Error during Deployment: "metrics-server" version "5.11.7" not found

[Danbondarenkomv]

Hi everyone,

I am trying to deploy Viya, but I get the error below.

Deployment command:

sudo docker run --rm \
--group-add root \
--user $(id -u):$(id -g) \
--volume /deployments:/data \
--volume /home/ec2-user/viya4-deployment/ansible-vars.yaml:/config/config \
--volume /home/ec2-user/viya4-iac-aws/terraform.tfstate:/config/tfstate \
--volume $HOME/.ssh/id_rsa:/config/jump_svr_private_key \
viya4-deployment --tags "baseline,viya,install"

The error I am getting:

TASK [baseline : Deploy metrics-server] ****************************************
fatal: [localhost]: FAILED! => changed=false
  command: /usr/local/bin/helm --version=5.11.7 --repo=https://charts.bitnami.com/bitnami/ show chart metrics-server
  msg: |-
    Failure when executing Helm command. Exited 1.
    stdout:
    stderr: Error: chart "metrics-server" version "5.11.7" not found in https://charts.bitnami.com/bitnami/ repository
  stderr: |-
    Error: chart "metrics-server" version "5.11.7" not found in https://charts.bitnami.com/bitnami/ repository
  stderr_lines: <omitted>
  stdout: ''
  stdout_lines: <omitted>

The Ansible summary:

PLAY RECAP *********************************************************************
localhost                  : ok=57   changed=4    unreachable=0    failed=1    skipped=38   rescued=0    ignored=0

Wednesday 18 January 2023  16:38:03 +0000 (0:00:01.838)       0:00:17.410 *****
===============================================================================
baseline : Deploy metrics-server ---------------------------------------- 1.84s
jump-server : jump-server - lookup groups ------------------------------- 1.32s
baseline : Check for metrics service ------------------------------------ 1.01s
baseline : Retreive K8s cluster information ----------------------------- 0.99s
jump-server : jump-server - create folders ------------------------------ 0.97s
baseline : Remove deprecated efs-provisioner namespace ------------------ 0.96s
baseline : Retrieve K8s cluster information ----------------------------- 0.88s
Gathering Facts --------------------------------------------------------- 0.86s
baseline : Retreive K8s cluster information ----------------------------- 0.83s
baseline : Deploy ingress-nginx ----------------------------------------- 0.70s
baseline : Deploy ebs-csi-driver ---------------------------------------- 0.68s
common : tfstate - export kubeconfig ------------------------------------ 0.67s
baseline : Deploy nfs-subdir-external-provisioner ----------------------- 0.67s
baseline : Deploy cluster-autoscaler ------------------------------------ 0.58s
baseline : Remove deprecated nfs-client-provisioner --------------------- 0.41s
global tmp dir ---------------------------------------------------------- 0.34s
baseline : Remove deprecated efs-provisioner ---------------------------- 0.30s
baseline role install --------------------------------------------------- 0.10s
jump-server role -------------------------------------------------------- 0.08s
common : Load config file ----------------------------------------------- 0.07s

I have previously provisioned the AWS infrastructure for the deployment using Terraform using this project https://github.com/sassoftware/viya4-iac-aws, and I am deploying Viya on that infrastructure.

Would anyone be able to help me resolve this error?

Thank you

[momiji]

I think last chart version is 3.8.3 (check here: https://artifacthub.io/packages/helm/metrics-server/metrics-server). But it does not work well if kubernetes version is < 1.25 - https://github.com/kubernetes-sigs/metrics-server/issues/1141. In that case use chart version 3.7.0.

[Danbondarenkomv]

Hi @momiji I changed the variable METRICS_SERVER_CHART_VERSION to use version 3.7.0, but it still can't find it:

[Danbondarenkomv]

Hi again @momiji I managed to resolve the issue. The deployment was getting the chart version from here https://artifacthub.io/packages/helm/bitnami/metrics-server, so the latest chart version there is 6.2.6. I have set the METRICS_SERVER_CHART_VERSION variable to the default value suggested here https://github.com/sassoftware/viya4-deployment/blob/main/docs/CONFIG-VARS.md#cert-manager which is 6.2.4 and then that allowed me to resolve the error.

Thank you

[momiji]

Great to hear it works, I didn't thought it was a bitnami chart, but you figured it out. Beware the constraint on the application version, that must not be 0.6.* so far, if you have a kubernetes 1.23 like me, unless the bitnami chart has a specific option that solves the issue, of course. I'll try this chart on our platform to see if the issue is still there.

[riragh]

Based on last comment, user managed to resolve this issue.