Created attachment 1399797 [details] File contains contents of inventory file, ansible-playbook log Description of problem: During rolling update playbook failing trying to restart (handler) mgr without copying restart script to mgr As per the observation, this issue is happening when the first node that playbook takes up is not the active mgr Version-Release number of selected component (if applicable): ceph-ansible-3.0.26-1.el7cp.noarch How reproducible: Always (2/2) Steps to Reproduce: 1. Configure containerized cluster of (latest-1) version 2. Upgrade to latest version running rolling_update Actual results: RUNNING HANDLER [ceph-defaults : restart ceph mgr daemon(s) - container] ** failed: [magna035 -> magna043] (item=magna043) => {"changed": false, "cmd": "/tmp/restart_mgr_daemon.sh", "item": "magna043", "msg": "[Errno 2] No such file or directory", "rc": 2} Expected results: Playbook must not try to restart daemon without copying respective script Additional info:
will be included in v3.0.27
Working fine with workaround - Mention active mgr's name in top of the mon group (when monitor and mgrs are collocated) in the inventory file and run rolling_update again.
Based on the workaround, retargeting to z2
1) Used - ceph-ansible-3.0.27-1.el7cp.noarch 2) Ensured that active mgr is not the first one to be listed in both mon group and mgr. 3) rolling updated worked fine upgrading cluster from 3.0 live to ceph-3.0-rhel-7-docker-candidate-99411-20180228192608 . Moving to VERIFIED state. Regards, Vasishta Shastry AQE, Ceph
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHBA-2018:0474