+++ This bug is a downstream clone. The original bug is: +++ +++ bug 1483078 +++ ====================================================================== Description of problem: Checking for upgrade seems doesn't work in role ovirt-cluster-upgrade. Manual check in webadmin will solve this issue and after that is possible to upgrade checked host via this role. Other hosts are still remain old. Version-Release number of selected component (if applicable): ovirt-ansible-roles-1.0.1-1.el7ev.noarch python-ovirt-engine-sdk4-4.1.6-1.el7ev.x86_64 ovirt-engine-restapi-4.1.5.2-0.1.el7.noarch ovirt-engine-4.1.5.2-0.1.el7.noarch How reproducible: always Steps to Reproduce: 1. add hosts to engine 2. add repositories with newer packages 3. run playbook with role ovirt-cluster-upgrade (Originally by Petr Kubica)
Yeah, we currently upgrade only hosts which reports that they have update available. (Originally by Ondra Machacek)
I don't consider that as urgent. As far as I know, this is due to the behavior in the engine. I don't expect the role to now trigger upgrade check on all hosts. We can maybe add a note clarifying that. (Originally by Oved Ourfali)
I think this role should help users to upgrade their huge environment without clicking on every single host and upgrade it. Role should check and upgrade every host in cluster in a simple way. Now user must click on every host and trigger upgrade check. After that user can use this role for upgrade. So from my perspective, the objective of upgrading hosts in cluster without difficult clicking (per host) isn't now available. (Originally by Petr Kubica)
(In reply to Petr Kubica from comment #3) > I think this role should help users to upgrade their huge environment > without clicking on every single host and upgrade it. Role should check and > upgrade every host in cluster in a simple way. > > Now user must click on every host and trigger upgrade check. After that user > can use this role for upgrade. So from my perspective, the objective of > upgrading hosts in cluster without difficult clicking (per host) isn't now > available. Check for upgrade is executed automatically each day (can be changed using engine-config -s HostPackagesUpdateTimeInHours) for all hosts in Up or NonOperational, so in real setups each host in a cluster should have some available updates (especially when cluster upgrade should be a manually scheduled process). Adding check for upgrade execution for each host prior to its upgrade can increase the time needed for upgrade, so we are still discussing what would be the best solution for that. (Originally by Martin Perina)
Verified in ovirt-ansible-roles-1.0.3-1.el7ev.noarch ansible-2.3.2.0-2.el7.noarch
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/RHEA-2017:2742