this post was submitted on 12 Jul 2023
7 points (88.9% liked)

Lemmy Support

4651 readers
1 users here now

Support / questions about Lemmy.

Matrix Space: #lemmy-space

founded 5 years ago
MODERATORS
 

I'm trying to run a federated lemmy instance using Ansible but I'm not sure what the minimal ram and storage requirements are. Let's say I want to support upto 100 active users. I prefer not to resort to trial and error.

top 3 comments
sorted by: hot top controversial new old
[–] [email protected] 4 points 1 year ago

Might find this post from lemmy.world interesting ^^ https://lemmy.world/post/1073599

[–] [email protected] 4 points 1 year ago

you know you can easily scale up and down instance sizes on aws, or move your instance to, say oracle. I've done both of these. The Ansible installer makes it very easy to move. you just down your containers and copy over the volumes directory for the move.

With these options you can start small and free. It may take a while to grow to 100 users. A 2 vcpu 2GB ram t4g.small on aws seems more than adequate and there's a free trial through the end of the year. It's arm64 though so small changes to the ansible lemmy.yml. The free tier t2.micro at 1 cpu 1 GB ram seemed too small.

Oracle has an always free arm64 image with 4 cpu and 24 GB ram plus a generous boot disk and 200GB block storage so you might as well start there. It will probably last quite a while. Maybe all the way through 100 users. The only thing is while I know it's very easy to change your instance type on AWS, I haven't looked into it on Oracle. But again moving your instance is very easy.

[–] [email protected] 3 points 1 year ago* (last edited 1 year ago)

Storage on the vm won't be too much of an issue, as long as you make sure to use Object Storage (s3) for pict-rs from the start. For Lemmit (just a few users, but hundreds of communities and over 150k posts) I'm doing fine with just 2 GB of memory, 1 vcpu, and 2 GB of disk storage for postgres. The storage bucket is sitting at 36 GB.

You might want to scale up cpu and memory for more users as you grow, but you'd be surprise with how little resources you can get away.

load more comments
view more: next ›