Add Helm deployment targets to Makefile when using helm plugin

[AlirezaPourchali]

What do you want to happen?

When using the Helm plugin (kubebuilder edit --plugins=helm/v2-alpha), a Helm chart is generated in dist/chart/, but no corresponding Makefile targets are added to work with it. This creates an inconsistency with the Kustomize workflow, which has dedicated targets like make deploy, make undeploy, etc.

Users are left to manually craft helm upgrade --install commands or write their own Makefile targets, which defeats the purpose of scaffolding and creates friction in adoption.

Current Behavior

After running kubebuilder edit --plugins=helm/v2-alpha:

• Helm chart is generated in dist/chart/
• No Makefile targets are added
• README mentions the chart exists but provides no usage examples
• Users must manually write helm commands with proper flags

Expected Behavior

Similar to Kustomize targets (deploy, undeploy, install, uninstall), Helm-specific targets should be scaffolded when the Helm plugin is used.

Proposed Solution

Add a new Makefile section when Helm plugin is active:

##@ Helm Deployment

HELM ?= helm
HELM_NAMESPACE ?= <project>-system
HELM_RELEASE ?= <project>
HELM_VALUES ?= dist/chart/values.yaml

.PHONY: helm-deploy
helm-deploy: ## Deploy controller via Helm
	$(HELM) upgrade --install $(HELM_RELEASE) dist/chart \
		--namespace $(HELM_NAMESPACE) \
		--create-namespace \
		--set manager.image.repository=$(shell echo ${IMG} | cut -d: -f1) \
		--set manager.image.tag=$(shell echo ${IMG} | cut -d: -f2) \
		--values $(HELM_VALUES) \
		--wait \
		--timeout 5m

.PHONY: helm-uninstall
helm-uninstall: ## Uninstall Helm release
	$(HELM) uninstall $(HELM_RELEASE) --namespace $(HELM_NAMESPACE)

.PHONY: helm-status
helm-status: ## Show Helm release status
	$(HELM) status $(HELM_RELEASE) --namespace $(HELM_NAMESPACE)

.PHONY: helm-history
helm-history: ## Show Helm release history
	$(HELM) history $(HELM_RELEASE) --namespace $(HELM_NAMESPACE)

.PHONY: helm-rollback
helm-rollback: ## Rollback to previous Helm release
	$(HELM) rollback $(HELM_RELEASE) --namespace $(HELM_NAMESPACE)

Benefits

1. Consistency - Aligns with existing Kustomize workflow patterns
2. Discoverability - Shows up in make help output
3. Best practices - Demonstrates proper helm upgrade/rollback usage
4. Reduced friction - Users can immediately use Helm after generation
5. Zero-downtime deployments - Makes rolling updates easy out of the box
6. Documentation by example - Makefile serves as executable documentation

Additional Context

The Helm plugin promotes zero-downtime deployments and proper CRD lifecycle management (as mentioned in the docs), but without Makefile integration, many users stick with make deploy (Kustomize) because it's simpler, even though Helm would be better for their use case.

Adding these targets would:

• Encourage Helm adoption for production deployments
• Reduce the learning curve
• Match user expectations (if there's a plugin, the Makefile should know about it)

Alternative Considered

Users can manually add these targets, but:

• It's not obvious what the best practices are
• Defeats the scaffolding benefit
• Creates inconsistency across projects

Related

• The Helm plugin already generates values.yaml with sensible defaults
• README could be updated to reference make helm-deploy instead of raw helm commands
• Could also add environment-specific targets like helm-deploy-dev, helm-deploy-prod (though that might be too opinionated)

Extra Labels

No response

[ashishkpathak-connect]

@camilamacedo86 May I know if I can pick this? Do you've any suggestions/add-ons for this?

[ashishkpathak-connect]

/assign

[camilamacedo86]

Hi @ashishkpathak-connect —

Thank you for looking on that.

I’m not certain we should do this one.

Helm is an optional plugin, and the Makefile can already be customized, so the only way to include this would be to append targets to the end of the Makefile post-scaffold. That might be OK, but if we go that route, I think we should keep it minimal and only add install/uninstall helpers, e.g.:

.PHONY: helm-deploy
helm-deploy: ## Deploy controller via Helm
	$(HELM) upgrade --install $(HELM_RELEASE) dist/chart \
		--namespace $(HELM_NAMESPACE) \
		--create-namespace \
		--set manager.image.repository=$(shell echo ${IMG} | cut -d: -f1) \
		--set manager.image.tag=$(shell echo ${IMG} | cut -d: -f2) \
		--values $(HELM_VALUES) \
		--wait \
		--timeout 5m

.PHONY: helm-uninstall
helm-uninstall: ## Uninstall Helm release
	$(HELM) uninstall $(HELM_RELEASE) --namespace $(HELM_NAMESPACE)

The other Makefile targets feel less related to the project itself and more like general Helm convenience commands, so I don’t think we should go there.

Even for helm-deploy / helm-uninstall, when I weigh the value vs. the cost/maintenance of adding Helm-specific targets by default, I’m still not 100% convinced we should add any. Helm is optional, and teams can already customize their Makefile if they want these helpers.

That said, if you want to take a look and decide, I’m totally fine either way.