]> git.apps.os.sepia.ceph.com Git - ceph.git/commit
mon/MDSMonitor: enforce mds_join_fs cluster affinity
authorPatrick Donnelly <pdonnell@redhat.com>
Tue, 11 Feb 2020 03:20:08 +0000 (19:20 -0800)
committerPatrick Donnelly <pdonnell@redhat.com>
Thu, 13 Feb 2020 15:51:10 +0000 (07:51 -0800)
commit3fac3b1236c4918e9640e38fe7f5f59efc0a23b9
treee0a803c9e0247509176c8bb12a7c5d9874db571d
parent58a6c4ccb5eb6b823bfa2dfd3eaa008c277ebc10
mon/MDSMonitor: enforce mds_join_fs cluster affinity

Before this, "mds_join_fs" config enforced a preference for the standby
the monitors would select. Now the monitors actively enforce this
by purposefully removing an MDS wither lower "affinity". An MDS standby
has highest affinity if its mds_join_fs is the file system in question
or a vanilla standby (no mds_join_fs).

Fixes: https://tracker.ceph.com/issues/43392
Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
PendingReleaseNotes
src/common/options.cc
src/mds/FSMap.cc
src/mds/FSMap.h
src/mds/MDSMap.cc
src/mds/MDSMap.h
src/mon/FSCommands.cc
src/mon/MDSMonitor.cc
src/mon/MDSMonitor.h