Forge Server Tools · every failure mode · guides
The server hung and the watchdog killed it — FATAL
One tick exceeded max-tick-time, so the watchdog assumed a deadlock and shot the server. On a heavy modded pack this is often a slow operation (worldgen, a big machine, chunk loading) rather than a real deadlock — but the stack trace in the crash report names whatever was running.
Short term: set max-tick-time=-1 in server.properties to disable the watchdog. Most modded packs ship this way on purpose. Real fix: read the stack trace in the crash report -- the top frames name the mod that took too long. If it recurs at the same place, that is your culprit.
The watchdog dumps the stack of the thread that was stuck. That trace is a real diagnosis, not a symptom — read it the way you would any crash.
Treated as a performance problem, so people raise the limit. The limit is not what broke.
Fix the mod the trace names. Raising max-tick-time hides it; setting it to -1 turns a crash into a server nobody can even disconnect from. See lag vs. a real stall.
Search your crash report or latest.log for any of these:
A single server tick took ... seconds ServerHangWatchdog Considering it to be crashed, server will forcibly shutdown
The free edition covers the six most common startup failures, and this is not one of them — it is in the paid tool, which checks all 24 and names the mod jar from the stack trace. The fix above is the same one it prints, so you do not need it to solve this.