Nick Baker

@cilindrox are you seeing this only on a specific set of instances types?

We're following a similar trail, but haven't encountered conditions where it seemingly spikes right away as reported here. Is there anything with notable disk I/O running on failing nodes? You...

/ci +workflow:os_distros al2

@khushboo121 are you still having any issues? You should be able to use kubelet flag `--container-log-max-size` on AL2 via [`--kubelet-extra-args`](https://github.com/awslabs/amazon-eks-ami/blob/38c24acb099d6823433cfce30ebf3736a041befe/templates/al2/runtime/bootstrap.sh#L34), or provide `containerLogMaxSize` as a value in the [`nodeadm` kubelet...

> Could we do the same for Neuron and possibly EFA > > ( on mobile so I can't direct link but see lines 83-84 https://github.com/aws-ia/terraform-aws-eks-blueprints/blob/main/patterns/aws-neuron-efa/eks.tf ) thanks for the...

/ci +workflow:k8s_versions 1.33 +build enable_accelerator=nvidia enable_efa=true +test --instance-types=g4dn.xlarge

i pulled up the node logs from the ci and confirmed the labels ```bash # nodeadm logs Jul 07 19:16:23 localhost nodeadm[1961]: {"level":"info","ts":1751915783.9248972,"caller":"kubelet/config.go:212","msg":"Adding node label","label":"nvidia.com/gpu.present=true"} # kubelet logs Jul 07...

@cartermckinnon scoped the version to start once we launch 1.34+ and updated docs, see the preview in https://ndbaker1.github.io/amazon-eks-ami/nodeadm/doc/api-concepts/

> Can you share more details on how you ran into the initial error? AFAICT, cloud-final.service file included in AL2023 does not include the referenced dependency from upstream cloud-init. >...

/ci build +build enable_accelerator=nvidia enable_efa=true