Add replay-safe logger support for orchestrations

.github/copilot-instructions.md

@@ -107,3 +107,18 @@ python -m pytest
 - `examples/` — example orchestrations (see `examples/README.md`)
 - `tests/` — test suite
 - `dev-requirements.txt` — development dependencies
+
+## Cross-Package Compatibility
+
+The `durabletask-azuremanaged` package extends the core `durabletask`
+package (e.g. `DurableTaskSchedulerWorker` subclasses
+`TaskHubGrpcWorker`). When adding or changing features in
+`durabletask/`, always verify that `durabletask-azuremanaged` still
+works correctly:
+
+- Check whether the azuremanaged worker, client, or tests override or
+  depend on the code you changed.
+- Run the azuremanaged unit tests if they exist for the affected area.
+- If a new public API is added to the core SDK (e.g. a method on
+  `OrchestrationContext`), confirm it is accessible through the
+  azuremanaged package and add a test or example if appropriate.

docs/features.md

@@ -150,6 +150,47 @@ Orchestrations can be suspended using the `suspend_orchestration` client API and
 
 Orchestrations can specify retry policies for activities and sub-orchestrations. These policies control how many times and how frequently an activity or sub-orchestration will be retried in the event of a transient error.
 
+### Replay-safe logging
+
+Orchestrator functions replay their history each time they are resumed,
+which can cause duplicate log messages. The `create_replay_safe_logger`
+method on `OrchestrationContext` returns a `ReplaySafeLogger` that wraps
+a standard `logging.Logger` and automatically suppresses output while
+the orchestrator is replaying. `ReplaySafeLogger` extends Python's
+`logging.LoggerAdapter`, which is the idiomatic way to add context or
+modify behavior on an existing logger.
+
+```python
+import logging
+
+logger = logging.getLogger("my_orchestrator")
+
+def my_orchestrator(ctx: task.OrchestrationContext, input):
+    replay_logger = ctx.create_replay_safe_logger(logger)
+    replay_logger.info("Starting orchestration %s", ctx.instance_id)
+    result = yield ctx.call_activity(my_activity, input=input)
+    replay_logger.info("Activity returned: %s", result)
+    return result
+```
+
+> [!NOTE]
+> Unlike the .NET SDK, where `CreateReplaySafeLogger` accepts a
+> category name string and internally creates the logger via
+> `ILoggerFactory`, the Python SDK requires you to pass an existing
+> `logging.Logger` instance. This is because Python's
+> `logging.getLogger(name)` already serves as the global factory and
+> is the standard way to obtain loggers.
+
+The replay-safe logger supports all standard log levels: `debug`,
+`info`, `warning`, `error`, `critical`, and `exception`, as well as
+the generic `log(level, msg)` method. It also exposes `isEnabledFor`
+which returns `False` during replay so callers can skip expensive
+message formatting.
+
+> [!TIP]
+> Create the replay-safe logger once at the start of your orchestrator
+> and reuse it throughout the function.
+
 ### Large payload externalization
 
 Orchestration inputs, outputs, and event data are transmitted through

durabletask/task.py

@@ -4,6 +4,7 @@
 # See https://peps.python.org/pep-0563/
 from __future__ import annotations
 
+import logging
 import math
 from abc import ABC, abstractmethod
 from datetime import datetime, timedelta, timezone
@@ -279,6 +280,51 @@ def new_uuid(self) -> str:
     def _exit_critical_section(self) -> None:
         pass
 
+    def create_replay_safe_logger(self, logger: logging.Logger) -> ReplaySafeLogger:
+        """Create a replay-safe logger that suppresses log messages during orchestration replay.
+
+        The returned logger wraps the provided logger and only emits log messages when
+        the orchestrator is not replaying. This prevents duplicate log messages from
+        appearing as a side effect of orchestration replay.
+
+        Parameters
+        ----------
+        logger : logging.Logger
+            The underlying logger to wrap.
+
+        Returns
+        -------
+        ReplaySafeLogger
+            A logger that only emits log messages when the orchestrator is not replaying.
+        """
+        return ReplaySafeLogger(logger, lambda: self.is_replaying)
+
+
+class ReplaySafeLogger(logging.LoggerAdapter):
+    """A logger adapter that suppresses log messages during orchestration replay.
+
+    This class extends :class:`logging.LoggerAdapter` and only emits log
+    messages when the orchestrator is *not* replaying. Use this to avoid
+    duplicate log entries that would otherwise appear every time the
+    orchestrator replays its history.
+
+    Obtain an instance by calling :meth:`OrchestrationContext.create_replay_safe_logger`.
+    """
+
+    def __init__(self, logger: logging.Logger, is_replaying: Callable[[], bool]) -> None:
+        super().__init__(logger, {})
+        self._is_replaying = is_replaying
+
+    def isEnabledFor(self, level: int) -> bool:
+        """Return whether logging is enabled for the given level.
+
+        Returns ``False`` while the orchestrator is replaying so that callers
+        can skip expensive message formatting during replay.
+        """
+        if self._is_replaying():
+            return False
+        return self.logger.isEnabledFor(level)
+
 
