Actions
Bug #62494
openLack of consistency in time format
Status:
Pending Backport
Priority:
Normal
Assignee:
Category:
Correctness/Safety
Target version:
% Done:
0%
Source:
Tags:
backport_processed
Backport:
reef,quincy,pacific
Regression:
No
Severity:
3 - minor
Reviewed:
Affected Versions:
ceph-qa-suite:
Component(FS):
mgr/snap_schedule
Labels (FS):
snapshots
Pull request ID:
Crash signature (v1):
Crash signature (v2):
Description
I wanted to test cephfs snapshots in latest Reef and noticed a discrepancy when it comes to time format, for example this works:
$ ceph fs snap-schedule add test-snap-dir1 1h Schedule set for path test-snap-dir1
While minutes are not allowed, but only if you specify capital "M", otherwise it's not parsed:
$ ceph fs snap-schedule add test-snap-dir1 1m Error ENOENT: schedule multiplier "m" not recognized $ ceph fs snap-schedule add test-snap-dir2 2M Error ENOENT: no minute snaps allowed
I would read capital "M" as "Month(s)", not "Minutes". Especially if one compares it with other schedules, e. g. rbd mirror:
$ rbd -p pool1 mirror snapshot schedule add 60m
Or the general 'ceph status' output for the uptime of daemons:
mon: 3 daemons, quorum host3,host2,host1 (age 2d) mgr: host2(active, since 17m), standbys: host1, host3 mds: 1/1 daemons up, 5 standby osd: 9 osds: 9 up (since 2d), 9 in (since 6M)
This is a little confusing and should be the consistent across all entities using time formats.
Actions