Feature #55126
open
Though we have MDS_HEALTH_SLOW_METADATA_IO and MDS_HEALTH_SLOW_REQUEST health alert, but those are not
precise nor accumulated. With slow reply counter compared to reply counter, we can find out the ratio
of slow requests through perf dump.
Reviewer suggested that the pr could be backported, so create the tracker here.
- Project changed from Stable releases to Ceph
- Tracker changed from Tasks to Feature
- Project changed from Ceph to CephFS
- Status changed from New to Fix Under Review
- Assignee set to yixing hao
- Backport set to quincy,pacific
- Component(FS) MDS added
- Status changed from Fix Under Review to Resolved
- Backport changed from quincy,pacific to quincy, pacific
- Status changed from Resolved to Pending Backport
- Copied to Backport #55412: pacific: mds: add perf counter to record slow replies added
- Copied to Backport #55413: quincy: mds: add perf counter to record slow replies added
- Tags set to backport_processed
Also available in: Atom
PDF