5491 - Admin links job on separate Heroku dyno - Add worker trigger #5618
Conversation
| if (ProcessEnv.nodeEnv === NodeEnv.development) { | ||
| void import('./worker/tasks/verifyLinks/triggerVerifyLinksWorker') | ||
| .then(({ triggerVerifyLinksWorker }) => triggerVerifyLinksWorker()) | ||
| .catch((error) => { | ||
| Logger.error(`[verifyLinks] failed to start local worker: ${JSON.stringify(error)}`) | ||
| }) | ||
| } |
There was a problem hiding this comment.
Starting the verify-links worker automatically so local dev env doesn’t get stuck in a stale state: the server restarts, the worker dies, but Redis still says a worker is “active”. When that happens, new verify requests won’t spin up a worker and the queue just sits there until you manually restart the worker or clear Redis.
We might need something similar for prod, we could use a watchdog mechanism that checks in an interval if there are queued jobs but no dyno worker. Though could be good to test first as it is to see if that's necessary at all.
There was a problem hiding this comment.
why do we need this in server start ?
There was a problem hiding this comment.
In dev we don’t have the same Heroku lifecycle (enqueue job → start dedicated dyno → dyno consumes queue), we run the worker in the same main process. So I'm starting it here so it cleans up any stale jobs that might happen because of server restarts during development.
Should we add a separate start.dev.ts maybe? what do you think?
| @@ -0,0 +1,25 @@ | |||
| import { LinkToVisit } from 'meta/cycleData/links/link' | |||
There was a problem hiding this comment.
Moved this file from src/server/controller/cycleData/links/visitCycleLinks/utils/mergeLinks.ts. No logic changes.
TODO: