Two games break more rented servers than almost anything else people self-host: Rust and Palworld. Not because they're badly written, but because they're memory-creep titles that grow their footprint the longer they stay up, and because a Rust wipe multiplies load in a way a flat monthly slot count never captures. This is a notes-from-the-rack look at how to do rust server hosting and Palworld hosting properly on a server you run yourself: sizing RAM per game, capping slots at the number the map actually supports, and scheduling restarts so RAM doesn't drift into swap by Thursday. Noded does not sell managed game plans; managed Rust and Palworld servers are sold by our sister brand ZeroPing.Cloud.
Why Rust and Palworld need different sizing than a Minecraft box
A quiet vanilla server sits at a predictable RAM baseline. Rust does not. Entity count climbs across a wipe as bases grow, and the server process holds more resident memory every day until the next wipe resets the map. Palworld has the same shape for a different reason — pal spawns, base structures, and save-state bloat push memory up steadily, and the dedicated server binary is notorious for leaking across long uptimes. Both are what we call memory-creep games: the RAM you need at hour one is not the RAM you need at hour seventy.
The fix isn't one heroic oversized box. It's honest per-game RAM plus a restart cadence that clears the creep before it hurts. Set slot caps that match what the map and netcode can actually hold, not an aspirational number that turns into rubber-banding at prime time.
Rust server hosting: wipe-aware, restart-scheduled
On Noded you run the Rust dedicated server yourself on a VPS or dedicated server. Set the max player count to what a wiped map with an active population actually holds on your RAM tier, so a full server on wipe day doesn't stall on entity updates. Schedule restarts for exactly the memory-creep reason above: a nightly or pre-wipe restart returns the process to a clean baseline instead of letting it grind toward swap. If you run a weekly or bi-weekly wipe, align the restart and map reset so players get a fresh server at a predictable hour rather than a mid-fight crash when RAM finally runs out.
For a Rust dedicated server in Europe — whether it's for a big modded map, Oxide plugins with heavy memory needs, or a growing community — our KVM VPS tiers give you full root, from VPS IV (16 GB) for a small vanilla map up to VPS V–VI (24–32 GB) for big or modded maps. Facepunch's own server guide asks for 12 GB of free RAM, which rules out the 8 GB tier. You tune the box yourself.
Palworld server RAM and slot reality
Palworld's dedicated server tops out at 32 players, and it gets unstable well before you'd hope on wishful player counts — the honest cap is where the save stays healthy and pals keep pathing. On Palworld server RAM: this is the title where scheduled restarts earn their keep the most. The binary's memory footprint drifts upward over long uptimes, so restart on a schedule to reclaim it before players notice stutter. Anyone who has run a self-hosted Palworld box for a week knows the "it was fine yesterday" slowdown — that's the creep, and a restart cadence is the boring, correct answer.
What it costs today
On Noded you run Rust, Palworld or any other title yourself on a VPS from EUR 4/mo (EUR 0.0056/hr) or a dedicated server from EUR 209/mo. For Rust and Palworld, start at VPS IV (16 GB, EUR 16/mo).
DDoS protection for game servers, not just web
Game servers get hit differently than websites. UDP floods, spoofed source packets, and reflection attacks target the exact ports your players connect on, and a generic scrubbing rule that works for HTTP will happily drop legitimate game traffic along with the attack. Our game server DDoS protection is always-on 100 Gbps Layer 3/4 mitigation included on every VPS and dedicated server, with higher DDoS protection tiers available as a per-server add-on.
Latency matters as much as uptime for these titles. We run on our own AS60982 with VPS in six EU locations plus Ashburn, so put the server in the one closest to most of your players. How the network is connected is on our network page if you want to check the path to your region before you commit.
How Noded Can Help
Rust and Palworld are the games that misbehave, not the games that behave. Run either yourself on an hourly VPS with root and KVM, billed by the hour from EUR 0.0222/hr (VPS IV, 16 GB): set your own restart schedule around your wipe cadence, and get 100 Gbps of Layer 3/4 DDoS protection included on our own network. If you would rather rent a managed server, our sister brand ZeroPing.Cloud sells both.
FAQ
How much RAM does a Palworld server need?
Palworld's own dedicated-server guide lists 16 GB with a 4-core CPU and recommends 32 GB or more (8 GB boots but risks out-of-memory crashes), which maps to our VPS IV (16 GB, 4 vCPU) as the floor and VPS VI (32 GB, 8 vCPU) for a busy server. The server drifts upward in memory over long uptimes, so schedule restarts to reclaim RAM before players notice stutter rather than relying on one oversized box.
How many players can a Rust server hold?
It depends on map size, plugins and RAM rather than a number on a plan. On a server you run yourself, you set the max player count: size it to what a wiped, active map holds on your RAM tier, schedule restarts, and align map resets with your wipe cadence so a full server doesn't stall on entity updates at prime time.
Do Rust and Palworld servers include DDoS protection?
Yes. Every VPS and dedicated server you run them on includes always-on 100 Gbps Layer 3/4 DDoS protection.
Can I get a dedicated Rust server in Europe instead of a shared slot?
Yes — that's how it works on Noded, which does not sell managed game slots. Run the Rust dedicated server on an hourly VPS with full root and KVM, billed by the hour on our own AS60982 network in six EU locations, or on a dedicated server from EUR 209/mo.
Why schedule restarts?
Rust and Palworld are memory-creep titles — the server process holds more RAM the longer it runs, drifting toward swap and stutter. Scheduled restarts return the process to a clean baseline at a predictable hour, which is far less disruptive than a mid-session crash when RAM finally runs out.