We need more 10Gbps or higher dedicated servers for hosting our OS and app updates. We have North America covered well enough via sponsored servers from ReliableSite in both Miami and Los Angeles and a sponsored server from Xenyth in Toronto but no longer have any left in Europe.
We can afford to pay for a couple 10Gbps servers in Europe if needed but it would be better if we could get them sponsored. For Europe, we use an average of 1Gbps to 2Gbps over the course of a month but it's very bursty so we need 10Gbps to 20Gbps total capacity for it in Europe.
@GrapheneOS do you need actual servers (with compute) or just a place to distribute traffic (so a CDN)? If the latter, @fastlydevs is usually happy to sponsor FOSS projects and has good connectivity across the globe.
@zhenech @fastlydevs We need actual servers since we host things ourselves where we can control how it's done and properly monitor it. We don't want to have any single points of failure. We host more than updates on these servers but updates are what uses the vast majority of the bandwidth and storage. Network time, connectivity checks, etc. don't consume much. Self-hosting geocoding instead of only a proxy will need a massive amount of memory and storage but it won't fit on the same servers.
@zhenech @fastlydevs We host 2 of our own anycast DNS networks and have servers around the world for our website/network services. We already essentially host a CDN ourselves. OS updates require a ton of bandwidth and a fair bit of storage so that's harder for us to afford with similarly low prices. We also host our own mail server, discussion forum, Matrix chat, Mastodon (this server), etc. We don't plan to move away from self-hosting but rather want to keep heading down the self-hosting path.
@zhenech @fastlydevs We have our own AS and IP space which we use to host 2 separate anycast DNS networks and are phasing in for other uses too including moving to it for email in the near future. Moving to a CDN is just not the direction we're going. Our approach is having a multi-provider setup with no single point of failure for DNS and the OS services where we aren't heavily reliant on any specific provider. Providers can go out of business, go down, change their mind about sponsorship, etc.