This repository has been archived by the owner on Jul 21, 2023. It is now read-only.
EKS cluster and node upgrades #1197
Milestone
Comments
|
We should do this after #1046, because applying a cluster version change affects the Kubernetes cluster resource we use to configure the Kubernetes provider. |
tgeoghegan
added a commit
that referenced
this issue
Jan 19, 2022
Plumbs variables for versions of a few EKS cluster components into
environment .tfvars files so that we can perform upgrades via Terraform.
Specifically we configure:
- The version of the EKS cluster itself. The EKS cluster version is derived
from the upstream Kubernetes version. See [1] for details.
- The version of the AMI used on worker nodes. The EKS API can figure out an
appropriate AMI from the cluster Kubernetes version. The `aws` provider
documentation[2] suggests we could omit this but we explicitly set
the same version as the cluster, since the EKS documentation[3]
states that we have to update the node groups after updating the
cluster.
- The VPC CNI add-on version. We must pick a version that is
compatible with the cluster version, per EKS documentation[4].
- The cluster autoscaler version. We must deploy a version that is
compatible with the cluster Kubernetes version. See [5] for
available versions.
[1]: https://docs.aws.amazon.com/eks/latest/userguide/update-cluster.html
[2]: https://registry.terraform.io/providers/hashicorp/aws/latest/docs/resources/eks_node_group
[3]: https://docs.aws.amazon.com/eks/latest/userguide/update-managed-node-group.html
[4]: https://docs.aws.amazon.com/eks/latest/userguide/managing-vpc-cni.html
[5]: https://github.com/kubernetes/autoscaler/releases
Resolves #1197
tgeoghegan
added a commit
that referenced
this issue
Jan 19, 2022
Plumbs variables for versions of a few EKS cluster components into
environment .tfvars files so that we can perform upgrades via Terraform.
Specifically we configure:
- The version of the EKS cluster itself. The EKS cluster version is derived
from the upstream Kubernetes version. See [1] for details.
- The version of the AMI used on worker nodes. The EKS API can figure out an
appropriate AMI from the cluster Kubernetes version. The `aws` provider
documentation[2] suggests we could omit this but we explicitly set
the same version as the cluster, since the EKS documentation[3]
states that we have to update the node groups after updating the
cluster.
- The VPC CNI add-on version. We must pick a version that is
compatible with the cluster version, per EKS documentation[4].
- The cluster autoscaler version. We must deploy a version that is
compatible with the cluster Kubernetes version. See [5] for
available versions.
[1]: https://docs.aws.amazon.com/eks/latest/userguide/update-cluster.html
[2]: https://registry.terraform.io/providers/hashicorp/aws/latest/docs/resources/eks_node_group
[3]: https://docs.aws.amazon.com/eks/latest/userguide/update-managed-node-group.html
[4]: https://docs.aws.amazon.com/eks/latest/userguide/managing-vpc-cni.html
[5]: https://github.com/kubernetes/autoscaler/releases
Resolves #1197
tgeoghegan
added a commit
that referenced
this issue
Feb 14, 2022
Plumbs variables for versions of a few EKS cluster components into
environment .tfvars files so that we can perform upgrades via Terraform.
Specifically we configure:
- The version of the EKS cluster itself. The EKS cluster version is derived
from the upstream Kubernetes version. See [1] for details.
- The version of the AMI used on worker nodes. The EKS API can figure out an
appropriate AMI from the cluster Kubernetes version. The `aws` provider
documentation[2] suggests we could omit this but we explicitly set
the same version as the cluster, since the EKS documentation[3]
states that we have to update the node groups after updating the
cluster.
- The VPC CNI add-on version. We must pick a version that is
compatible with the cluster version, per EKS documentation[4].
- The cluster autoscaler version. We must deploy a version that is
compatible with the cluster Kubernetes version. See [5] for
available versions.
[1]: https://docs.aws.amazon.com/eks/latest/userguide/update-cluster.html
[2]: https://registry.terraform.io/providers/hashicorp/aws/latest/docs/resources/eks_node_group
[3]: https://docs.aws.amazon.com/eks/latest/userguide/update-managed-node-group.html
[4]: https://docs.aws.amazon.com/eks/latest/userguide/managing-vpc-cni.html
[5]: https://github.com/kubernetes/autoscaler/releases
Resolves #1197
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
On GKE, cluster and node upgrades are automatic. On EKS, upgrades are mostly managed but must be manually triggered by the customer. We can manage EKS cluster and worker node versions via Terraform. We also have to make sure to update the
cluster-autoscalercontainer image version to make sure its minor version matches the Kubernetes control plane minor version. We should plumb variables for cluster andcluster-autoscalerversion up into.tfvarsand document a process for cluster updates.The text was updated successfully, but these errors were encountered: