My solution for Atlan Backend Engineering Internship Task

To build the API, run docker-compose up --build inside atlan-challenge directory.

The solution uses worker threads and transactions. Every time a submit request is made, a new worker thread is created to run the query. The query will run in a transaction.

Once the query completes or fails, the query is cancelled and the worker thread dies.

The query runs inside a transaction to keep the update atomic.

• If user pauses the update, the connection in that new thread is paused.
• If the user then clicks on resume, that connection can now be resumed.
• If user cancels, the connection will be destroyed, and then worker thread will be killed. (In such a case, the transaction will anyway be rolled back by the nodejs-mysql module, so partial update isn't an issue here)
• If the user does not pause, cancel or resume the transaction, and the query successfully completes, it will be committed to the database.

So I have kept a SLEEP(1000) query before and after the UPDATE .. query in the transaction in the getDataForDateRange(..) function, to give time to test the pause, resume and cancel features.

• These features can be checked by observing that if pause/resume/cancel are pressed- even after more than 20 seconds (enough time for the entire transaction to have run), the logs would not show a successfully committed message.
• For resume, the transaction would continue, and the changes are successfully committed- unless it was a cancel operation, in which case it would have rolled back and not displayed successful commit message.

After creating the containers, use docker exec -it mysql-db-container mysql -u root -p and enter the password. Then run query SELECT * FROM sakila.customers in the sakila db to check the active field to check if the changes have been reflected in the database.

This was a fun challenge! Hope to hear back from you soon!