The server was entirely reachable, throughout the outage every 1 minute I could refresh the page and get a response. What fell over was the NodeJS and/or Rust code, nginx was failing to get lemmy-ui and lemmy-server to respond to API calls. They didn’t deny service to the network, but to overload resources on the server. Could be Rust trying to do HTTP connections for peer servers, could be PostgreSQL, lots of opportunity with both. I hope we get details.
Isn’t the domain protected by Cloudflare? Is it linked to the downtime of lemmy.world too?
The server was entirely reachable, throughout the outage every 1 minute I could refresh the page and get a response. What fell over was the NodeJS and/or Rust code, nginx was failing to get lemmy-ui and lemmy-server to respond to API calls. They didn’t deny service to the network, but to overload resources on the server. Could be Rust trying to do HTTP connections for peer servers, could be PostgreSQL, lots of opportunity with both. I hope we get details.