* add mgr.x to roles
* to setup mgr and set the require-osd-release bit in osdmap
* do not restart an osd for waiting for healthy: the cluster is not
healthy until the require-osd-release=luminous is set in osdmap, and a
mgr is up and running.
Fixes: http://tracker.ceph.com/issues/20342
Signed-off-by: Kefu Chai <kchai@redhat.com>
- - mon.a
- osd.0
- osd.1
+ - mgr.x
- - mon.b
- mon.c
- osd.2
daemons: [mon.a, mon.b, mon.c, osd.0, osd.1, osd.2, osd.3]
wait-for-healthy: false
wait-for-osds-up: true
- - exec:
- mon.a:
- - ceph osd set require_kraken_osds
- - ceph.restart:
- daemons: [osd.0]
- print: "**** done ceph.restart all"
daemons: [osd.2, osd.3]
wait-for-healthy: false
wait-for-osds-up: true
- - exec:
- mon.a:
- - ceph osd set require_kraken_osds
- - ceph.restart: [osd.3]
- sleep:
duration: 60
+++ /dev/null
-../../../../releases/kraken.yaml
\ No newline at end of file
--- /dev/null
+../../../../releases/luminous.yaml
\ No newline at end of file