aws / aws/containers-roadmap

[EKS] [CoreDNS add-on]: Respect existing number of replicas during add-on update

Open
#2,540 2 comments 11 reactions 0 assignees View on GitHub
EKS EKS Add-Ons EKS Networking Proposed
Dominant language
Shell
Stars
5.4k
Forks
334
PR merge metrics
No merged PRs in 30d

Description

### Community Note

* Please vote on this issue by adding a 👍 [reaction](https://blog.github.com/2016-03-10-add-reactions-to-pull-requests-issues-and-comments/) to the original issue to help the community and maintainers prioritize this request
* Please do not leave "+1" or "me too" comments, they generate extra noise for issue followers and do not help prioritize the request
* If you are interested in working on this issue or have submitted a pull request, please leave a comment

**Tell us about your request**
EKS add-on for CoreDNS to respect the current number of replicas of the deployment during updates.

**Which service(s) is this request for?**
EKS

**Tell us about the problem you're trying to solve. What are you trying to do, and why is it hard?**
I would like the EKS add-on updates for CoreDNS to not revert to the default replica count of 2, but instead to respect the existing number of replicas, and complete a rolling update based off that value.

**Are you currently working around this issue?**
You can set the "replicaCount" add-on configuration to override the default value of 2.
For example, setting it to 5 would override the default value of 2. If "autoscaling.minReplicas" is set to 10 and "replicaCount" is set to 5, it will terminate 5 pods to meet replicaCount, then scale up to 10 to meet autoScaling minReplicas.

This would not be required if the deployment respected the existing number of replicas.

**Additional context**
To replicate the issue:
1. Update the EKS add on to use autoscaling with minReplicas 5
2. Observe it auto-scale to 5 replicas (from default 2)
3. Make an update to the EKS add-on configuration (can be whitespace change)
4. Observe that it drops the replica count down to 2 (default value), and then back up to 5 (minReplicas value) in < 1 min

**Attachments**
N/A

Contributor guide

Open the contributing guide

Research direction

Start with the EKS CoreDNS add-on update behavior described in the issue, focusing on how the existing replica count, replicaCount, and autoscaling.minReplicas interact. Reproduce the listed update sequence and verify that an update preserves the current count and completes without the temporary drop to the default of two replicas.

Written by the indexing model from the issue text.

Assessment

Tech stack
aws, kubernetes
Domain
cloud, devops
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Active
Clarity
Mostly clear
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.