 class FailureDetails:
     def __init__(self, message: str, error_type: str, stack_trace: Optional[str]):

examples/activity_sequence.py

@@ -1,5 +1,6 @@
 """End-to-end sample that demonstrates how to configure an orchestrator
 that calls an activity function in a sequence and prints the outputs."""
+import logging
 import os
 
 from azure.identity import DefaultAzureCredential
@@ -8,6 +9,8 @@
 from durabletask.azuremanaged.client import DurableTaskSchedulerClient
 from durabletask.azuremanaged.worker import DurableTaskSchedulerWorker
 
+logger = logging.getLogger("activity_sequence")
+
 
 def hello(ctx: task.ActivityContext, name: str) -> str:
     """Activity function that returns a greeting"""
@@ -16,10 +19,15 @@ def hello(ctx: task.ActivityContext, name: str) -> str:
 
 def sequence(ctx: task.OrchestrationContext, _):
     """Orchestrator function that calls the 'hello' activity function in a sequence"""
+    # Create a replay-safe logger to avoid duplicate log messages during replay
+    replay_logger = ctx.create_replay_safe_logger(logger)
+
+    replay_logger.info("Starting activity sequence for instance %s", ctx.instance_id)
     # call "hello" activity function in a sequence
     result1 = yield ctx.call_activity(hello, input='Tokyo')
     result2 = yield ctx.call_activity(hello, input='Seattle')
     result3 = yield ctx.call_activity(hello, input='London')
+    replay_logger.info("All activities completed")
 
     # return an array of results
     return [result1, result2, result3]

tests/durabletask/test_orchestration_executor.py

@@ -1501,6 +1501,170 @@ def orchestrator(ctx: task.OrchestrationContext, _):
     assert complete_action.result.value == encoded_output
 
