Satisfactory lag: diagnose a large factory on a server

Separate rendering from simulation

If the picture stutters when facing a dense area but interactions remain responsive, compare client graphics settings. Have a second player observe the same scene. A substantial FPS difference points toward the client machines.

If several players wait for the same interactions or experience position corrections, investigate networking and the server too. The number of buildings visible on one screen is not a direct server-load measurement.

Create comparable test sessions

Retain a world copy and compare ordinary play with a session containing only the necessary test players. Record visited areas and installed mods. A separate lightweight save can provide a comparison without replacing the main factory.

If the issue began with a mod or update, record its version and date. Do not remove a content mod from your only save because dependent objects may be lost.

Inspect hosting measurements during lag

Record memory, CPU load, logs and restarts during the incident. Distinguish sustained pressure from a brief save-related spike. Compare autosave times with pauses players experience.

A factory stopping production only when everyone disconnects may simply be using automatic pause. Check that behavior separately; it does not demonstrate insufficient processing power.

Choose a correction based on evidence

Persistent memory pressure calls for examining capacity and world growth. A network issue affecting one player warrants testing that connection first. A mod-specific regression calls for a compatible test or recovery using an appropriate backup.

If an instance limit is established, give support the measurements and a description of the factory. Compare prospective hosting with an identical copy and the same actions. A generic claim of lag-free hosting is not a benchmark result.

Frequently asked questions