Forge Server Tools · every failure mode · guides
This message is normal during chunk generation and alarming at 20 seconds. How to tell routine stutter from a real problem, and how to find what is eating the tick.
Can't keep up! Is the server overloaded? Running 2150ms or 43 ticks behind
A Minecraft tick should take 50ms. When one takes longer, the server falls behind and prints this, with how far behind it got.
This is the part that matters: a few thousand milliseconds during chunk generation is normal. A player flying into unexplored terrain generates chunks, which is expensive, and the server catches up immediately afterwards.
| Behind by | Meaning |
|---|---|
| Under 5,000ms, occasional | Normal. Chunk loading or a world save. |
| 5,000–20,000ms, repeating | Something is wrong. Worth investigating. |
| Over 20,000ms | A real stall. Find it. |
| 60,000ms | The watchdog kills the server. See watchdog timeouts. |
If you alert on the first row you will train yourself to ignore the alerts entirely, which is worse than having none.
Guessing from mod names does not work. Measure:
/spark profiler start, then
/spark profiler stop) gives you a tree of where tick time went,
by mod and by method. It is the single most useful tool here and it is free.Do not raise max-tick-time to silence it, and never set it to
-1 outside of debugging. That disables the watchdog, which converts a
crash into a server that hangs forever and that nobody can even disconnect from.
Forge Server Doctor performs the whole procedure above — reads the report, finds the real cause under the wrappers, skips the vanilla frames, and names the mod jar. 24 failure modes, each with its fix. The free edition covers the six most common startup failures.