Bug #55088
closed
Manager is failing to keep updated metadata in daemon_state for upgraded MON(s) and takes unpredictable time to get correct metadata.
Added by Prashant D about 2 years ago.
Updated about 1 year ago.
Category:
Administration/Usability
Component(RADOS):
Manager (RADOS bits)
Description
The ceph manager updates mon metadata through handle_mon_map which gets triggered
less frequently, mostly in the case of addition/removal of MONs or on mgr restart or mgr failover.
The daemon_state still holds old metadata for upgraded mon which is being incorrectly reflected in prometheus metrics.
- Description updated (diff)
- Status changed from New to Fix Under Review
- Pull request ID set to 45670
- Status changed from Fix Under Review to Pending Backport
- Copied to Backport #55305: quincy: Manager is failing to keep updated metadata in daemon_state for upgraded MON(s) and takes unpredictable time to get correct metadata. added
- Backport changed from quincy to pacific,quincy
- Copied to Backport #55308: pacific: Manager is failing to keep updated metadata in daemon_state for upgraded MON(s) and takes unpredictable time to get correct metadata. added
- Tags set to backport_processed
- Copied to Bug #57678: Mon fail to send pending metadata through MMgrUpdate after an upgrade resulting in prometheus metrics showing old ceph version for upgraded mon added
- Copied to deleted (Bug #57678: Mon fail to send pending metadata through MMgrUpdate after an upgrade resulting in prometheus metrics showing old ceph version for upgraded mon)
- Status changed from Pending Backport to Resolved
Also available in: Atom
PDF