[Metrics UI] Host name and kubernetes node name are used to join metrics, but don't always match up

[weltenwort]

:lady_beetle: Problem description

Recent investigations into failures to display Kubernetes node metrics suggest that there is a mismatch between the node name added by the add_kubernetes_metadata and the host name added by the add_host_metadata in metricbeat. The mismatch has been observed to be caused by different characteristics:

1. the host name is not lower-cased while the k8s node name is
2. the host name is the FQDN while the k8s node name isn't

Neither is "techically incorrect" according to the respective specs, but causes a problem for us because the inventory and node details use them to join the data.

:question: Potential solutions

It's not clear which would be the correct solution:

1. somehow try to normalize the values in beats
2. find a different field to join the metrics in the UI

[elasticmachine]

Pinging @elastic/infra-monitoring-ui (Team:Infra Monitoring UI)

[miltonhultgren]

@endorama Do you have any more thoughts to add or options we can consider for a solution?

[roshan-elastic]

Prioritising low - we will most likely pick this up when we revisit the asset views for K8s

[elasticmachine]

Pinging @elastic/obs-ux-infra_services-team (Team:obs-ux-infra_services)

[botelastic[bot]]

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.

[smith]

We're not prioritizing this work at this time.