Releases: kubernetes/autoscaler
Releases · kubernetes/autoscaler
Cluster Autoscaler 1.0.4
- Fixes around metrics in the multi-master configuration.
- Fixes for unready nodes issues when quota is overrun.
Cluster Autoscaler 0.6.3
- Fixes around metrics in the multi-master configuration.
Cluster Autoscaler 1.1.0
- Added Azure support.
- Added support for pod priorities. More details here.
Image: gcr.io/google_containers/cluster-autoscaler:v1.1.0
Addon Resizer 1.8.1
- Re-introduce flags that are used to configure Addon Resizer resources if not specified by config map
Image: gcr.io/google_containers/addon-resizer:1.8.1
Cluster Autoscaler 1.0.3
- Adds support for safe-to-evict annotation on pod. Pods with this annotation can be evicted even if they don't meet other requirements for it.
- Fixes issue when too many nodes with GPUs could be added during scale-up (kubernetes/kubernetes#54959).
Image: gcr.io/google_containers/cluster-autoscaler:v1.0.3
Addon Resizer 1.8.0
- Use Component Config to configure Addon Resizer
Image: gcr.io/google_containers/addon-resizer:1.8.0
Cluster Autoscaler 1.0.2
- Fixes issues with scaling node groups using GPU from 0 to 1 on GKE (#401) and AWS (#321).
- Fixes a bug where goroutines performing API calls were leaking when using dynamic config on AWS (#252).
- Node Autoprovisioning support for GKE (the implementation was included in 1.0.0, but this release includes some bugfixes and introduces metrics and events).
Image: gcr.io/google_containers/cluster-autoscaler:v1.0.2
Cluster Autoscaler 1.0.1
- Fixes a bug in handling nodes that, at the same time, fail to register in Kubernetes and can't be deleted from cloud provider (#369).
- Improves estimation of resources available on a node when performing scale-from-0 on GCE(#326).
- Bugfixes in the new GKE cloud provider implementation.
Image: gcr.io/google_containers/cluster-autoscaler:v1.0.1
Cluster Autoscaler 1.0.0
With this release we are graduating Cluster Autoscaler to GA.
Notable features:
- Support for 1000 nodes running 30 pods each.
- Support for 10 min graceful termination.
- Improved eventing and monitoring.
- Node allocatable support.
- Removed Azure support.
cluster-autoscaler.kubernetes.io/scale-down-disabled
annotation for marking nodes that should not be scaled down.
Compatibility notes:
scale-down-delay-after-delete
andscale-down-delay-after-failure
flags replacedscale-down-trial-interval
Image: gcr.io/google_containers/cluster-autoscaler:v1.0.0
Cluster Autoscaler 0.6.2
- Better behaviour and eventing in case of scale-up problems (like stock-outs).
Image: gcr.io/google_containers/cluster-autoscaler:v0.6.2