+++ This bug was initially created as a clone of Bug #1649502 +++ Description of problem: ----------------------- When upgrade of RHVH host is initiated from RHV Manager UI, the host is first moved in to maintenance, redhat-virtualization-host-image-update is updated, then the host is rebooted. As part of this upgrade/update procedure, in case, if moving the host in to maintenance fails, there are no proper events or messages notified to the user Version-Release number of selected component (if applicable): ------------------------------------------------------------- RHHI-V 1.5 ( RHV 4.2.7 & RHGS 3.4.1 ) How reproducible: ----------------- Always Steps to Reproduce: ------------------- 1. Stop one of the brick of the volume in host1 2. Try to upgrade host2 from RHV Manager UI Actual results: --------------- Upgrade initiated from RHV Manager UI silently fails without throwing any errors or events Expected results: ----------------- Upgrade should fail with meaningful error or event, so that user will be aware the reason behind the failure. Additional info: ---------------- While moving the host in to maintenance, by stopping gluster service, there are proper error messages that were thrown. The same should be implemented for upgrade/update procedure initiated from RHV Manager UI
The upgrade flow seems to be infra related - moving to infra team
I don't see any flow handling gluster bricks when moving host to maintenance, we are doing only 2 flows: 1. Enabling local maintenance if host can be used to run hosted engine VM 2. Migrate all VMs out of host We have fixed error propagation, when some VM migration will fail as a part of BZ1631215. So how exactly is the flow around stopping gluster bricks affecting moving host to maintenance? Are only VM migration unsuccessful or is there a different error?
(In reply to Martin Perina from comment #2) > I don't see any flow handling gluster bricks when moving host to > maintenance, we are doing only 2 flows: > > 1. Enabling local maintenance if host can be used to run hosted engine VM > 2. Migrate all VMs out of host We also have the flow to stop gluster services when a host is moved to maintenance. > > We have fixed error propagation, when some VM migration will fail as a part > of BZ1631215. > > So how exactly is the flow around stopping gluster bricks affecting moving > host to maintenance? Are only VM migration unsuccessful or is there a > different error? If one of the bricks on another host (h2) is stopped before moving host (h1) to maintenance, the validation fails that the quorum is lost. I think this validation failure is not propogated to the UpgradeHost flow
Martin, is there anything further that needs to be done to ensure validation failures are propogated?
Ravi, could you please verify if this issue is also fixed by BZ1631215
After the recent change for BZ1631215 all validation failures should be picked up by UpgradeHostCallback. So BZ1631215 should fix the issue in this BZ
Aligning status with BZ1631215
INFO: Bug status wasn't changed from MODIFIED to ON_QA due to the following reason: [No relevant external trackers attached] For more info please contact: infra
Same reproduction as https://bugzilla.redhat.com/show_bug.cgi?id=1649502 moving to verified.
This bugzilla is included in oVirt 4.2.8 release, published on January 22nd 2019. Since the problem described in this bug report should be resolved in oVirt 4.2.8 release, it has been closed with a resolution of CURRENT RELEASE. If the solution does not work for you, please open a new bug report.