kubernetes-sigs/cluster-api-provider-aws

AWSMachinePool graceful scale down

Open

#2,574 opened on Jul 8, 2021

View on GitHub
 (19 comments) (0 reactions) (1 assignee)Go (689 forks)auto 404
help wantedkind/featurelifecycle/stalepriority/important-soon

Repository metrics

Stars
 (723 stars)
PR merge metrics
 (PR metrics pending)

Description

/kind feature /help /priority important-soon /milestone v0.7.x

Describe the solution you'd like When using AWSMachinePool and the pool is scaled down (i.e. the ASG is scaled down) i would like to ensure that the pods running on the node are evicted gracefully and that no new pods are scheduled on the node to be terminated. Currently the instance is terminated and the pods are not shutdown gracefully.

We should use a lifecycle hook on the ASG and use the notification to cordon the node and evict the running pods.

Anything else you would like to add: This does not apply to AWSManagedMachinePool as this is handled already by AWS.

Discussed briefly in slack

Environment:

  • Cluster-api-provider-aws version: 0.6.6

Contributor guide

Research direction
Tech stack
Domain
Issue type
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
Prerequisites
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.