oxheadalpha / tezos-pulumi Goto Github PK

The type of TezosK8sHelmChart.config is pulumi Helm chart config's type ChartOpts | LocalChartOpts. This type is the input properties for a pulumi Helm chart component. This should not be the value of .config on our component. It only needs to be a subset of properties, such as values which is an object of the values that will be applied to the chart. config shouldn't be exposing fields such as skipCRDRendering rendering.

tutorial for deploying rpc nodes on EKS with Pulumi

https://docs.github.com/en/repositories/releasing-projects-on-github/automatically-generated-release-notes

Nginx: https://kubernetes.github.io/ingress-nginx/
I found a commit in another repo of ours that might be a helpful reference: https://github.com/oxheadalpha/pilotcrm/commit/49866f2bd0caa5583020807474414aad883d34b2

Traefik:
https://github.com/traefik/traefik
https://doc.traefik.io/traefik/providers/kubernetes-ingress/
https://traefik.io/blog/aws-elastic-kubernetes-traefik/

We have the ability to deploy custom helm charts.

We need to preserve the ability to deploy custom images as well.

Allow passing custom paths for all tezos-k8s-images. When a custom path is given, use docker.buildAndPushImage to build it and pass the resulting image URL to the helm chart values.

This way we can keep having submodules in oxheadinfra whenever needed.

CI should validate

Currently it automatically deploys the tezos chart. But we have pyrometer, faucet, and the upcoming snapshot charts that should all be able to be deployed.

When I tried to upgrade my cluster to join jakartanet, it failed with the following error:

Diagnostics:
  pulumi:pulumi:Stack (tezos-aws-jakarta):
    error: unable to recognize "/tmp/tmp-1154702CkN2h267UQqZ.tmp": no matches for kind "CustomResourceDefinition" in version "apiextensions.k8s.io/v1beta1"
 
    error: update failed
 
  eks:index:VpcCni (tezos-aws-jakarta-vpc-cni):
    error: Command failed: kubectl apply -f /tmp/tmp-1154702CkN2h267UQqZ.tmp
    error: unable to recognize "/tmp/tmp-1154702CkN2h267UQqZ.tmp": no matches for kind "CustomResourceDefinition" in version "apiextensions.k8s.io/v1beta1"
 
Resources:
    59 unchanged

Duration: 9s

Possibly use the Medium post tutorial

https://cert-manager.io/docs/

https://aws.amazon.com/blogs/security/tls-enabled-kubernetes-clusters-with-acm-private-ca-and-amazon-eks-2/

https://github.com/aws/eks-charts/blob/master/stable/aws-load-balancer-controller/README.md
https://stackoverflow.com/questions/65934606/what-does-eksctl-create-iamserviceaccount-do-under-the-hood-on-an-eks-cluster

One of the benefits is that we are not ever blocked by a release as we may choose to implement the module as it is at some commit.

I had experimented with this. I used npm script "postinstall" to compile the ts files after the module was installed via a git commit url. It worked fine for me locally. Jimmy and Nicolas had issues with it however. It didn't seem that all of the necessary files were being compiled.

We should be able to use the Release component which would avoid us having to run a transformation that prevents the tls cert from regenerating on every pulumi deployment.

We might need to use this as well: https://github.com/kubernetes-sigs/aws-load-balancer-controller/blob/main/helm/aws-load-balancer-controller/values.yaml#L153

https://docs.github.com/en/communities/using-templates-to-encourage-useful-issues-and-pull-requests/configuring-issue-templates-for-your-repository

• tests (none right now)
• formatting / linting
• pkg publishing on release