(Note that I got the data from agent, but probably the same with metricbeat)

There are two visualizations on the [Metrics Kubernetes] Overview dashboard that are overriding the index pattern but shouldn't be (they have "yes" checked, but don't specify it). This results in broken visualizations:


(note that the logs-* in the picture isn't actually the default-- there isn't one. That's just what it picked.

the visualizations in question are:

• Desired pods [Metrics Kubernetes]
• Unavailable pods [Metrics Kubernetes]

The radio button for override is set, but if the default index pattern isn't metrics-* (or, as was in my case, isn't set at all) it breaks. It should probably just be set to "no" for the override. Here's a before and after screenshot:

For confirmed bugs, please report:

• Version: 7.15.1 & 7.16
• Operating System: running in kubernetes
• Discuss Forum URL: N/A, slack
• Steps to Reproduce:
  • load kubernetes data (I did so via a policy with k8s integration and Elastic Agent / Fleet)
  • make sure that the default index pattern isn't metrics
  • go to the Kubernetes overview dashboard
  • You should see the errors as above