kibana
kibana copied to clipboard
[Metrics UI] Inconsistent metrics across panels
Kibana version: 8.4.3
Elasticsearch version: 8.4.3
Server OS version: Elastic Cloud Belgium GCP N2
Browser version: Latest Chrome
Browser OS version: Mac OSX Monterrey latest
Original install method (e.g. download page, yum, from source, etc.): Elastic Agent installed via rpm
Describe the bug: Inconsistent Metrics across Panels in Infrastructure view, in particular about CPU
Steps to reproduce:
- Install Elastic-Agent with Linux and System integrations on Centos7 machines
- Install a few processes such as Logstash, shell scripts
- Observe Infrastucture > Inventory panels, play around, enable autorefresh
**Bug Description ** In the main page, my server is reporting continuously 78~79% of CPU usage If I click on the server, I can see now in the popped up panel, in Metrics tab, that the CPU is switching all the time between [0%-1%] and [40% - 50] ranges even if I display on the last hour. The scale should be constant and this switching avoided... The result is that we dont know which figure to trust
In this same popped up panel, if I click on Processes, I see only a few processes whereas I expect to see more like htop Also, all processes seems to be in Sleeping whereas they are Running, which means that there may be an issue with those processes.
Conclusion, the popped up panel does not make the job and present inconsistent misleading metrics
Screenshots (if relevant): See attached video
https://user-images.githubusercontent.com/48563314/195881995-e0473ebd-1b95-4909-a386-5b90e5c4467c.mov
Pinging @elastic/unified-observability (Team:Observability)
cc @pmeresanu85
Pinging @elastic/infra-monitoring-ui (Team:Infra Monitoring UI)
@smith @neptunian could we add this to our Bug backlog for 8.7 ?
The first part of the issue is a duplicate of https://github.com/elastic/kibana/issues/133127 Could we rewrite this issue to focus on the Processes tab issues you're seeing?
@pmeresanu85 I added it to refining to make sure it gets rewritten to focus on the part that's not already written in the other issue Milton specified.
@smith thank you for the heads up
Hey @smith should this one go back into the backlog?
Looks like it probably needs redefining/reprioritising? I'm happy to pick this up
🙏🏼 bump to keep this on Product/Dev's radar & requesting release candidate label be updated
Pinging @elastic/obs-ux-infra_services-team (Team:obs-ux-infra_services)
This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.
Closing - we'll be superseding the existing inventory with a new one where we can ensure consistency with metrics