Description

What problem are you trying to solve?

When a subnet is almost out of IPs, Karpenter will continue to launch nodes in it, leading to the VPC CNI failing to become ready, and the node becoming unready as well. In many cases, there's nothing we can do, but if another subnet has IP addresses, and the workload does not have scheduling constraints that prevent it from running in those zones, we should launch in the subnets with available IPs.

How important is this feature to you?

Managing the ipv4 space is hard, and anything we can do to alleviate these pains would help customers on their path to ipv6.

• Please vote on this issue by adding a 👍 reaction 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