Hébergement, bande passante et ingénierie réseau, écrits par l’équipe qui exploite l’AS 25198 — capacité illimitée, peering, stockage et les arbitrages qui vont avec. Les articles eux-mêmes sont publiés en anglais.
A streaming server does three jobs with three different bottlenecks. Ingest wants path quality, transcode wants cores, egress wants a port. Size them separately, work a real build end to end, and see why the blended ABR bitrate — not the top rung — sets the uplink.
Monthly transfer is the wrong starting point, because a port is a rate. Size from peak concurrency, per-unit bitrate and headroom — then check the cost per gigabit of each tier, because that curve is not monotonic and the cheapest gigabit is not where most people buy.
A virtual machine is a slice of a box; a dedicated server is the box. The useful question is where the slice stops being enough. Four ceilings decide it — single-thread speed, disk latency, packet rate and RAM — and an afternoon of measurement settles it better than a year of guessing.
The machine is the easy part. Address reputation, geolocation and an abuse process that answers decide whether the service works at all — then conntrack, MSS clamping and IPv6 decide whether it scales past one node.
A game server is a simulation loop with a fixed deadline, and no amount of cores makes one thread finish sooner. What to buy for a single instance, what to buy for density, and why jitter beats port size.
The edges take the load until a purge, a deploy or a cold region sends every one of them to the origin at once. Hit-ratio arithmetic, the cache headers that are the whole contract, and which machine to actually build.
The failures in colocation are rarely technical: missing rails, a dual-PSU server on one feed, nobody with the IPMI password. Space, power, network, remote access and paperwork, in the order they cost you a day.
All three are still shipping because they solve different problems. The interface ceilings, the queue models, the DWPD and power-loss specifications that actually matter, and how to pick the mix for your working set.
The port sets the ceiling; disk, TLS and connection state decide the number you actually reach. Bitrate arithmetic for every tier, a worked 10 Gbps live example, and the three resources that run out before the network card does.
Transit buys reach to everywhere by the longest road. Peering buys a short path to the networks you actually send to. What each one costs, how BGP picks, and how to read the difference from your own shell.
Same network, same team, a site that finally explains it. Specs and prices in the open, live stock on every server table, and eight languages.
Unmetered is not unlimited-with-an-asterisk. It means you pay for the port speed and never for the bytes that cross it — here is what that changes in practice.