mirror of
https://github.com/kubernetes/kubernetes.git
synced 2026-05-28 04:04:39 -04:00
68a66f454d
The default size (100) is not enough for large clusters and results in unnecessary restarts of kube-controller-manager watcher for leases, see http://perf-dash.k8s.io/#/?jobname=gce-5000Nodes&metriccategoryname=APIServer&metricname=LoadRequestCount&Resource=leases&Scope=cluster&Subresource=&Verb=LIST This PR will make it match what we have for nodes. |
||
|---|---|---|
| .. | ||
| BUILD | ||
| cachesize.go | ||
| OWNERS | ||