+++ This bug is a downstream clone. The original bug is: +++ +++ bug 1676461 +++ ====================================================================== Description of problem: rhevm-setup plugins should point to KB in case customers hit issues with https://bugzilla.redhat.com/show_bug.cgi?id=1665197 Version-Release number of selected component (if applicable): 4.2.8-1 (Originally by Lukas Svaty)
Let's check if problematic rpm is installed at the beginning of the setup, issue an error pointing to KB and exit. (Originally by Sandro Bonazzola)
(In reply to Sandro Bonazzola from comment #1) > Let's check if problematic rpm is installed at the beginning of the setup, > issue an error pointing to KB and exit. I looked at the referenced bug 1665197 and do not see there any specific problematic rpm we can check about. AFAIU all current versions of ovirt-vmconsole are problematic to _upgrade_ (perhaps also remove?), and they are required by the engine, so we can't exit. A possible workaround for now can be to patch ovirt-engine-setup to require the currently-latest ovirt-vmconsole. This way, it will be upgraded while upgrading engine-setup, and if this fails, it's hopefully less dramatic than if engine-setup fails. (Originally by didi)
(There isn't anymore rhevm-setup) (Originally by didi)
Let's try with the solution proposed by didi. (Originally by Sandro Bonazzola)
(In reply to Yedidyah Bar David from comment #3) > A possible workaround for now can be to patch ovirt-engine-setup to require > the currently-latest ovirt-vmconsole. Francesco, does this make sense? If so, which version should we require? I do not see a bump in the git repo after merging [1]. Can you please bump, and build for 4.2? Thanks. [1] https://gerrit.ovirt.org/97062 > This way, it will be upgraded while > upgrading engine-setup, and if this fails, it's hopefully less dramatic than > if engine-setup fails. (Originally by didi)
Didi, I think it's a great workaround. You can require any version of ovirt-vmconsole. As far as I've seen the error is ignored when invoked on commandline. Francesco, it makes sense to proceed with that regardless when/how you fix ovirt-vmconsole for future (Originally by michal.skrivanek)
(In reply to Michal Skrivanek from comment #9) > Didi, I think it's a great workaround. You can require any version of > ovirt-vmconsole. Not really - if I require e.g. 1.0.4 and the user already has that, engine-setup will still try to upgrade (to 1.0.6.) and might fail. > As far as I've seen the error is ignored when invoked on > commandline. Indeed. We decided to not ignore it in otopi, to be on the safe side - see bug 1493160. > Francesco, it makes sense to proceed with that regardless when/how you fix > ovirt-vmconsole for future (Originally by didi)
Verified on ovirt-engine-4.2.8.5-0.1.el7ev.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/RHBA-2019:0459