chore(performance-metrics): update dev documentation

api/src/opentrons/util/performance_helpers.py

@@ -103,7 +103,7 @@ def _track_a_function(
     state_name: "RobotActivityState",
     func: _UnderlyingFunction[_UnderlyingFunctionParameters, _UnderlyingFunctionReturn],
 ) -> typing.Callable[_UnderlyingFunctionParameters, _UnderlyingFunctionReturn]:
-    """Track a function.
+    """Wrap a passed function with RobotActivityTracker.track.
 
     This function is a decorator that will track the given state for the
     decorated function.

performance-metrics/README.md

@@ -13,29 +13,7 @@ It is assumed that you already have the other projects in the monorepo setup cor
 make -C performance-metrics setup
 ```
 
-### Testing against OT-2 Dev Server
-
-```bash
-make -C robot-server dev-ot2
-```
-
-### Testing against real OT-2
-
-To push development packages to OT-2 run the following commands from the root directory of this repo:
-
-```bash
-make -C performance-metrics push-no-restart host=<ot2-ip>
-make -C api push-no-restart host=<ot2-ip>
-make -C robot-server push host=<ot2-ip>
-```
-
-### Testing against Flex Dev Server
-
-```bash
-make -C robot-server dev-flex
-```
-
-### Testing against real Flex
+### Pushing performance-metrics package to Flex
 
 ```bash
 make -C performance-metrics push-no-restart-ot3 host=<flex-ip>
@@ -69,3 +47,34 @@ To disable it run:
 ```bash
 make unset-performance-metrics-ff host=<ip>
 ```
+
+## Available features
+
+### Robot activity tracking
+
+#### Description
+
+Developers are able to track when the robot is in a block of code they choose to monitor. Looking at
+`api/src/opentrons/util/performance_helpers.py` you will see a class called `TrackingFunctions`. This class
+defines static methods which are decorators that can be used wrap arbitrary functions.
+
+As of 2024-07-31, the following tracking functions are available:
+
+- `track_analysis`
+- `track_getting_cached_protocol_analysis`
+
+Looking at `TrackingFunctions.track_analysis` we see that the underlying call to \_track_a_function specifies a string `"ANALYZING_PROTOCOL"`. Whenever a function that is wrapped with `TrackingFunctions.track_analysis` executes, the tracking function will label the underlying function as `"ANALYZING_PROTOCOL"`.
+
+To see where tracking function is used look at `robot_server/robot-server/protocols/protocol_analyzer.py`. You will see that the `ProtocolAnalyzer.analyze` function is wrapped with `TrackingFunctions.track_analysis`. Whenever `ProtocolAnalyzer.analyze` is called, the tracking function will start a timer. When the `ProtocolAnalyzer.analyze` function completes, the tracking function will stop the timer. It will then store the function start time and duration to the csv file, /data/performance_metrics_data/robot_activity_data
+
+#### Adding new tracking decorator
+
+To add a new tracking decorator, go to `performance-metrics/src/performance_metrics/_types.py`, and look at RobotActivityState literal and add a new state.
+Go to `api/src/opentrons/util/performance_helpers.py` and add a static method to the `TrackingFunctions` class that uses the new state.
+
+You can now wrap your functions with your new tracking decorator.
+
+### System resource tracking
+
+performance-metrics also exposes a tracking application called `SystemResourceTracker`. The application is implemented as a systemd service on the robot and records system resource usage by process. See the `oe-core` repo for more details.
+You can configure the system resource tracker by modifying the environment variables set for the service. The service file lives at `/lib/systemd/system/system-resource-tracker.service`. You can change the defined environment variables or remove them and define them in the robot's environment variables. See `performance-metrics/src/performance_metrics/system_resource_tracker/_config.py` to see what environment variables are available.

performance-metrics/src/performance_metrics/_data_shapes.py

@@ -34,10 +34,9 @@ class RawActivityData(CSVStorageBase):
     """Represents raw duration data with activity state information.
 
     Attributes:
-    - function_start_time (int): The start time of the function.
-    - duration_measurement_start_time (int): The start time for duration measurement.
-    - duration_measurement_end_time (int): The end time for duration measurement.
     - state (RobotActivityStates): The current state of the activity.
+    - func_start (int): The start time of the function.
+    - duration (int): The start time for duration measurement.
     """
 
     state: RobotActivityState