Read logs and measure backend performance

Read the backend logCopy Anchor Link

For Docker Compose, run:

docker compose logs --tail=200

Use docker compose logs -f for live output. For Java, preserve the process console and any configured log files. The automated setup provides a log action in its management menu.

Record the timestamp, affected instance, bot name, and full error. The first error can explain later reconnect or task errors.

Identify which server failedCopy Anchor Link

A healthy SoulFire backend can have bots rejected by Minecraft. A healthy Minecraft server can receive too little load from an overloaded bot backend. Compare logs and resource metrics from both machines before changing settings.

Measure before tuningCopy Anchor Link

Record CPU, memory, network throughput, connected bots, active scripts, and tasks. Add load in stages with the testing procedure. Inspect a profiler capture if CPU remains high or a repeated action stalls. The backend's spark command exposes its profiling tools. Use spark for the live help.

Tune one factor at a timeCopy Anchor Link

• Increase join spacing if bursts cause login failures.
• Reduce unrelated scripts, camera captures, and plugins to isolate their cost.
• Change the Java heap limit only with evidence of memory pressure.
• Keep memory available for the operating system and other processes.

Bot capacity depends on world activity, server versions, plugins, hardware, and workload. Use measured results for your deployment instead of treating a fixed bots-per-gigabyte table as a guarantee.

For startup failures, use dedicated server troubleshooting.