MTA Tracker

Introduction

MTA Tracker is a small web service to monitor NYC MTA subway service. The web service provides two endpoints /status and /uptime that can be used to check delays and uptime for a given subway line.

Given Requirements

• Continuously monitors the status of MTA service to see whether a line is delayed or not.
  • When a line transitions from not delayed → delayed, you should print the following message to console or to a logfile: “Line <line_name> is experiencing delays”.
  • Similarly, when a line transitions from delayed → not delayed, you should print the following message to console or to a logfile: “Line <line_name> is now recovered”.
• Exposes an endpoint called /status, which takes the name of a particular line as an argument and returns whether or not the line is currently delayed.
• Exposes an endpoint called /uptime, which also takes the name of a particular line as an argument and returns the fraction of time that it has not been delayed since inception.
  • More concretely, “uptime” is defined as 1 - (total_time_delayed / total_time)

Getting Started

1. clone this repo locally
2. cd into project dir with Dockerfile and run... (Note: celery worker is started right away)

docker-compose up --build

3. In another terminal window, use cli cmd to create tables in postgres db

docker-compose exec api mta_tracker db reset

4. (Optional) check to see tables have been created in postgres db

docker-compose exec postgres psql -U mta_tracker -d mta_tracker

6. Ping the status endpoint for a subway line's status

http://localhost:8000/mta_tracker/api/v1.0/subway/lines/status?line=L

7. Ping the uptime endpoint for a subway line's uptime

http://localhost:8000/mta_tracker/api/v1.0/subway/lines/uptime?line=L

8. Check logs for messages related to subway line status changes

docker-compose exec api cat logs/line_status.log

Testing

• Run the following command to test endpoints

docker-compose exec api py.test mta_tracker/tests

Clean up

docker-compose stop

docker-compose rm -f

docker rmi -f $(docker images -qf dangling=true)

Implementation details

• A celery worker runs the task in mta_tracker/tasks/tasks.py every minute. All details regarding the continuous monitoring of MTA service status can be found in that file, including logging and where data is stored in the db.
• To understand endpoints, see mta_tracker/resources/subway/lines.py.

Known Issues

• Celery config should be stored in same file as flask configs...this way instance folder can be used for prod configs.
• Times are all in UTC