[Bug]: Cannot override .Values.services.neo4j.cleanup.image.tag

[senyahnoj]

Contact Details

senyahnoj@gmail.com

What happened?

Background

I'm using the "neo4j" helm chart and trying to override the version of the bitnami kubectl image which is used in the delete hook for the load balancer service.

To reproduce

To demonstrate the problem in the root of the checked out repository I am running

helm template neo4j --set neo4j.name=test --set volumes.data.mode=dynamic --set neo4j.minimumClusterSize=3 --set neo4j.edition=enterprise --set neo4j.acceptLicenseAgreement=yes --set services.neo4j.cleanup.image.tag=1.33.1 | grep image:

Here is the output, you can see that it isn't using the 1.33.1 tag as per my configuration

The line to fix

There is a variable scoping bug in the template function "cleanup.image"

https://github.com/neo4j/helm-charts/blob/dev/neo4j/templates/_image.tpl#L36

Specifically, within the following section

{{- $termination := printf "%s.%s" $.Capabilities.KubeVersion.Major (regexReplaceAll "\\D+" $.Capabilities.KubeVersion.Minor "") -}}
{{- if not (empty (.tag | trim)) -}}
    {{- $termination := .tag | toString -}}
{{- end -}}

the bit inside the if block should be modifying the outer $termination not creating a new local $termination variable.

So instead of

{{- $termination := .tag | toString -}}

it should be

{{- $termination = .tag | toString -}}

When I make that change in my locally checked out copy, it fixes the problem:

Chart Name

Neo4j

Chart Version

5.12.0

Environment

Local Machine

Relevant log output

Code of Conduct

• I agree to follow this project's Code of Conduct

[colin-telos]

This is blocking our deployment as well. Do you know when the fix will be incorporated and released?

[christopher-wheatley-telos]

Any ETA on a completion for this? It is affecting my project as well.

[senyahnoj]

Also, just for the benefit of anyone following this thread, you'll also be likely affected by the fact bitnami moved all their images yesterday (29th September 2025) bitnami/containers#83267

So docker.io/bitnami/kubectl no longer works (images moved to docker.io/bitnamilegacy/kubectl)

[bfeshti]

@senyahnoj We would suggest to manage LB creation and cleanup with the separate load-balancer chart instead. Using the load-balancer that gets created by default is a legacy setup.

[christopher-wheatley-telos]

@senyahnoj We would suggest to manage LB creation and cleanup with the separate load-balancer chart instead. Using the load-balancer that gets created by default is a legacy setup.

So it is a legacy set up for the 5.x LTS version? Since this is LTS version, I would expect it to still be supported, even if it is legacy code, as it is LTS.

[bfeshti]

@christopher-wheatley-telos This is not about versions, we now have a separate helm chart for managing loadbalancers offering a more robust and dynamic configuration instead of the basic default one.