Eco performance, why your server slows down at month 3 and what to do

Most Eco admins meet the same wall: the server runs smoothly for a month or two, then starts to drag. Specific symptoms: stutter during peak play hours, slow responses to crafting interactions, market price updates that take seconds, animals spawning slowly.

This is not your imagination. Eco accumulates load in predictable ways. Understanding why lets you intervene before it kills the playthrough.

What's happening under the hood

Eco's simulation tracks more state than most multiplayer games. As the world matures, all of the following grow:

• Explored area. Every loaded chunk persists.
• Built structures. Each building's blocks contribute to world calculations.
• Animals and plants. The biosphere simulates beyond just "spawn / kill."
• Pollution. Tracked per-region, accumulates over time.
• Contracts and laws. Each active contract is checked.
• Economic transactions. Markets process recurring orders.
• Skill data per player. Grows linearly with player count and progression.
• Government data. Elections, terms, vote records.

Most of these are O(1) per item, but at month 3 of an active server, the total can balloon to thousands of objects each tick.

The symptoms in detail

Match what you're seeing to its likely cause:

Symptom	Likely cause
Slow market responses	Many active contracts evaluating each tick
Stutters during industry use	Pollution / emission propagation math
Slow chunk loading on teleport	Save file is large, disk I/O bottleneck
Animals scarce or slow to respawn	Biosphere simulation throttled
Government UI laggy	Accumulated election / term / vote data

The tick budget

At 70 ms average, players will mostly enjoy it but one more load (new district, new player) tips you over. At 100+ ms, intervention is urgent.

Solutions, ranked

Solution 1: Audit and prune contracts

Eco contracts persist. Players create them and forget. By month 3, your server may have hundreds of inactive contracts evaluating every tick.

Admin tool: open the contracts panel, sort by activity. Anything inactive for 14+ days can be cancelled with a note to the originator.

Easier alternative: pass a law that cancels contracts inactive for X days. Self-cleaning.

Solution 2: Reduce simulation granularity

Eco config exposes simulation tick rates for various systems:

• Plant growth
• Animal AI
• Pollution propagation
• Climate

For a struggling server, you can lower the frequency of non-critical systems. Plants don't need to tick every second; once every 5 seconds is invisible to players.

Specific config keys vary by Eco version. Look in Gameplay.eco for TargetUPS and in EcoSim.eco for simulation tick settings.

Solution 3: Reduce world area

If the world is sprawling (players have explored 5x the population can use), consider:

• Setting a world border to a sensible playable area.
• Pruning chunks that have no built content.

This dramatically reduces the simulation surface area.

Solution 4: Restart more often

Eco's per-tick work is partly cumulative state that resets on restart. Long-running servers benefit from daily or every-other-day restarts to clear in-memory accumulation.

Set up an automatic nightly restart. Players adjust quickly.

Solution 5: Hardware upgrade

If the server is genuinely CPU-bound after the above optimizations, more CPU helps. Eco uses cores reasonably well (better than most games), so adding cores helps somewhat. Single-thread speed still matters for the hot path.

This is the most expensive solution. Try the others first.

A diagnostic process

Before fixing, figure out which subsystem is hot.

1. Check the server log. Eco logs tick times when they exceed budget. The system that's slow usually names itself.
2. Watch the simulation panel if your version exposes one.
3. Toggle suspect systems off temporarily (in a test setting) to see which one changes the picture.

Don't apply five fixes at once. You won't know which one helped.

When to take a server-wide pause

Sometimes the best fix is a planned downtime for cleanup:

1. Announce a 24-hour maintenance window.
2. During downtime: prune contracts, prune chunks, audit and optimize.
3. Restart fresh.
4. Players come back to a noticeably smoother experience.

Communicate clearly. Players understand maintenance if you frame it as "to keep the server alive longer."

A long-term schedule for healthy Eco servers

Adapted from what's worked on the servers we host:

When	Action
Week 1	Setup, mild monitoring
Week 2-3	Smooth play, light maintenance
Week 4	First contract audit, performance baseline
Week 6	Second contract audit, check pollution patterns
Week 8	Mid-playthrough maintenance window. Prune inactive content.
Week 10	Assess: approaching endgame or stalling?
Week 12	Meteor approach, performance matters more

Most pacing problems trace to skipping the week 4 / week 6 audits.

A surprising fact

Server slowdowns in Eco are mostly admin-fixable. Few problems require hardware upgrades. Most are content accumulation that pruning can clear.

30 minutes of contract audit can recover 20 percent of tick budget. The reason most servers never improve: admins haven't tried the prune steps.

Common mistakes

Doing nothing and hoping. Performance only gets worse.

Restarting without pruning. Restart helps with in-memory state but doesn't clear persistent state. Both are needed.

Resetting the world to fix performance. A massive overreach. Try the lighter fixes first.

Blaming hardware first. Hardware is usually fine. Content is usually the issue.

Conclusion

Eco servers slow down predictably. The slowdown is content accumulation, not engine deterioration. Audit contracts. Prune chunks. Restart regularly. Adjust simulation rates. Most "the server died" stories are actually "the admin gave up before doing maintenance." With light recurring effort, an Eco server runs fast through the entire 60 to 90 day arc.