Unable to set partialFailureMode to best_effort on VirtualMCPServer

[robk1234]

Bug description

I attempted to set partialFailureMode on my VirtualMCPServer by adding the following to spec.config

    operational:
      logLevel: debug
      failureHandling:
        partialFailureMode: best_effort

The corresponding pod ends up in CrashLoopBackOff status. This message is in the logs:

{"time":"2026-04-15T17:17:05Z","level":"ERROR","msg":"Configuration validation failed: invalid configuration:\n - operational.failureHandling: partialFailureMode must be one of: fail, bestEffort"}
Error: validation failed: invalid configuration:

• operational.failureHandling: partialFailureMode must be one of: fail, bestEffort
  {"time":"2026-04-15T17:17:05Z","level":"ERROR","msg":"Error executing command: validation failed: invalid configuration:\n - operational.failureHandling: partialFailureMode must be one of: fail, bestEffort"}

bestEffort is not a valid value from the CRD: https://github.com/stacklok/toolhive/blob/main/deploy/charts/operator-crds/files/crds/toolhive.stacklok.dev_virtualmcpservers.yaml#L1328

I'm using this image:
ghcr.io/stacklok/toolhive/vmcp:v0.18.0

Steps to reproduce

Set partialFailureMode to best effort on VMCP Service

Expected behavior

The pod starts and just excludes the failed pods

Actual behavior

The pod fails to start

Environment (if relevant)

• OS/version:
• ToolHive version: 0.18.0

Additional context

Any additional information or logs you think might help.