Bug 1962276
| Summary: | VM inventory tree is slow to display if the source provider has more than 4000 VMs | ||
|---|---|---|---|
| Product: | Migration Toolkit for Virtualization | Reporter: | David Vaanunu <dvaanunu> |
| Component: | General | Assignee: | Fabien Dupont <fdupont> |
| Status: | CLOSED ERRATA | QA Contact: | Ilanit Stein <istein> |
| Severity: | high | Docs Contact: | Avital Pinnick <apinnick> |
| Priority: | unspecified | ||
| Version: | 2.0.0 | CC: | apinnick, dagur, fdupont, ibragins, istein, jortel, mguetta, mturley |
| Target Milestone: | --- | ||
| Target Release: | 2.0.0 | ||
| Hardware: | Unspecified | ||
| OS: | Unspecified | ||
| Whiteboard: | |||
| Fixed In Version: | Doc Type: | If docs needed, set a value | |
| Doc Text: | Story Points: | --- | |
| Clone Of: | Environment: | ||
| Last Closed: | 2021-06-10 17:12:05 UTC | Type: | Bug |
| Regression: | --- | Mount Type: | --- |
| Documentation: | --- | CRM: | |
| Verified Versions: | Category: | --- | |
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |
| Cloudforms Team: | --- | Target Upstream Version: | |
| Embargoed: | |||
|
Description
David Vaanunu
2021-05-19 16:27:07 UTC
Additional info: There is a slowness also in loading the "Migration details by VM" page for the first time, when migrating from source provider with many VMs (VMware with 4K VMs) In my case, I started a cold migration via API -> opened the "Migration details by VM" page => page loading took ~1.5 minutes * Moving to "Migration plans" page and reopening the "Migration details by VM" => the page is shown immediately * Reloading the "Migration details by VM" page (F5) => takes again too much time Maayan, that's also concerning and I don't believe my PR would fix it. Looking into it now. HI @mturley It was on mgn06 but the cluster is not in a good shape now. I will try to reproduce it on another cluster and will update you with its connection details I found a/the culprit for the high cpu usage when the controller should be idle. Basically it's a situation that results in a hot (infinite) loop in the goroutine that searches for VMs to validate. The search runs every 10 min just in case the watch notification get's overwhelmed. It's a catch-all. I think we pretty much always get into this situation when 1+ large providers are loaded. More detail in the PR description: https://github.com/konveyor/forklift-controller/pull/269 This issue was tested by QE using MTV upstream May 25 2021 latest and we do not see any slowness in VMs discovery or building the inventory tree. We consider this bug as solved. @Fabien, Can this bug move to MODIFIED? @jortel, I assume any backend pieces of this are merged/backported already? I know the UI parts are, so I think we can move it to MODIFIED. (In reply to Mike Turley from comment #10) > @jortel, I assume any backend pieces of this are > merged/backported already? I know the UI parts are, so I think we can move > it to MODIFIED. Yes. Please verify with build 2.0.0-20 / iib:77891 Provider creation passed with no delay at all, picking VMs up when creating a new plan took around 10-12 sec. Validated on build 2.0.2-23. 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 (MTV 2.0.0 images), 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-2021:2381 |