> ## Content Index > Fetch the complete content index at: https://dedicatedmc.io/blog/llms.txt > Use this file to discover other available public pages before exploring further. # Fixing 'Can't keep up! Is the server overloaded?' - URL: https://dedicatedmc.io/blog/fixing-cant-keep-up/ - Published: 2026-10-30T13:43:24.000Z - Updated: 2026-10-30T13:43:25.000Z - Author: Mark E - Tags: Guide, Minecraft, Performance If your server console is printing "Can't keep up! Is the server overloaded? Running Xms or Y ticks behind", you have found the single most-searched Minecraft error string, and also one of the most misread. It is not a disk warning, it is not a corruption alert, and it is not telling you to buy more RAM. It is the server saying it fell behind the clock, and this post explains exactly what that means, what usually causes it, and how to find the cause rather than guess at it. ## What the message actually means A Minecraft server aims to run at 20 ticks per second. A tick is one step of the simulation: every mob moves, every hopper checks itself, redstone updates, chunks load and save, player actions resolve. Twenty ticks a second gives the server 50 milliseconds to finish each one, because a second divided by twenty is 50ms. That is the budget for a single tick. When a tick takes longer than 50ms, the server has missed its slot against the wall clock. It notices that it is now behind real time, and it prints "Can't keep up" to tell you it is skipping ahead to catch up. The number it quotes, "Running 2000ms or 40 ticks behind", is how far behind it had drifted before it caught itself. So the message is a symptom of one thing: a tick that ran over 50ms. Everything in the game is timed off ticks, so while the server is behind, the world runs in slow motion for everyone. Mobs move in slow steps, crops grow slower, hoppers crawl, blocks break with a delay. This is the same underlying problem as low TPS, described from the server's point of view. The tick got heavy, and it could not finish in time. What the message is not matters just as much. It is not a disk error, even though the server touches the disk during a tick. It is not about running out of memory, though a server thrashing in garbage collection can trigger it as a side effect. And it is almost never fixed by adding RAM. The tick was not waiting on memory capacity, it was waiting on compute or on work it could not get through fast enough. ## This is a server problem, not your connection Before you change anything, rule out the two things people confuse with this. "Can't keep up" is printed by the server, about the server, and it affects everyone on it at once. If the whole world is stuttering for every player, that is this problem. High ping is different. Ping is the round-trip time between one player's machine and the server. If a single player is rubber-banding while everyone else is smooth, the server is probably ticking along fine at 20 TPS and that player's connection is the issue. You will not see "Can't keep up" in the console for a ping problem, because the server is keeping up perfectly well. Client-side FPS is different again. Frames per second is how fast a player's own computer draws the game. Low FPS makes the picture stutter even when the server is healthy, and no server tuning touches it. The console tells you which of these you have: if it is printing "Can't keep up", the server missed its tick budget, and that is where to look. We pull these three apart in more detail in [what causes Minecraft server lag](https://dedicatedmc.io/blog/what-causes-minecraft-server-lag). ## What commonly causes it A short list of things eats most tick budgets, and the cause is almost always one of them. Entities and mob farms. Every mob, item, minecart and armour stand is ticked. A large mob farm, a floor covered in dropped items, or a big animal pen from breeding can put thousands of entities into the tick, and each one costs time. This is the most common single cause of a heavy tick. Chunk loading and generation. Loading and generating chunks is heavy work that lands right in the tick when players explore into fresh terrain, when chunk loaders keep distant areas active, or when view and simulation distance are set high. One player flying into unexplored land can spike a tick on its own. Heavy plugins or mods. A plugin or mod doing expensive work on the main thread, scanning chunks, hitting a database synchronously, or looping every tick, takes its time straight out of your 50ms. A single badly-behaved one can dominate the whole budget. Modpacks stack this effect, because each tick simply has more to do. Redstone and hoppers. Redstone updates and hopper item checks are surprisingly expensive at scale. A large contraption or a wall of hoppers polling for items adds real milliseconds every tick, whether or not anything is moving through them. An undersized or shared CPU. The tick runs on a single thread, so what finishes it in time is per-core speed and cache, not core count. On a slow core, or a core you are sharing with noisy neighbours on an oversold box, the same workload that would fit inside 50ms elsewhere runs over it here. If the workload is already lean and the tick is still late, this is where to look. ## How to diagnose and fix it The tools are good, so you do not have to guess. Start with `/tps` in-game for a quick read of the current ticks per second. At or near 20 the server is keeping up. Down in the teens, the tick is over budget, which is what the console is already telling you. Then watch MSPT, the milliseconds each tick actually takes. Most modern server software shows it. A low average with nasty spikes usually points at chunk loading or a periodic task, while a steadily high average points at a constant cost like entities or redstone. To find the specific cause, use the [spark](https://spark.lucko.me/?ref=dedicatedmc.io) profiler. It samples the server while the problem is happening and shows exactly where the tick's time is going, down to the plugin, the entity type, or the chunk. Take the profile while the console is printing "Can't keep up", not afterwards, because the spike is the evidence. spark runs on Paper, Purpur, Fabric, Forge and NeoForge, and Paper-family servers from 1.21 onward ship it built in. Once spark has named the cause, the fix is usually obvious, and it follows the same order every time: measure, drop view and simulation distance, thin entities and mob caps, track down hopper and redstone hotspots, pre-generate the world, audit plugins one at a time, then tune the JVM, and only then look at hardware. Our [Minecraft lag checklist](https://dedicatedmc.io/blog/minecraft-lag-checklist) walks that sequence step by step, and it is the right next stop once you know what is eating your tick. Hardware is deliberately last, because a faster chip papers over a lot but cannot outrun a runaway hopper array or a plugin scanning the world every tick. Fix the workload first. If you have done that and the tick is still late, the honest answer is per-core speed and cache, which [choosing a CPU for Minecraft](https://dedicatedmc.io/blog/choosing-a-cpu-for-minecraft) covers in full. An honest footnote. The tuning we do underneath the floor is more than most servers need, and none of it outruns a bad workload. What it does is remove the host from the list of things you have to rule out, which is exactly why hardware sits last on that list rather than first. If you would rather run on hardware picked for a low, flat MSPT with reserved cores that no neighbour can steal, that is what our [Minecraft hosting](https://dedicatedmc.io/minecraft-server-hosting?ref=dedicatedmc.io) is built for.