[Housekeeping] Use simplified grpc probes in helm charts

[Sovietaced]

Describe the issue

Current grpc probes are defined using a curl script but could be simplified to: https://kubernetes.io/docs/tasks/configure-pod-container/configure-liveness-readiness-startup-probes/#define-a-grpc-liveness-probe

What if we do not do this?

Less elegant code

Related component(s)

No response

Are you sure this issue hasn't been raised already?

• [x] Yes

Have you read the Code of Conduct?

• [x] Yes

[runllm[bot]]

Flyte's current Helm charts define gRPC probes for flyteadmin using a shell script that combines an HTTP health check (via curl) and a gRPC health probe, rather than using Kubernetes' native gRPC probe syntax. This approach is visible in the default liveness and readiness probe configuration in the Helm values file and deployment manifests, where the probe is an exec command running both checks in sequence, not a native gRPC probe block as described in the Kubernetes documentation (flyte-core/values.yaml, flyte_helm_generated.yaml, Kubernetes docs).

There is currently no evidence in the provided data sources that Flyte's Helm charts support the simplified, native gRPC probe syntax (i.e., using grpc: blocks in the probe definition). The current method is more complex and less idiomatic than the native Kubernetes gRPC probe. Would you like more detail on how to update the Helm chart to use the native gRPC probe syntax?

Sources:

• flyte_helm_generated.yaml
• Kubernetes: Define a gRPC liveness probe
• flyte-core/values.yaml

---

This is an AI-generated response and your feedback is appreciated! Please leave a 👍 if this is helpful and 👎 if it is not.

Mention @runllm to ask follow-up questions. See runllm.com for more!