That‘s what I had in mind. To run nginx on a seperate vps, so I can scale it easier. Run fediverse instances in the back, either all on one vps or on different vps. This way I could provide a hub while increase performance (due to compression and caching) and provide redundancy/load balancing if necessary.
What‘s the typical traffic you experience? Peak (Gbit/s) and average/daily traffic (GB)
Yeah I saw that. I‘m a big fan of minimalistic, yet super performant architectures and I‘m just trying to get a feeling on how I could solve this problem. I try to avoid any downtime, whenever possible
So I‘m currently on planning to host an instance myself. This graph helped me quite a lot to get an idea what system resources are required.
Do you use any reverse proxy in front of it?
Nginx runs on the server , proxying to the lemmy docker containers
That‘s what I had in mind. To run nginx on a seperate vps, so I can scale it easier. Run fediverse instances in the back, either all on one vps or on different vps. This way I could provide a hub while increase performance (due to compression and caching) and provide redundancy/load balancing if necessary.
What‘s the typical traffic you experience? Peak (Gbit/s) and average/daily traffic (GB)
Thanks, that’s super helpfu!
Lemmy world has a lot of users. So your instance initially will require a lot less resources ✌️
Yeah I saw that. I‘m a big fan of minimalistic, yet super performant architectures and I‘m just trying to get a feeling on how I could solve this problem. I try to avoid any downtime, whenever possible