage/garbagecollect_test.go, I'd like to see a test added for this where a link file is created and then checked...
```bash === RUN TestGCWithUnusedLayerLinkPath unusedlayerlink manifest tags path of repository unusedlayerlink does not exist 0 blobs marked, 4 blobs and 0 manifests eligible for deletion blob eligible for deletion: sha256:91321a43b4fc8f906ebcdf10343a8179773d63df38ab5b7326536129b29239b4...
> @microyahoo is it normal that there is no dry-run protection when deleting the `deleteLayerSet`? > > like adding inside the loop : > > ``` > if opts.DryRun {...