Don't inspect request headers in cron jobs

[Vinnl]

References:

Jira: MNTOR-5071
Figma:

Description

I could not reproduce the exact issue that we're seeing in our logs, but I could trigger the code path that imports next/headers locally, and that shouldn't be possible.

Possibly the error in the logs are because next isn't installed for some reason, or we're using a version of Node that doesn't support importing from next/headers rather than just next?

How to test

Follow the readme instructions on pubsub, but for the curl request, find a user's primary_sha1, take the first six characters and use that as the hashPrefix, and the rest as a hashSuffix.

Also, after you do this once, be sure to DELETE FROM "email_notifications"; if you want to try again.

Checklist (Definition of Done)

• Localization strings (if needed) have been added.
• Commits in this PR are minimal and have descriptive commit messages.
• I've added or updated the relevant sections in readme and/or code comments
• I've added a unit test to test for potential regressions of this bug.
• If this PR implements a feature flag or experimentation, I've checked that it still works with the flag both on, and with the flag off.
• If this PR implements a feature flag or experimentation, the Ship Behind Feature Flag status in Jira has been set
• Product Owner accepted the User Story (demo of functionality completed) or waived the privilege.
• All acceptance criteria are met.
• Jira ticket has been updated (if needed) to match changes made during the development process.
• Jira ticket has been updated (if needed) with suggestions for QA when this PR is deployed to stage.

[kschelonka]

You correctly identified here that the default value of E2E_TEST_ENV=local was leaking into prod (it wasn't overridden by prod config). I'm concerned about how much the runtime and config is entangled with the business logic, and I know it's caused incidents in the past. I think rather than adding more, we should try to address the issue of unclear domain separation/boundaries and hidden dependencies.

[Vinnl]

@kschelonka As you know, I agree :) That said, I think with that in mind too it can be helpful to merge this - when we get around to untangling it, the checks for NEXT_RUNTIME will help us identify which code paths go into the Next.js codebase and can be left out of the cron job/PubSub codebase. WDYT?

[kschelonka]

Reading up on this so I can give a more educated reply

[kschelonka]

Ok, I've done some more reading and thinking. I'm cautious with "magical" stuff like next because I want to make sure I understand what's going on under the hood...

getExperiments as written before had a hidden dependency on the next environment due to importing and invoking loadNextHeaders. I'm resistant to adding more layers of state-dependent logic, when the issue could be solved more simply by accepting actual headers in getExperiments function args (or using DI similarly)