How to configure AWS Observability addon to record pod metrics but not API metrics

I am focusing on the the `/aws/containerinsights//performance` log group as it has alot of putlogevents in the cloudwatch log metrics and is resulting the highest costs.

When I look at the log events they don't look like stuff that I am ever going to query/require (certainly not in my non production environment such as

`apiserver_request_duration_seconds`
`etcd_request_duration_seconds`
`apiserver_request_total`

BUT i do want pod metrics, these in non-prod are very useful

SO basically I want

https://aws-observability.github.io/observability-best-practices/guides/containers/oss/eks/best-practices-metrics-collection/#podcontainer-metrics

but I don't want

https://aws-observability.github.io/observability-best-practices/guides/containers/oss/eks/best-practices-metrics-collection/#control-plane-metrics

How do I configure the Amazon Cloudwatch Observability addon to do this?

I wondered about disabling API server Control Plane logs but the log group that uses (according the AWS Console is `/aws/eks//cluster` . This was confirmed by experimentation as it had no effect on the log events in the performance log group.

It would be good to have a performance logs equivalent to


something like

**Performance Logging**

- enable/disable API Server
- enable/disable rest client
- enable/disable etcd

Start with the Amazon CloudWatch Observability addon configuration and the linked pod/container and control-plane metrics guide; compare the requested metric groups with the addon's documented controls. Done means establishing whether API-server, REST-client, and etcd performance metrics can be independently disabled while pod metrics remain enabled, then documenting the supported configuration or required controls.

Written by the indexing model from the issue text.