arq VS celery

I need to process a lot of long running IO heavy jobs with background workers. I've been using ARQ for a while but decided to take a crack at writing my own distributed queue.

Lately I've been writing asynchronous python code and yes, the resource integration problem has come again. Because now from version 1.4 SQLAlchemy has become asynchronous a new boilerplate was created. Now, as a scheduler, I took a completely asynchronous Arq. Considering the specifics of the service, long I/O operations, it seems that the service turned out to be more optimal in asynchronous execution. I haven't measured the performance yet, but I think I'll write another post about it.

So this might be my understanding, but stuff like celery is more like an orchestrator that chunks up workloads (think Hadoop with multiple nodes).

In the Django world, you'd probably rely on Celery to do this for you. You're probably looking for something similar that works with Go. https://github.com/celery/celery

I've been doing some research and there doesn't seem to be much information on this issue, aditionally there's this but without a fix yet or any workaround: https://github.com/celery/celery/issues/7792

RabbitMQ Classic mirror queues are very fragile to network partitioning. They are deprecated in favor of Quorum queues, but Celery doesn't support them yet : https://github.com/celery/celery/issues/6067

The Celery package provides some number of (undocumented!) result backends to store task results in different local, network, and cloud storages. The django-celery-result package adds options to use Django-specific ORM-based result storage, as well as Django-specific cache subsystem.

I spent 3 years building a high scale crawler on top of Celery.

I can't recommend it. We found many bugs in the more advanced features of Celery (like Canvas) we also ran into some really weird issues like tasks getting duplicated for no reason [1].

The most concerning problem is that the project was abandoned. The original creator is not working on it anymore and all issues that we raised were ignored. We had to fork the project and apply our own fixes to it. This was 4 years ago so maybe things improved since them.

Celery is also extremely complex.

I would recommend https://dramatiq.io/ instead.

[1]: https://github.com/celery/celery/issues/4426

You can use celery to process the file for extraction, saving and creating rar/zip.

Celery not usefull for long term future tasks (far future) · Issue #4522 · celery/celery (github.com)