ServiceDiscover - No DNS records were found for service

### Is there an existing issue for this?

- [x] I have searched the existing issues

### Describe the bug

We have a local development with Aspire which works well with service discovery.
In production we have a k8s cluster with headless gRPC (code first) services and named "grpc" port with service discovery registered with `AddDnsSrvServiceEndpointProvider()`.
On k8s for unknow reasons even without changes we have a lot of error "No DNS records were found for service" (pods have no restarts or any sort of updates).
Normally it works well but each time the workload increase this error occur often and 5 / 10 seconds after the error the gRPC call returns to work, and then, the error occurr again.
I don't know if it is generate due to race condition, new httpclient instances creation, concurrency requests or something else.
Now I hard-code the url with service name and port, and the services return "ClusterIP" but if I can do some steps to find the "issue" I will try.

### Expected Behavior

_No response_

### Steps To Reproduce

_No response_

### Exceptions (if any)

_No response_

### .NET Version info

_No response_

### Anything else?

_No response_