Which exposes a metrics endpoint for fetching ipam logs and information.
Approach
We should check out this tool and consider installing it on our cluster(s) as it looks like it will be useful is helping us track down and understand better potential ip starvation events during EKS upgrades and node recycles, as we have been bitten by these on several occasions now.
Background
The AWS VPC CNI repo contains the tool
cni-metrics-helper
:https://github.com/aws/amazon-vpc-cni-k8s/blob/master/cmd/cni-metrics-helper/README.md
Which exposes a metrics endpoint for fetching ipam logs and information.
Approach
We should check out this tool and consider installing it on our cluster(s) as it looks like it will be useful is helping us track down and understand better potential ip starvation events during EKS upgrades and node recycles, as we have been bitten by these on several occasions now.
Questions / Assumptions
Definition of done
Reference
How to write good user stories