dashboard
dashboard copied to clipboard
rancher
pod counts inconsistent
Setup
- Rancher version: V2.6.4
- Browser type & version: Firefox 103.0.1 (linux)
Describe the bug The pod count shown on the home page, the cluster dashboard page, and in the cluster workload count are all different
.... and, in my case, wildly and worryingly so.
To Reproduce
Connect to rancher web UI, and click the the appropriate locations
Result
For my case, I have 68/220 on the home page; 1093/220 from the cluster dashboard; and 55 pods listed under the "Workload" menu.
(the cluster has 3 control nodes and 2 worker nodes)
Expected Result I would expect the three values to be reasonably close.
I can also see no way to force a recount (assuming a page-load doesn't do that)
I note that kubectl get pods returns a lesser number, however that will be for the context rather than the whole cluster
In the top nav of cluster explorer, is All Namespaces selected in the top nav? Also, can you share a screenshot of where you see the 1093/220?
Top level "home" screen has a table entry that reads:
| State | Name | Provider | Kubernetes Version | CPU | Memory | pods |
|---|---|---|---|---|---|---|
| (active) | foo | rke | v1.20.9 | 5.38/16 cores | 9.27GiB/47GiB | 65/220 |
(edited to remove all other clusters)
In the "Cluster Dashboard" (and I checked: 'Only User Namespaces' is selected - and changing it makes no difference to the numbers) and "workload" numbers can be seen google-hosted image of cluster dashboard web-page: https://photos.app.goo.gl/WGArWDWiabzFGhNp8
Please can you retest with the latest version (as of writing 2.8.3) and re-open if still applicable