 
+def test_replay_safe_logger_suppresses_during_replay():
+    """Validates that the replay-safe logger suppresses log messages during replay."""
+    log_calls: list[str] = []
+
+    class _RecordingHandler(logging.Handler):
+        def emit(self, record: logging.LogRecord) -> None:
+            log_calls.append(record.getMessage())
+
+    inner_logger = logging.getLogger("test_replay_safe_logger")
+    inner_logger.setLevel(logging.DEBUG)
+    inner_logger.addHandler(_RecordingHandler())
+
+    activity_name = "say_hello"
+
+    def say_hello(_, name: str) -> str:
+        return f"Hello, {name}!"
+
+    def orchestrator(ctx: task.OrchestrationContext, _):
+        replay_logger = ctx.create_replay_safe_logger(inner_logger)
+        replay_logger.info("Starting orchestration")
+        result = yield ctx.call_activity(say_hello, input="World")
+        replay_logger.info("Activity completed: %s", result)
+        return result
+
+    registry = worker._Registry()
+    activity_name = registry.add_activity(say_hello)
+    orchestrator_name = registry.add_orchestrator(orchestrator)
+
+    # First execution: starts the orchestration. The orchestrator runs without
+    # replay, so both log calls should be emitted.
+    new_events = [
+        helpers.new_orchestrator_started_event(datetime.now()),
+        helpers.new_execution_started_event(orchestrator_name, TEST_INSTANCE_ID, encoded_input=None),
+    ]
+    executor = worker._OrchestrationExecutor(registry, TEST_LOGGER)
+    result = executor.execute(TEST_INSTANCE_ID, [], new_events)
+    assert result.actions  # should have scheduled the activity
+
+    assert log_calls == ["Starting orchestration"]
+    log_calls.clear()
+
+    # Second execution: the orchestrator replays from history and then processes the
+    # activity completion. The "Starting orchestration" message is emitted during
+    # replay and should be suppressed; "Activity completed" is emitted after replay
+    # ends and should appear exactly once.
+    old_events = new_events + [
+        helpers.new_task_scheduled_event(1, activity_name),
+    ]
+    encoded_output = json.dumps(say_hello(None, "World"))
+    new_events = [helpers.new_task_completed_event(1, encoded_output)]
+    executor = worker._OrchestrationExecutor(registry, TEST_LOGGER)
+    result = executor.execute(TEST_INSTANCE_ID, old_events, new_events)
+    complete_action = get_and_validate_complete_orchestration_action_list(1, result.actions)
+    assert complete_action.orchestrationStatus == pb.ORCHESTRATION_STATUS_COMPLETED
+
+    assert log_calls == ["Activity completed: Hello, World!"]
+
+
+def test_replay_safe_logger_all_levels():
+    """Validates that all log levels are suppressed during replay and emitted otherwise."""
+    log_levels: list[str] = []
+
+    class _LevelRecorder(logging.Handler):
+        def emit(self, record: logging.LogRecord) -> None:
+            log_levels.append(record.levelname)
+
+    inner_logger = logging.getLogger("test_replay_safe_logger_levels")
+    inner_logger.setLevel(logging.DEBUG)
+    inner_logger.addHandler(_LevelRecorder())
+
+    def orchestrator(ctx: task.OrchestrationContext, _):
+        replay_logger = ctx.create_replay_safe_logger(inner_logger)
+        replay_logger.debug("debug msg")
+        replay_logger.info("info msg")
+        replay_logger.warning("warning msg")
+        replay_logger.error("error msg")
+        replay_logger.critical("critical msg")
+        return "done"
+
+    registry = worker._Registry()
+    orchestrator_name = registry.add_orchestrator(orchestrator)
+
+    new_events = [
+        helpers.new_orchestrator_started_event(datetime.now()),
+        helpers.new_execution_started_event(orchestrator_name, TEST_INSTANCE_ID, encoded_input=None),
+    ]
+    executor = worker._OrchestrationExecutor(registry, TEST_LOGGER)
+    result = executor.execute(TEST_INSTANCE_ID, [], new_events)
+    complete_action = get_and_validate_complete_orchestration_action_list(1, result.actions)
+    assert complete_action.orchestrationStatus == pb.ORCHESTRATION_STATUS_COMPLETED
+
+    assert log_levels == ["DEBUG", "INFO", "WARNING", "ERROR", "CRITICAL"]
+
+
+def test_replay_safe_logger_direct():
+    """Unit test for ReplaySafeLogger — verifies suppression based on is_replaying flag."""
+    log_calls: list[str] = []
+
+    class _RecordingHandler(logging.Handler):
+        def emit(self, record: logging.LogRecord) -> None:
+            log_calls.append(record.getMessage())
+
+    inner_logger = logging.getLogger("test_replay_safe_logger_direct")
+    inner_logger.setLevel(logging.DEBUG)
+    inner_logger.addHandler(_RecordingHandler())
+
+    replaying = True
+    replay_logger = task.ReplaySafeLogger(inner_logger, lambda: replaying)
+
+    replay_logger.info("should be suppressed")
+    assert log_calls == []
+
+    replaying = False
+    replay_logger.info("should appear")
+    assert log_calls == ["should appear"]
+
+
+def test_replay_safe_logger_log_method():
+    """Validates the generic log() method respects the replay flag."""
+    log_calls: list[str] = []
+
+    class _RecordingHandler(logging.Handler):
+        def emit(self, record: logging.LogRecord) -> None:
+            log_calls.append(record.getMessage())
+
+    inner_logger = logging.getLogger("test_replay_safe_logger_log_method")
+    inner_logger.setLevel(logging.DEBUG)
+    inner_logger.addHandler(_RecordingHandler())
+
+    replaying = True
+    replay_logger = task.ReplaySafeLogger(inner_logger, lambda: replaying)
+
+    replay_logger.log(logging.WARNING, "suppressed warning")
+    assert log_calls == []
+
+    replaying = False
+    replay_logger.log(logging.WARNING, "visible warning")
+    assert log_calls == ["visible warning"]
+
+
+def test_replay_safe_logger_is_enabled_for():
+    """Validates isEnabledFor returns False during replay."""
+    inner_logger = logging.getLogger("test_replay_safe_logger_enabled")
+    inner_logger.setLevel(logging.DEBUG)
+
+    replaying = True
+    replay_logger = task.ReplaySafeLogger(inner_logger, lambda: replaying)
+
+    # During replay, isEnabledFor should always return False
+    assert replay_logger.isEnabledFor(logging.DEBUG) is False
+    assert replay_logger.isEnabledFor(logging.INFO) is False
+    assert replay_logger.isEnabledFor(logging.CRITICAL) is False
+
+    # After replay, delegates to the inner logger
+    replaying = False
+    assert replay_logger.isEnabledFor(logging.DEBUG) is True
+    assert replay_logger.isEnabledFor(logging.INFO) is True
+
+    # If a level is below the inner logger's level, should return False
+    inner_logger.setLevel(logging.WARNING)
+    assert replay_logger.isEnabledFor(logging.DEBUG) is False
+    assert replay_logger.isEnabledFor(logging.WARNING) is True
+
+
 def test_when_any_with_retry():
     """Tests that a when_any pattern works correctly with retries"""
     def dummy_activity(_, inp: str):