Add PDB and probe support to MCP Helm chart

[kkhristenko51]

Add PDB and startup/readiness/liveness probes support to MCP Helm chart

This PR extends the rosetta-mcp-server Helm chart to support configurable Kubernetes health probes on the Deployment and an optional PodDisruptionBudget (PDB), and bumps the chart version to reflect the new functionality.

Changes:

• Add optional livenessProbe, readinessProbe, and startupProbe values and render them into the Deployment when configured.
• Add an optional PodDisruptionBudget template controlled by podDisruptionBudget.* values (and gated by replica/minReplica count).
• Bump chart version from 0.2.0 to 0.3.0.

[Copilot]

Pull request overview

This PR extends the rosetta-mcp-server Helm chart to support configurable Kubernetes health probes on the Deployment and an optional PodDisruptionBudget (PDB), and bumps the chart version to reflect the new functionality.

Changes:

• Add optional livenessProbe, readinessProbe, and startupProbe values and render them into the Deployment when configured.
• Add an optional PodDisruptionBudget template controlled by podDisruptionBudget.* values (and gated by replica/minReplica count).
• Bump chart version from 0.2.0 to 0.3.0.

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 2 comments.

File	Description
helm-charts/rosetta-mcp-server/values.yaml	Adds values for health probes and PDB configuration.
helm-charts/rosetta-mcp-server/templates/deployment.yaml	Renders optional container probes from values.
helm-charts/rosetta-mcp-server/templates/poddisruptionbudget.yaml	Adds a PDB resource template controlled by chart values.
helm-charts/rosetta-mcp-server/Chart.yaml	Version bump to 0.3.0.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[github-actions]

Rosetta Triage Review

Summary: This PR extends the rosetta-mcp-server Helm chart to support configurable Kubernetes health probes (liveness, readiness, startup) and an optional PodDisruptionBudget (PDB), then bumps the chart version from 0.2.0 to 0.3.0.

Findings:

• All three probe types (livenessProbe, readinessProbe, startupProbe) are rendered using idiomatic Helm {{- with }} blocks — clean and correct pattern for optional fields.
• The PDB template correctly gates creation on podDisruptionBudget.enabled AND replica count > 1, with a {{- fail }} guard that prevents both minAvailable and maxUnavailable being set simultaneously — good defensive design.
• Edge case to verify: If a user overrides both minAvailable and maxUnavailable to null/unset in their values override, the PDB spec would render with neither field, producing an invalid Kubernetes resource. The defaults in values.yaml protect against this (maxUnavailable: 1 is set), but it is worth noting for documentation or an explicit validation.
• values.yaml provides helpful commented-out examples for all probe types and the PDB — good discoverability.
• Version bump from 0.2.0 to 0.3.0 is appropriate for a minor feature addition.
• No Helm unit tests (e.g., helm-unittest) are included for the new templates. Existing chart test coverage level is unknown, but adding tests for the new PDB template and probe injection would strengthen confidence.

Suggestions:

• Consider adding a helm-unittest test case to verify the PDB renders when enabled: true and replicaCount: 2, and that it does NOT render when replicaCount: 1.
• Consider documenting the edge case where neither minAvailable nor maxUnavailable is set (after user overrides) to make the behavior explicit in the values comments.

Automated triage by Rosetta agent