Description of problem: - Currently "overcloud ceph-upgrade run" triggers ceph-ansible playbooks - that cause ceph-osd nodes downtime also on fully updated nodes upon re-execution. - There seems to be no detection on ceph-osd nodes for container version checking - and skipping OSDs shutdown in case they already run the latest container version. - At the moment "overcloud ceph-upgrade run" is unnecessary intrusive - when re-running it (after partial update failure for example). Version-Release number of selected component (if applicable): RHOSP 13 How reproducible: Execute overcloud ceph-upgrade run Actual results: There seems to be no detection on ceph-osd nodes for container version checking - and skipping OSDs shutdown in case they already run the latest container version. Expected results: We should have some detection mechanism/provision in place in case to check the container version, and the upgrades should not re-run on already upgraded ceph nodes.
Can you say what version of ceph-ansible are you using?
@Giulio: [root@os2-prd-director01 ~]# rpm -qa|grep ceph-ansible ceph-ansible-3.1.5-1.el7cp.noarch
This is already possible. A ceph-ansible playbook will run its tasks on the nodes in its inventory. If you don't want it to run those tasks on certain nodes, then the recommendation from the ceph-asnible team is to omit those nodes from the inventory. TripleO will omit nodes from the ceph-ansi8ble inventory if you black list them as described in the following document: https://access.redhat.com/documentation/en-us/red_hat_openstack_platform/13/html-single/director_installation_and_usage/#Scaling-Blacklisting_Nodes
The needinfo request[s] on this closed bug have been removed as they have been unresolved for 1000 days