exam moderation service cron

[ShaunSHamilton]

Exam Moderation Service

Dockerfile: https://github.com/freeCodeCamp/exam-services/blob/main/moderation-service/Dockerfile
Compose File: https://github.com/freeCodeCamp/exam-services/blob/main/docker-compose.yml

Currently, this is set to build the image on GHA, and push to DOCR.

Ideally, the container will be run on cron.

Eventually, I expect it to run once per day, but the timing might end up being completely changed depending on when we decide attempts should be awarded etc.

[raisedadead]

Questions:

• Is this a singleton service, that is if one instance is not finished executing, another instance is triggered should the new one fail?
• What is the minimum duration a job will be alive?
• What is the retry logic? That is how does the service tell the orchestrator its done its job? Exit codes? Log messages?

[ShaunSHamilton]

Is this a singleton service, that is if one instance is not finished executing, another instance is triggered should the new one fail?

Yes, this is a singleton service. Nothing catastrophic happens if two of these run at the same time, but duplicate records are created.

What is the minimum duration a job will be alive?

(time to connect to Atlas) + (time to query moderation records) + (time to query exams in db) + (time to query attempts not moderated) + (time to write to moderation collection)

If that is ever greater than 5 seconds, I would be worried. But, minimum is difficult to say. I will run some tests and report back.

What is the retry logic?

No retry logic. If there are errors, error is sent to Sentry, and the process exists with 0.

That is how does the service tell the orchestrator its done its job? Exit codes?

Yes. The exit code is always 0. Errors are sent to Sentry.