Is fly fine if we are not chasing performance ?


I’m evaluating some PAAS, lately I’ve tested open source self hosted solutions like caprover/coolify and render/fly for hosted ones, I also consider testing DO’s one.

I wanted to deploy a django + Celery app and I have to say I had a better experience with fly than with render so far, I have had more errors with the later using the same DockerFile and same with auto python environment detection mode. “Performance” wise (for what I can tell, I’m not yet in production) with no traffic I couldn’t even run the django shell on the the 1cpu/1GB render box due to lack of memory, with fly I only had to upgrade from the 1 shared cpu 256mb to 512 mb to get into it.

For the rest it seems that Render has a more capable dashboard while fly gives more possibilities solutions via its cli.

However Fly is basing its marketing around being closer to the user while render and other PAAS around being Heroku’s successors and thus easy to use, I must admit I’m not really chasing performance for the moment and reliability/ease of use etc is what I aim for, yet I had a better experience with fly in that regard.

I really want to stick to a platform and gain experience with for futur deployments so is fly just going to focus on performance on the future or does the PASS DX matters too ?

Thanks for reading and congratulations on your product.

Thanks for the complient! In short, we’re focused on both performance and PaaS DX. We want devs who migrate to feel good about today, and to grow up with us when performance becomes more of a concern.

The nice thing now is you can just deploy in a single region - whichever is most useful to you.

Thanks sounds good, good continuation