Description of problem: Since https://bugzilla.redhat.com/show_bug.cgi?id=1408942 host-deploy is requiring (it installs it at execution time since host-deploy get deployed as a tar.gz and not as an rpm) cockpit-ovirt-dashboard and tries to start it. hosted-engine-setup is a bit special on this context: since host-deploy is running while the engine VM is already running on the host, it must avoid any package upgrade since it could kill the running engine VM, that's why it forces host-deploy to run in offline mode. So, if host-deploy requires something to be installed/upgraded, hosted-engine-setup should require it as well to ensure that the package is there before starting the engine VM. Version-Release number of selected component (if applicable): ovirt-hosted-engine-setup-2.1.0.4 ovirt-host-deploy-1.6.2 How reproducible: 100% Steps to Reproduce: 1. run hosted-engine-setup and ensure that the engine appliance contains > ovirt-host-deploy-1.6.2-1.el7ev.noarch Actual results: host-deploy fails: 2017-03-06 20:05:20 DEBUG otopi.plugins.otopi.services.systemd plugin.execute:926 execute-output: ('/bin/systemctl', 'start', 'cockpit.service') stderr: Failed to start cockpit.service: Unit not found. 2017-03-06 20:05:20 DEBUG otopi.context context._executeMethod:142 method exception Traceback (most recent call last): File "/tmp/ovirt-fiObiWABAk/pythonlib/otopi/context.py", line 132, in _executeMethod method['method']() File "/tmp/ovirt-fiObiWABAk/otopi-plugins/ovirt-host-common/cockpit/packages.py", line 70, in _closeup self.services.state('cockpit', True) File "/tmp/ovirt-fiObiWABAk/otopi-plugins/otopi/services/systemd.py", line 141, in state service=name, RuntimeError: Failed to start service 'cockpit' 2017-03-06 20:05:20 ERROR otopi.context context._executeMethod:151 Failed to execute stage 'Closing up': Failed to start service 'cockpit' Expected results: host-deploy successfully completes
How did we miss this?
This will also need a change in cockpit-ovirt in order to avoid circular dependency
Its interesting, but although I've had cockpit-ovirt-dashboard-0.10.7-0.0.11.el7ev.noarch on host, the backup still failed on engine during HE deployment as appears in https://bugzilla.redhat.com/show_bug.cgi?id=1235200
I'm reverting changes from https://bugzilla.redhat.com/show_bug.cgi?id=1408942, breaking too many things at once.
(In reply to Yaniv Dary from comment #1) > How did we miss this? Why missed? Here I've seen that its not working and reported back and reassigned: https://bugzilla.redhat.com/show_bug.cgi?id=1235200#c8
(In reply to Yaniv Dary from comment #1) > How did we miss this? It's just with ovirt-host-deploy-1.6.2 which is from this week. The engine-appliance is probably outdated and so it includes an older ovirt-host-deploy which was fine.
We changed the dependency tree introducing ovirt-host package. Please check everything is ok now.
With manually pre-installed rhvm-appliance-1:4.1.20170811.0-1.el7, ovirt-hosted-engine-setup-2.2.0-0.0.master.20170913082146.git120d800.el7.centos.noarch deployed successfully, with ovirt-host-deploy-1.6.6-1.el7ev.noarch on HE-VM. Moving to verified.
Refining my previous comment #9: Deployment was successful with ovirt-engine-appliance-4.2-20170916.1.el7.centos.noarch installed on RHEL7.4 host with ovirt-hosted-engine-setup-2.2.0-0.0.master.20170913082146.git120d800.el7.centos.noarch and ovirt-host-deploy-1.7.0-0.0.master.20170912090102.git1eeb5a2.el7.centos.noarch was installed on the appliance.
This bugzilla is included in oVirt 4.2.0 release, published on Dec 20th 2017. Since the problem described in this bug report should be resolved in oVirt 4.2.0 release, published on Dec 20th 2017, it has been closed with a resolution of CURRENT RELEASE. If the solution does not work for you, please open a new bug report.