geomaMay 23

System Resource (RAM and CPU) usage for self hosting fediverse social platforms

https://lemmy.ml/post/47746527

System Resource (RAM and CPU) usage for self hosting fediverse social platforms - Lemmy

I want to host a fediverse instance but I need it to be light (limited system resources). Mastodon seems to be kinda heavy. GotoSocial is supposedly very light? (but limited on functionality?) is Pixelfed much heavier? Bonfire is built on elixir so should also be light? Akkoma? Pleroma? Also, I need it to be engaging to people migrating mostly from IG, so I do need some bells/whistles I guess thanks!

Show threadabeorch

@geoma I'm really interested in what you are doing. Are you aware of coopcloud.tech ? this is designed to make deloyment of apps across multiple servers. My thought was that perhaps you could go wider but deeper and offer to setup a series of Yunhost instances - and then federating them (Nextcloud, ActivityPub etc) for each organisation?

You could then use more basic tech at each location / for each organisation and get the benefit of them providing backup to each other.

Id actually been looking at doing this for sailingclubs - They need Web/Customer Relationship -Membership Management as well social media integration.

If you are engaging with people who are currently on Insta etc. you may also need to look at something like #Postiz postiz.com/ - to message out to FB , Insta etc and include links back to ActivityPub/Websites to encourage people to migrate.

Postiz: The All-in-One agentic social media scheduling tool

Streamline your social media with Postiz. Schedule posts, analyze performance, and manage all accounts in one place.

Postiz
May 26 at 3:33pmWeb
Show threadgeomaJun 10

So coopcloud.tech is like a yunohost of sorts but made for deploying on multiple servers so you have mirrors for duplication/high availability?

I’ll check out postiz. thanks!

Show threadabeorchJun 13
@geoma I think its more that you can run it across multiple compute instances - and for multiple organisations. So its supporting the case where your compute requirements have grown beyond one vm (or you want redundancy ) and where you have multiple organisations that you want to support multiple independent instances of the same (or different) suite of applications.