Create missing service saturation metrics for services in K8s

The services that we moved over to K8s are missing some saturation metrics that we had available when they were deployed in our VM fleet.

We should get those saturation metrics (or meaningful equivalents in K8s) added back to services deployed in K8s.

Acceptance criteria:

Add (or find a K8s equivalents for) following metrics:

• cpu/single node cpu saturation (gitlab-com/runbooks!3642 (closed))
• node_schedstat_waiting saturation
• disk_space saturation
• disk_inodes saturation
• go_memory saturation: gitlab-com/runbooks!3669 (merged)