When ServiceType is ClusterIP, ConnectionString should use "localhost" instead of an IP address

[alaye-ms]

[documentdb-triage-tool]

🤖 Auto-triaged by documentdb-triage-tool.

Applied: go, bug
Project fields suggested: Component api · Priority P2 · Effort S · Status In Progress
Confidence: 0.60 (llm)

Reasoning

LLM: Functional bug where ClusterIP service type generates an incorrect ConnectionString with an IP address instead of "localhost", which would break connectivity for users relying on that string.

If a label is wrong, remove it manually and ping @patty-chow so the rules can be tuned. The bot will not re-label items that already have component labels.

[WentingWu666666]

Hi @alaye-ms, I'm working on this issue and wanted to clarify the intended use case before proceeding.

The current behavior puts the ClusterIP address (e.g., 10.96.x.x) in status.connectionString. This IP is actually valid and routable for any pod running inside the cluster.

Using localhost would be correct for users connecting from outside the cluster via kubectl port-forward, but would be wrong for in-cluster applications that could use the ClusterIP directly.

Could you clarify:

1. Is the connectionString primarily intended as a user-facing convenience (for kubectl get documentdb output)?
2. Or could it also be consumed by in-cluster workloads?

This will help determine the right fix. Thanks!