Feature
The following changes were added in 1.36.200-gke.118:
The following changes were added in 1.36.200-gke.118:
- Fixed an issue in clusters using bundled BGP load balancing where control plane upgrades or configuration updates could cause up to 20 seconds of API server connection failures while a control plane node restarted. After the fix, failover to a healthy control plane node completes in approximately 4 seconds.
- Fixed an issue where node drain in maintenance mode did not wait for attached storage volumes to detach.
Read the original announcement