Upscaling speed and worker config in kubernetes deployment of ray-cluster.yaml

asm582 · April 26, 2021, 8:37pm

Hello,

I see that we specify replicas in ray-cluster.yaml to deploy ray cluster on kubernetes, can you please help where do we specify upscaling speed, min_workers and max_workers in such a deployment?

Thanks

Dmitri · April 28, 2021, 4:47pm

A full example configuration for launching with the Operator can be found here:

github.com

ray-project/ray/blob/master/python/ray/autoscaler/kubernetes/operator_configs/example_cluster.yaml#L7-L12


maxWorkers: 3
# The autoscaler will scale up the cluster faster with higher upscaling speed.
# E.g., if the task requires adding more nodes then autoscaler will gradually
# scale up the cluster in chunks of upscaling_speed*currently_running_nodes.
# This number should be > 0.
upscalingSpeed: 1.0

For launching with the cluster launcher:

github.com

ray-project/ray/blob/master/python/ray/autoscaler/kubernetes/example-full.yaml#L4-L12


# The maximum number of workers nodes to launch in addition to the head
# node.
max_workers: 2


# The autoscaler will scale up the cluster faster with higher upscaling speed.
# E.g., if the task requires adding more nodes then autoscaler will gradually
# scale up the cluster in chunks of upscaling_speed*currently_running_nodes.
# This number should be > 0.
upscaling_speed: 1.0

Topic		Replies	Views
Kubernetes YAML files for manual deployment Ray Clusters	0	351	August 10, 2021
[Autoscaler] Autoscaler behavior for changes to min_workers for deployed cluster Ray Clusters	2	312	June 3, 2021
Autoscaler not scaling up the worker node when using image rayproject/ray:1.11.0-py38 Kubernetes	3	849	July 2, 2022
Scale up from 0 Ray Clusters	7	548	July 15, 2021
Autoscaler doesn't scale workers on K8s	5	669	February 15, 2021

Upscaling speed and worker config in kubernetes deployment of ray-cluster.yaml

Related topics