Task Queue Issues: Redis with Replit

Hi,

I’m building an automated video generator app. Since generating videos can take time, I wanted to use a queuing mechanism. AI suggested Redis, I went for it. But I’m not experienced with Redis, especially in the Replit environment.

My questions are:
1- Is Redis free?
2- Is there a simpler alternative for the Replit Autoscale server? AI suggests using Replit database, is it good enough until [PUT A NUMBER HERE] amount of video generations in an hour/day?
3- If Redis is the only way to go, how can I activate (and keep it running) Redis and worker script in the production? Is there an easy way to do it via .replit configuration (tried .sh script etc. but couldn’t make it work)?

Thanks a lot in advance,
Mutlu

2 Likes

I believe so.


I would suggest just using Python, as it should be fast enough for your needs and has a wide variety of libraries. Replit DB would not be fast or large enough to store or process videos. I suggest Cloudinary, or as @boston2029 suggested, Firebase Cloud Storage for video storage.


I don’t have much experience in Redish, so I’m not sure. Again, I recommend Python.

1 Like

For storing videos, you might want to use Firebase Cloud Storage.

1 Like

Thanks. Would it be better than Replit Autoscale?

Replit Autoscale is for hosting your project. It’s not for any storage; in fact, it doesn’t even support persistent file systems, so you can’t use it for databases.

1 Like

Instead, you can use it to connect to databases to then store data in those databases. Autoscale is intended for web servers.

2 Likes

Thanks. Sorry for not providing all info; it’s a webservice, when it’s called it generates the video and returns the publicly accessible URL of it. Generating the video could take a few minutes, so queuing it makes more sense to me. But if you have good tips about generating the videos in a few seconds, please let me know, in that scenario I believe queing is not necessary.

You could try using python Multiproccesing to run multiple functions at once, utilizing more than one CPU core.

2 Likes