[service] [request]: Ability to pull Pause Image if deleted by user/garbage collector

### 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**
In the event a user deletes the pause container, or the garbage collector somehow inadvertently deletes the pause container, the user is able to pull a new pause image without the need to terminate and redeploy worker node(s).

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

**Tell us about the problem you're trying to solve. What are you trying to do, and why is it hard?**
With new K8s 1.24 now being used by EKS, "support for Container Runtime Interface (CRI) for Docker (also known as Dockershim) is removed from Kubernetes 1.24. Amazon EKS official AMIs have containerd as the only runtime. Before moving to Amazon EKS 1.24 or higher, you must remove any reference to bootstrap script flags that aren't supported anymore."

If a user should somehow delete the pause container, "the Kubernetes level provides credentials for pulling Pod images, but these credentials are not used by containerd when it pulls the sandbox image." If the user attempts to pull these images, they will receive an authentication error due to lack of credentials being stored by containerd. This forces the user to have to redeploy nodes to attach/schedule pods.

**Are you currently working around this issue?**
One solution to this is to add the '--pod-infra-container-image ' flag to the EKS 'pause' image. This also ensures the images is not pruned by the image garbage collector.

**Additional context**
None at this time.

**Attachments**
None at this time.

Start with the EKS 1.24 and containerd behavior described in the issue, including the --pod-infra-container-image flag and pause-image credential handling. Done means a deleted pause image can be pulled again without redeploying worker nodes, while the image remains available to the garbage collector.

Written by the indexing model from the issue text.