[EKS] [Feature]: Allow specifying pause version as part of managed node group
- 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**
What do you want us to build?
A setting allowing setting a non-default (newer) version of the pause container when creating a managed node group.
**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?**
What outcome are you trying to achieve, ultimately, and why is it hard/impossible to do right now? What is the impact of not having this problem solved? The more details you can provide, the better we'll be able to understand and solve the problem.
Would like to be able to create a managed node group where a newer pause container version can be specified, allowing AMI updates while maintaining the desired pause version.
**Are you currently working around this issue?**
How are you currently solving this problem?
Managed node group creation from launch template with specific AMI (EKS latest) and user data for the cluster to include `--pause-container-version 3.x` at the end of the `/etc/eks/bootstrap.sh` command. This requires creating new node groups each time the AMI changes and is closer in behavior to a self-managed node group.
Alternatively, automation that updates a node after it starts using scripts to re-run the bootstrap, restart kubelet, and stop all the pause containers so they will be replaced with the desired version. Container stop accomplished with docker in eks 1.23, but non-trivial with containerd/ctr in 1.24+.
**Additional context**
Anything else we should know?
**Attachments**
If you think you might have additional information that you'd like to include via an attachment, please do - we'll take a look. (Remember to remove any personally-identifiable information.)
Contributor guide
Research direction
Start by reviewing the EKS managed node group request and the documented `/etc/eks/bootstrap.sh` workaround, including the `--pause-container-version 3.x` setting. Done means a managed node group can specify a non-default pause container version while still receiving AMI updates without recreating node groups or rerunning bootstrap automation.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- aws, kubernetes
- Domain
- cloud, infrastructure
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 25/100