Direct answer
Host CPU percentage and memory use alone do not identify the cause of server lag. Compare Linux process metrics, JVM heap and a Minecraft profiler during the same window.
Recommended procedure
- Run
uptime,free -handtop -b -n 1to inspect load averages, available RAM and busy processes. - Use
ps -eo pid,comm,%cpu,%mem,rss --sort=-%cpu | headto identify the Java process and competing workloads. - On containers compare
memory.max,memory.eventsandcpu.statwhere available; match these readings with MSPT and player count.
What to change
uptime
free -h
top -b -n 1
ps -eo pid,comm,%cpu,%mem,rss --sort=-%cpu | head- If an external backup or another Java process consumes CPU, schedule it outside peak hours.
- If memory pressure or swapping is evident, adjust the real workload/heap only after accounting for container limits.
- If the Linux host appears healthy but server MSPT is high, profile entities, plugins or chunk generation rather than guessing.
How to verify the result
Collect the same command outputs and spark profile during another comparable peak. Identify whether host contention, GC or server tick work changed.
Version and safety notes
Linux load average is not directly CPU percentage. free may show host RAM inside a container; MemAvailable differs from MemFree. Folia/Paper tick models also differ.
Official references
Related guides
Help us improve this article
Tell us what is missing, unclear, outdated, or incorrect. You can also suggest something we should add.