// Hosting

Server location and ping: why the Netherlands is a strong hub

Server location and ping: why the Netherlands is a strong hub

Anyone renting a game server looks at RAM, CPU and price, and rarely at the question players ask first: what is the ping? Ping is the time a packet needs to travel from the player to the server and back, in milliseconds. Below 30 ms a game feels instant, between 30 and 60 ms it is fine for survival and building, above 80 ms players in shooters and PvP notice that their actions arrive late. The location of the server determines the largest part of that, but not in the way most people think: distance is only one factor, and often not the most important.

What really determines ping

Light in fibre travels about 200 kilometres per millisecond. Amsterdam–Berlin is 650 kilometres, so the physical floor for a round trip is roughly 7 ms. In practice you measure 15 to 25 ms, and that difference is in three things:

  • The route. Traffic does not follow the shortest path but the path networks have agreed among themselves. A provider without a direct link to the data centre's network sends traffic on a detour, sometimes via Frankfurt or London, and that costs tens of milliseconds.
  • Peering. Data centres connected to large internet exchanges swap traffic directly with hundreds of providers. The more direct links, the fewer detours.
  • The last mile. Wi-Fi, an overloaded cable connection or a 4G link add 5 to 40 ms that no server location can compensate.

Why the Netherlands is a favourable hub

The Netherlands has, with AMS-IX, one of the largest internet exchanges in the world, and almost every European provider is connected to it directly or through a partner. For players in the Netherlands and Belgium that means pings of 5 to 15 ms; for Germany, northern France and the United Kingdom 15 to 25 ms; for Poland, Denmark and Austria 25 to 35 ms. A server in the Netherlands thus sits "in the middle" for all of north-western Europe, while a server in Finland or Spain, for example, goes over 50 ms for half of those players.

Our servers are in Hengelo (PDC2), with redundant power and network and direct connections to the large Dutch and German providers. That is no coincidence: most of our players are in the Benelux, Germany and Poland, and for those three markets the east of the Netherlands is one of the best-placed locations there are.

Per game: how much ping is acceptable?

  • Minecraft (survival, building): up to 60 ms nobody notices; PvP and minigames feel noticeably sharper under 40 ms.
  • FiveM: under 50 ms for smooth driving and sync; above 100 ms players see cars "jump".
  • Counter-Strike 2, Rust, DayZ: every millisecond counts; under 30 ms is the goal for competitive play.
  • Palworld, Enshrouded, Valheim, 7 Days to Die: co-op is forgiving, playable up to 80 ms, but fights feel better under 50 ms.

Measuring yourself: ping, traceroute and the difference

Never measure only the ping to the server IP; measure the route as well. On Windows tracert ip and on Linux and macOS traceroute ip show the delay per hop. If you see a jump from 5 to 40 ms at one hop, there is a detour or an overloaded link, and that is information your host (or your own provider) can act on. Measure at three moments: during the day, in the evening between eight and ten (the busiest hour) and late at night. A location that has 30 ms extra in the evening has an uplink or peering problem, not a distance problem.

With us you can ping a test IP before ordering; it is on the store pages of the games. If you want to know what your players measure, ask three people from different countries to send a traceroute: that says more than any marketing claim about "low ping".

Two properties of a data centre do not show on a map. The uplink of the node: a server sharing a 1 Gbit connection with dozens of others suffers packet loss in the evening, which players experience as "rubberbanding" even though the ping is low. And the DDoS filtering: some protection services route all traffic through a scrubbing centre in another country, which permanently adds 10 to 30 ms. Filtering done in the data centre itself keeps the route short. Ask about it; a good host can explain. How our protection works is in the knowledge base under DDoS protection.

What you can do about ping yourself

The server location is half the story; the other half is with the player. A wired connection easily saves 5 to 20 milliseconds compared to Wi-Fi and above all a lot of fluctuation (jitter), which in shooters is worse than a slightly higher but stable ping. Close downloads and streams on the same network while playing. Providers with a link at AMS-IX or direct peering with the data centre give the shortest route; a traceroute shows whether your provider takes a detour, and that is sometimes a reason to switch providers. And look at the game's own client settings: interpolation and network buffers are adjustable in many games and compensate for part of the delay.

Serving several countries

If your community plays in two regions far apart, for example the Netherlands and southern Europe, nobody wins with one server in the middle: everyone then has 40 ms. Better is one server where most players are and accepting that the rest has a bit more ping, or — for a network — two servers with a proxy such as Velocity that sends players to the nearest server. For most servers that is unnecessary; for a network with hundreds of players it is.

In short

If you want to experience it yourself: every game server at MC-Node is in Hengelo, online within a minute and cancellable monthly, so you can test your own group's ping before committing to anything longer.

// TRY IT YOURSELF
Your Minecraft server online in 60 seconds
View packages →