Open hemanthnakkina opened 3 months ago
Hm, I couldn't immediately reproducing it using charm-microceph, cf. below
Do you still see this behaviour? If so, would you be able to provide steps for reproducing and logs (juju logs, microceph logs and /v/l/syslog ideally)?
juju deploy microceph -n2 --channel reef/candidate --config default-pool-size=1
juju run microceph/2 add-osd loop-spec=4G,1 --wait=2m
juju run microceph/3 add-osd loop-spec=4G,1 --wait=2m
...
ubuntu@cm-0:~/src/charm-microceph$ juju ssh 2 -- sudo microceph status
MicroCeph deployment summary:
- juju-79f944-2 (10.73.128.231)
Services: mds, mgr, mon, osd
Disks: 1
- juju-79f944-3 (10.73.128.25)
Services: mds, mgr, mon, osd
Disks: 1
Connection to 10.73.128.231 closed.
ubuntu@cm-0:~/src/charm-microceph$ juju status
Model Controller Cloud/Region Version SLA Timestamp
microceph-test localhost-localhost localhost/localhost 3.5.3 unsupported 08:37:35Z
App Version Status Scale Charm Channel Rev Exposed Message
microceph active 2 microceph reef/candidate 73 no
Unit Workload Agent Machine Public address Ports Message
microceph/2* active idle 2 10.73.128.231
microceph/3 active idle 3 10.73.128.25
Machine State Address Inst id Base AZ Message
2 started 10.73.128.231 juju-79f944-2 ubuntu@22.04 Running
3 started 10.73.128.25 juju-79f944-3 ubuntu@22.04 Running
ubuntu@cm-0:~/src/charm-microceph$ juju remove-unit microceph/3
WARNING This command will perform the following actions:
will remove unit microceph/3
Continue [y/N]? y
ubuntu@cm-0:~/src/charm-microceph$ juju status
Model Controller Cloud/Region Version SLA Timestamp
microceph-test localhost-localhost localhost/localhost 3.5.3 unsupported 08:39:08Z
App Version Status Scale Charm Channel Rev Exposed Message
microceph active 1 microceph reef/candidate 73 no
Unit Workload Agent Machine Public address Ports Message
microceph/2* active idle 2 10.73.128.231
Machine State Address Inst id Base AZ Message
2 started 10.73.128.231 juju-79f944-2 ubuntu@22.04 Running
ubuntu@cm-0:~/src/charm-microceph$ juju ssh 2 -- sudo microceph status
MicroCeph deployment summary:
- juju-79f944-2 (10.73.128.231)
Services: mds, mgr, mon, osd
Disks: 1
Removing a microceph unit does not remove the node from microceph cluster.
Deployed sunbeam 2 node with all roles. Removed the non-primary node using [1] However
microceph status
on primary node shows the non-primary node (sunbeam-2) as part of cluster.[1] https://microstack.run/docs/scaling-down