Add retry logging for replication task send attempts

[jiechenz]

What changed?

Added throttled warn logging inside the replication task send retry loop in stream_sender.go. Error messages are now prefixed with the step that failed (convert:, rate_limit:, send:).

Sample log output:

  {
    "level": "warn",
    "msg": "Replication task send retry",
    "wf-namespace-id": "abc123",
    "wf-id": "my-workflow",
    "task-id": 54321,
    "counter": 2,
    "error": "convert: context deadline exceeded"
  }

Why?

The replication_task_send_attempt inner retry loop was completely silent — no logs. The only logging happened after all retries were exhausted, making it impossible to diagnose transient retry.

How did you test it?

• built
• run locally and tested manually
• covered by existing tests
• added new unit test(s)
• added new functional test(s)

Potential risks

n/a