[Stack Monitoring] Hide Disk IO charts on when the data is unavaiable.

[simianhacker]

Currently the Disk IO charts are only available if the user is running Elasticsearch natively on Linux (not containerized). For every other deployment the chart will be blank. With the proliferation of Docker and Kubernetes, it probably makes sense to remove this metric from the Node Details page, unless the data actually exists.

See #45870 (comment)

[jasonrhodes]

I would also be okay with displaying a message about how these data are not available in all systems. The situation I think we should avoid would be this chart appearing and disappearing for a user, or a user knowing they have that chart in one cluster and being confused about why it's not there in the other cluster. I'm not sure exactly what the best solution is here, but one or the other, probably.

[simianhacker]

I would support either hiding it OR graying it out with a message that says "Disk IO is not available on his platform" (or something along those lines). Whatever solution we pick, it shouldn't look like a bug like it does today.

[elasticmachine]

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