3 ms·
We were using Thoonk. Jobs were originally pushed from our Django app, into a Celery job, which consumed the message, and published into Thoonk. Thoonk was the
by mattrobenolt 13y ago
We were using Thoonk.
Jobs were originally pushed from our Django app, into a Celery job, which consumed the message, and published into Thoonk. Thoonk was then consumed by the old Python backend.
With this rewrite, we've also removed Thoonk and are using a fanout exchange in Rabbit directly. Go consumes the Celery job from the Rabbit queue.
- lysium 13y agoSo, Thoonk could have been the bottleneck and it had nothing to do with Go? Just kidding.
- mattrobenolt 13y agoIt definitely wasn't helping. But it was not our bottleneck, I assure you.