A microservice to receive tracks and feedback from users.
The easiest way to run the tracking service is to use the contained docker-compose:
docker-compose up
Via the following build arguments, the build can be configured:
WORKER_MODE: If in worker mode or manager mode.
WORKER_HOST: The host of the worker.SYNC_KEY: The key used when syncing.
MANAGER_HOST: The host of the manager.SYNC_KEY: The key used when syncing.
This service can run in two modes: manager and worker. The worker mode is designed to face user traffic and can be scaled horizontally.
These are the exact tasks of each role:
- The worker receives and checks tracks (can be scaled). Data stored in the worker DB can be lost at any time.
- The manager fetches tracks from the worker, tells them to flush their DB once new tracks are finished, and exposes a download API for all tracks. This service should be persisted.
See docker-compose.yml for an example setup.
Get the contents of a specific track. This request is performed against the manager.
curl "http://localhost:8000/tracks/fetch/?pk=\[012345\]&key=secret"
Example response:
{"metadata": {"startTime": 0, "endTime": 1, "debug": true, "backend": "staging", "positioningMode": "gnss", "userId": "[012345]", "sessionId": "[012345]", "deviceType": "iPhone 11,2", "deviceHeight": 812, "deviceWidth": 375, "appVersion": "1.0.0", "buildNumber": "1234", "statusSummary": {}, "taps": [], "selectedWaypoints": [], "routes": [], "predictions": []}, "gpsCSV": "timestamp,longitude,latitude,speed,accuracy\n0,0,0,0,0", "accelerometerCSV": "timestamp,x,y,z\n0,0,0,0", "gyroscopeCSV": "timestamp,x,y,z\n0,0,0,0", "magnetometerCSV": "timestamp,x,y,z\n0,0,0,0", "pk": 9}
If anything else happens, the server will respond with a response code other than 200.
Parameters:
key- The API key to use.pk- The primary key of the track to get.
Get a list of tracks. This request is performed against the manager.
curl "http://localhost:8000/tracks/list/?key=secret"
Response:
[
{
"results": [
{
"pk": track.pk,
"startTime": track.start_time,
"endTime": track.end_time,
"debug": track.debug,
"backend": track.backend,
"positioningMode": track.positioning_mode,
"deviceType": track.device_type,
"userId": track.user_id,
},
...
],
"page": page,
"pageSize": page_size,
"totalPages": tracks.paginator.num_pages,
},
...
]
If anything else happens, the server will respond with a response code other than 200.
Parameters:
key- The API key to use.from- The start date to get tracks from. Format: unix milliseconds. Default:None(Include all).to- The end date to get tracks to. Format: unix milliseconds. Default:None(Include all).debug- If set tofalse, the server won't return debug tracks. Default:None(Include all).backend- The kind of backend to look for. Default:None(Include all).positioning- The kind of positioning to look for. Default:None(Include all).deviceType- The kind of device type to look for. Default:None(Include all).userId- The kind of user ID to look for. Default:None(Include all).page- The page to get. Default:1.pageSize- The page size to get. Default:10. Limited to100.
Perform an example multipart request:
curl -X POST \
-F gps.csv.gz='@example-gps.csv.gz' \
-F accelerometer.csv.gz='@example-accelerometer.csv.gz' \
-F magnetometer.csv.gz='@example-magnetometer.csv.gz' \
-F gyroscope.csv.gz='@example-gyroscope.csv.gz' \
-F metadata.json.gz='@example-metadata.json.gz' \
http://localhost/tracks/post/
Response:
{
"success": true
}
If anything else happens, the server will respond with a response code other than 200.
Publish a new answer. The body of the POST request should contain at least the following information:
{
"userId": <The id of the user. Max length: 100>,
"questionText": <The text of the question. Max length: 300>
}
However, there are more fields that can be passed:
{
"userId": <The id of the user. Max length: 100>,
"questionText": <The text of the question. Max length: 300>,
"questionImage": <The base 64 encoded image of this question, if provided. Max length: 10MB>,
"sessionId": <The id of the session, if provided. Max length: 100>,
"value": <The value of the answer, if provided. Max length: 1000>
}
Perform an example POST request:
curl --data "@example.json" http://localhost/answers/post/
Response:
{
"success": true
}
If anything else happens, the server will respond with a response code other than 200.
Prometheus metrics are generated on every sync call if the track count has changed.
Metric generation can be tested by adding tracks. Therefore different example-metadata-x.json can be used.
The metrics.txt is stored under /backend/data/ and can be requested with:
curl "http://localhost:8000/monitoring/metrics?api_key=secret"
The metric only contains tracks, that did not charge the battery during tracks.
The Track attribute has_battery_data is also false for tracks that did charge during the track.
We highly encourage you to open an issue or a pull request. You can also use our repository freely with the MIT license.
Every service runs through testing before it is deployed in our release setup. Read more in our PrioBike deployment readme to understand how specific branches/tags are deployed.
Help us improve this documentation. If you have any problems or unclarities, feel free to open an issue.