I create a rancher k8s cluster of 284 nodes, 280 deployments and 100 pods per deployment. After the all 28000 pods up and went in stable state for hours. I found there were logs on rancher-server web page showed the cluster contiuously create and delete pods. But when I login worker nodes to check the containers state, I found they were fine.
So the pod number used is contiuously increasing, till now it’s 192085. But the pod capacity of this cluster is 30910.
Anybody knows something about this?