I'm making a dashboard in GCP stackdriver. We have a autoscaling node pool which the pods I'm interested in monitoring run in. What I'm wondering is how do I monitor the amount of nodes the are currently running in the pool?
I've had a look at log based metrics, but I can't find anywhere in the logs where it actually says how many nodes are currently running.
There's no metric for number of nodes per se. We can get something similar though with the sum of Total cores
grouped by cluster_name
: