CloudWatch Metrics Fails on EKS with ContainerD Runtime
- Dominant language
- Mustache
- Stars
- 1.3k
- Forks
- 1.1k
- Avg merge
- 22m
- Merged PRs (30d)
- 5
Description
https://github.com/aws/eks-charts/blob/5b07b8623f1a3d35f046c8a8c46bf773521c0a64/stable/aws-cloudwatch-metrics/templates/daemonset.yaml#L39
EKS now supports the containerd runtime. However the CloudWatch metrics agent helm charts are missing the required volume mounts. Which results in this error:
`W! No pod metric collected, metrics count is still 19 is containerd socket mounted? https://github.com/aws/amazon-cloudwatch-agent/issues/188`
This link shows what needs to be added:
https://github.com/aws/amazon-cloudwatch-agent/issues/188#issuecomment-803764697
Contributor guide
Research direction
Start at stable/aws-cloudwatch-metrics/templates/daemonset.yaml near line 39 and read the linked amazon-cloudwatch-agent issue and comment for the required containerd volume mounts. Confirm the chart mounts the containerd socket and that the CloudWatch metrics agent no longer reports that the socket is missing.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- aws, helm, kubernetes
- Domain
- infrastructure, observability
- Issue type
- Bug
- Difficulty
- 2/5
- Estimated time
- 1-3 hours
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 55/100