http://jenkins-product.srt.mirantis.net:8080/job/7.0.system_test.ubuntu.thread_3/62/testReport/%28root%29/ceph_ha_restart/ceph_ha_restart/?
Scenario:
1. Revert from ceph_ha
2. Waiting up galera and cinder
3. Check ceph status
4. Run OSTF
5. Destroy osd-node
6. Check ceph status
7. Run OSTF
8. Destroy one compute node
9. Check ceph status
10. Run OSTF
11. Cold restart
12. Waiting up galera and cinder
13. Run single OSTF - Create volume and attach it to instance
14. Run OSTF
In the test we destroy 2 nodes with ceph role. Sometimes it can happen that destroying can be done wile second node was not synced. In the test env replication factor is 2. Obviously necessary up replication factor till 3 to avoid problems like in the test.
Fix proposed to branch: master /review. openstack. org/227173
Review: https:/