Watchdog Crash
The server is killed by its own Watchdog after a single tick ran far too long. Something froze the main thread, an infinite loop, a stuck operation, or a massive lag spike, and the Watchdog shut the server down to prevent a permanent hang.
What does this error mean?
The Watchdog monitors how long each tick takes. If one tick exceeds the configured timeout (60 seconds by default), it assumes the server is hung and forcibly stops it, dumping the thread state. The dump shows what the main thread was doing when it froze.
A single server tick took 60.00 seconds
Most Common Causes
- An infinite loop in a plugin or mod.
- A heavy synchronous operation blocking the main thread.
- A worldgen or chunk operation that stalled.
- A deadlock between threads.
- Extreme lag (huge redstone/entity load) pushing one tick over the limit.
How To Diagnose
- Read the Watchdog thread dump, the main-thread stack shows what froze it.
- Identify the plugin/mod or operation named in that stack.
- Check what happened just before the freeze in the console.
- Use spark to catch long ticks before the timeout.
Recommended Fixes
- Fix the blocking code
Update or remove the plugin/mod whose stack appears in the Watchdog dump. - Resolve heavy lag
Cut the redstone/entity load or operation that pushes a tick over the limit. - Profile with spark
Catch the long tick in advance and address its root cause. - Adjust the timeout cautiously
Only raise the Watchdog timeout if a long operation is legitimate, never to mask a real freeze.
Frequently Asked Questions
No, it protects you from permanent hangs. Fix the freeze instead of hiding it.
Look at the main server thread's stack trace, the top entries show what was executing when it stalled.
Both are possible, an infinite loop is a bug, while a 60s tick from sheer load is extreme lag.