[updatecli] Bump jenkins-infra helmfile docker image

[jenkins-infra-updatecli]

[updatecli] Bump jenkins-infra helmfile docker image

Report

Source:
	✔ [lastRelease] Get latest jenkinsciinfra/helmfile docker image version(githubrelease)


Condition:
	✔ [checkDockerImage] Ensure that the image "jenkinsciinfra/helmfile:<found_version>" is published(dockerimage)
	✔ [checkPodTemplate] Is container 'helmfile' correctly defined in PodTemplates.yaml(yaml)

Target:
	✔ [updatePipelineImage] Update docker-helmfile in Jenkinsfile(file)
	✔ [updatePodImage] Update helmfile container's image in PodTemplates.yaml(yaml)

Changelog

Click to expand

Release published on the 2022-06-08 15:27:59 +0000 UTC at the url https://github.com/jenkins-infra/docker-helmfile/releases/tag/2.4.28

<!-- Optional: add a release summary here -->
## 👻 Maintenance

* cleanup: remove aws-iam-authenticator (not needed anymore) and fix updatecli manifests for 0.25.0 (#114) @dduportal

Remark

This pull request was automatically created using Updatecli.

Please report any issues with this tool here

[dduportal]

Interesting: the checks are failing for the EKS cluster with Failed to get release autoscaler in namespace autoscaler: exit status 1: Error: Kubernetes cluster unreachable: exec plugin: invalid apiVersion "client.authentication.k8s.io/v1alpha1"

[dduportal]

This PR fails the check on cik8s which is hosted on EKS with the error exit status 1: Error: Kubernetes cluster unreachable: exec plugin: invalid apiVersion "client.authentication.k8s.io/v1alpha1"

It seems that we are hit by https://discuss.kubernetes.io/t/error-exec-plugin-invalid-apiversion-client-authentication-k8s-io-v1alpha1/19916 (aws/aws-cli#6920 + aws/aws-cli#6916).

Haven't checked more.

[dduportal]

Blocking while experimenting

[dduportal]

After succesfully reproducing the issue locally, following actions to verify in the context of Jenkins:

• Manually changed the kubeconfig credential associated to cik8s (s/apiVersion: client.authentication.k8s.io/v1alpha1/apiVersion: client.authentication.k8s.io/v1beta1) as per aws eks update-kubeconfig invalid apiVersion aws/aws-cli#6920 (comment)
• Launched a new job (Checkout scm from Jenkinsfile #4) on this PR to validate that the new credential works as expected

[dduportal]

Build failed with a new error: the aws cli version that we have on the Docker image is outdated: It's the latest 1.19 from Alpine Linux 3.15 (https://pkgs.alpinelinux.org/packages?name=aws-cli&branch=v3.15), while I'm using the 2.7.4.

Latest v1 version is 1.22.87 as Pip: let's try with this one

[dduportal]

ok, with a manual test, it seems that everything works as expected with the version 1.25.4 (latest available in py-pip):

+ aws --version
aws-cli/1.25.4 Python/3.9.7 Linux/5.4.0-1078-azure botocore/1.27.4
+ kubectl cluster-info
<redacted>

Let's bump the AWS AWS cli tools versions in the image jenkinsciinfra/helmfile a trigger a new release

[dduportal]

• AWS CLI bump: feat(aws-cli) Bump from 1.19 native package to 1.25.4 pip package docker-helmfile#112. lets review, approve, merge and wait for the release to be ^picked here.
• As seen with @smerle33 , I've rollbacked the manual change on the cik8s kubeconfig change as it was breaking all the main branch builds.

[dduportal]

• Tested with the latest aws-cli in 1.25.4 + manually updated credential: works as expected
• Commit https://github.com/jenkins-infra/charts-secrets/commit/f7a2e2f519ae97460d67b4af7e02ec06b3bc4673 (private repo) persists the api version change in the credential for cik8s

[dduportal]

LGTM 🚀

• Code change is ok
• Checks are ok
• Diff is ok