chore(k8s): hold off pipelines in dev (replicaCount=0)

[thatcatfromspace]

Summary

• Set replicaCount: 0 in k8s/dev/values.yaml so the dev Deployment scales to zero.
• Pipelines is not required to be operational in dev right now and recent rollouts have been failing health checks — scaling to zero keeps ArgoCD reconciling cleanly without churning unhealthy pods.
• HPA is disabled in the base chart (and not enabled by the dev overlay), so replicaCount: 0 is not overridden.

To bring the service back: set replicaCount: 1 (or remove the override) and resync the ArgoCD app.

Test plan

• helm lint k8s/chart -f k8s/dev/values.yaml clean
• After merge, confirm ArgoCD pipelines app reports Healthy with 0/0 pods

Co-Authored-By: Claude Opus 4.7 (1M context) noreply@anthropic.com

Greptile Summary

Sets replicaCount: 0 in the dev Helm overlay to scale the pipelines Deployment to zero while health checks are stabilising in dev. No code, schema, or API changes are included.

• Adds a single replicaCount: 0 line to k8s/dev/values.yaml with an inline comment explaining the hold-off and how to reverse it.
• The dev-only overlay means prod and staging are unaffected; the base chart's HPA is not enabled for dev, so the zero replica count is not overridden at runtime.

Confidence Score: 5/5

Safe to merge — single-line Helm overlay change affecting only the dev environment with no impact on production or staging.

The change touches one value in a dev-only Helm overlay. It scales the dev Deployment to zero, which is intentional and documented. Prod and staging are unaffected. The comment in the file and the PR description both explain the rationale and how to reverse it.

No files require special attention.

Important Files Changed

Filename	Overview
k8s/dev/values.yaml	Adds replicaCount: 0 to scale the dev deployment to zero; change is minimal, well-commented, and scoped only to the dev overlay.

Flowchart

%%{init: {'theme': 'neutral'}}%%
flowchart TD
    A[ArgoCD reconcile loop] --> B{replicaCount in dev overlay}
    B -- "0 (current)" --> C[Deployment scaled to 0 pods]
    C --> D[ArgoCD reports Healthy — no pods churning]
    B -- "1 (to restore)" --> E[Deployment scaled to 1 pod]
    E --> F[Readiness probe passes?]
    F -- Yes --> G[Pod Running / ArgoCD Healthy]
    F -- No --> H[Pod fails health check / ArgoCD Degraded]

Loading

_{Reviews (1): Last reviewed commit: "chore(k8s): hold off pipelines in dev (r..." | Re-trigger Greptile}