[Bug]: Helm uninstalls Vault on failure

[itzsaga]

Prior Search

• I have already searched this project's issues to determine if a bug report has already been made.

What happened?

When working through the bootstrapping guide the docs against main the docs state that the Vault deployment may not fail. This happened to me, however, afterwards the chart uninstalled so I was left without any pods or statefulset. It removed everything. I do not believe this is the intended behavior.

Steps to Reproduce

Follow the Deploy Vault instructions.

Have no pf_stack_version or version defined in the environment.yaml.

Relevant log output

╷
│ Error: release vault failed, and has been uninstalled due to atomic being set: context deadline exceeded
│
│   with helm_release.vault,
│   on main.tf line 140, in resource "helm_release" "vault":
│  140: resource "helm_release" "vault" {

[itzsaga]

This isn't explicitly a bug in the module the more I dig in. I think this is a documentation issue. On initial setup we should pass an input of wait = false. Then if it fails as it might the pods stay in they Running state until manually initialized. After they're healthy the input should be removed or changed to true.

I've tested this and confirmed the desired behavior.

[fullykubed]

Yes, you are correct. I'd consider that a documentation bug if the bootstrapping guide doesn't have wait = false set. Can you fix?

[fullykubed]

Might actually be good to add a bootstrap_mode_enabled flag to kube_vault to align with aws_eks and then this input can be set globally earlier in the guide.

[fullykubed]

I updated the wait documentation. We can circle back and see how the installation steps can be even further simplified once the automated installer is ready.