Breaking change in helm 3.3.2

[githje]

Hi,

I wanted to make you aware of a breaking change introduced in helm version 3.3.2 (https://github.com/helm/helm/releases/tag/v3.3.2):
"The default behavior of helm repo add was changed: It no longer overwrites repositories by default. "

As a result, the deployment scripts deploy_monitoring_cluster.sh and deploy_logging_open.sh will fail with this error:
"Error: repository name (stable) already exists, please specify a different name"

To work around this issue, one can edit the scripts and add --force-update to the helm repo add commands. (however this would cause errors for helm version below 3.3.2)

See here for a discussion: helm/helm#8771

[BryanEllington]

This was fixed in 21a5c2c and cfbd934

[githje]

Hi Brian, thanks for the response, I see the changes you added. However, in my case the scripts actually errored out at that point (git repo add).

[BryanEllington]

Could you provide more information on the error? Do you have the output? Have you pulled the latest version of the repo?

[githje]

Yes, sure. I cloned the repo this ~3hrs ago (all greenfield).

[me@jumphost ops4viya]$ viya4-monitoring-kubernetes/monitoring/bin/deploy_monitoring_cluster.sh
INFO Helm client version: v3.3.3+g55e3ca0
INFO Kubernetes client version: v1.17.9-eks-4c6976
INFO Kubernetes server version: v1.17.9-eks-4c6976
INFO Loading user environment file: /home/centos/ops4viya/site-config-monitoring/monitoring/user.env

Deploying monitoring to the [monitoring] namespace...
INFO Adding 'stable' helm repository
Error: repository name (stable) already exists, please specify a different name

[BryanEllington]

Apologies for the back-and-forth, but can you get the output of these three commands:

• git log -n 5 --format=oneline (from the viya4-monitoring-kubernetes directory)
• helm repo list
• LOG_DEBUG_ENABLE=true viya4-monitoring-kubernetes/monitoring/bin/deploy_monitoring_cluster.sh

[githje]

Sure.

[me@jumphost viya4-monitoring-kubernetes]$ git log -n 5 --format=oneline
2021c33 Refactor samples buried in monitoring and logging directories to a top-level, merged set of samples
cfbd934 Need to 'helm repo ls' not supported in helm 2.x, use 'helm repo list'
21a5c2c Helm 3.3.3+ 'repo add' behavior changed - now fails if repo exists. Add conditional.
efc33b5 Revised logging README
7331a46 Revised removal information in logging README
[me@jumphost viya4-monitoring-kubernetes]$ helm repo list
NAME URL
stable https://kubernetes-charts.storage.googleapis.com
[me@jumphost viya4-monitoring-kubernetes]$ cd ..
[me@jumphost ops4viya]$ LOG_DEBUG_ENABLE=true viya4-monitoring-kubernetes/monitoring/bin/deploy_monitoring_cluster.sh
DEBUG Working directory: /home/centos/ops4viya/viya4-monitoring-kubernetes
DEBUG User directory: /home/centos/ops4viya/site-config-monitoring
INFO Helm client version: v3.3.3+g55e3ca0
INFO Kubernetes client version: v1.17.9-eks-4c6976
INFO Kubernetes server version: v1.17.9-eks-4c6976
DEBUG Temporary directory: [/tmp/sas.mon.uLHx5Zg0]
INFO Loading user environment file: /home/centos/ops4viya/site-config-monitoring/monitoring/user.env

DEBUG TLS is disabled. Skipping verification of cert-manager.
DEBUG Checking for Helm 2 release of [prometheus-monitoring]
DEBUG Found default storageClass: [gp2]
Deploying monitoring to the [monitoring] namespace...
INFO Adding 'stable' helm repository
Error: repository name (stable) already exists, please specify a different name
DEBUG Deleted temporary directory: [/tmp/sas.mon.uLHx5Zg0]

[BryanEllington]

This should now be fixed. The regex for helm repo checks now deals with the tab characters in the output.

[githje]

I did a git pull and tried again. All good, thank you.