I saw some folks posting that they were doing Lemmy instances with cheap Vultr instances. Are you using something similar? And how’s the bandwidth going with peering to other nodes? I’ve toyed around with the idea of starting my own node.
yes, I’m using a Vultr instance. I was on the lowest tier of the High Frequency plan until last night. I upgraded to 3vCPUs and it’s going a lot better. Biggest problem I think was that lots of other communities are federating and so we are up to 104 GB of outbound traffic in only like 5 days. We were hitting swap which I think was causing lots of turmoil on writes. Only 17gb of inbound traffic so far. You could easily start your own node on the lowest instance, if no one else is federating your communities. As soon as big servers start federating is where we started to see trouble.
Good to know. You make a good point about the direction of traffic: folks will definitely pull more from here, than folks here requesting from other servers. And TIL inbound traffic is free on Vultr, so you don’t have to worry about that cost at least.
Was playing around with a small in memory cache as well as materialized views to prevent the swap hits. Hard to prevent the inbound traffic though, maybe a CDN could help, but need to see what the traffic patterns look like.
I saw some folks posting that they were doing Lemmy instances with cheap Vultr instances. Are you using something similar? And how’s the bandwidth going with peering to other nodes? I’ve toyed around with the idea of starting my own node.
yes, I’m using a Vultr instance. I was on the lowest tier of the High Frequency plan until last night. I upgraded to 3vCPUs and it’s going a lot better. Biggest problem I think was that lots of other communities are federating and so we are up to 104 GB of outbound traffic in only like 5 days. We were hitting swap which I think was causing lots of turmoil on writes. Only 17gb of inbound traffic so far. You could easily start your own node on the lowest instance, if no one else is federating your communities. As soon as big servers start federating is where we started to see trouble.
Good to know. You make a good point about the direction of traffic: folks will definitely pull more from here, than folks here requesting from other servers. And TIL inbound traffic is free on Vultr, so you don’t have to worry about that cost at least.
Was playing around with a small in memory cache as well as materialized views to prevent the swap hits. Hard to prevent the inbound traffic though, maybe a CDN could help, but need to see what the traffic patterns look like.