The “Can’t keep up! Is the server overloaded?” message means your Minecraft server took longer than its 50 ms tick budget to process game logic. One slow tick can be harmless; repeated slow ticks make movement, block actions, redstone and mob behavior feel delayed. Measure the server first, identify what occupies the long ticks, then change one workload or configuration variable at a time.
Playing with friends? Godlike.Host runs hosted Minecraft Java servers from 10 data-center locations, with free DDoS protection.
This guide explains what the milliseconds and “ticks behind” values mean, how to profile a live server with spark, how to isolate plugins safely, and which Paper controls are useful after you have evidence for a particular bottleneck.
| # | Preview | Product | Price | |
|---|---|---|---|---|
| 1 |
|
Minecraft | Java & Bedrock Deluxe Collection | Windows Digital Code | $39.99 | Buy on Amazon |
| 2 |
|
Minecraft | Java & Bedrock Ultimate Collection | Windows Digital Code | $49.99 | Buy on Amazon |
| 3 |
|
Minecraft | Standard Edition | XBOX Digital Code | $19.99 | Buy on Amazon |
| 4 |
|
Minecraft | $6.99 | Buy on Amazon |
| 5 |
|
Minecraft - Bedrock Edition PS4 | $42.98 | Buy on Amazon |
As an Amazon Associate I earn from qualifying purchases.
Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Scan for outdated or missing drivers - takes under a minute3Clear out junk files and repair common Windows errorsWhat “Can’t keep up” means
The 20-tick-per-second target
Minecraft’s server loop aims to execute 20 game ticks every second. Each tick has a 50 ms time budget. As spark’s Tick Loop documentation puts it, “The Minecraft server aims to run exactly 20 of these game ‘ticks’ every second, or in other words, one tick every 50 milliseconds.” A tick processes player packets, entity updates, mob spawning and artificial intelligence, pathfinding, redstone, chunk work, saves and plugin tasks.
If a tick finishes in less than 50 ms, the server waits for the remainder before starting the next tick. If it takes longer, the next tick cannot start on time: Minecraft’s game ticks do not execute in parallel. The loop therefore falls behind, and the console prints the warning.
#1 Best Overall
- DELUXE COLLECTION — Includes Minecraft: Java & Bedrock Edition, three Bedrock add-ons, three exclusive Character Creator items, and 700 Minecoins.
- CREATE YOUR WORLD — Build whatever you imagine in an infinite world that’s unique in every playthrough.
- EXPLORE AND CRAFT — Discover biomes, resources, and mobs, then craft your way through a sandbox world filled with surprises.
- SURVIVE THE ADVENTURE — Face mysterious foes, travel across exciting landscapes, and venture into perilous dimensions.
- PLAY TOGETHER — Play cross-platform with friends in Bedrock Edition on console, mobile, and PC, or join community servers in Java Edition on PC, Mac, and Linux. Online console multiplayer requires a platform-specific subscription (sold separately).
TPS, MSPT and the backlog
- TPS (ticks per second): the rate the server is actually completing ticks. A healthy server is close to 20 TPS during the measured interval.
- MSPT (milliseconds per tick): how long each tick takes. An average at or below 50 ms leaves enough time for 20 TPS; values above 50 ms indicate that the server cannot sustain the target rate.
- Ticks behind: the amount of scheduled work the loop has missed at that moment.
- Milliseconds in the warning: the size of the slow tick or backlog reported in that message.
The last two values describe the symptom, not its cause. They do not tell you whether chunk generation, entities, redstone, a plugin, saving or another task consumed the time. That requires a profile captured while the problem is occurring.
Is one warning serious?
| Pattern | What it usually means | What to do |
|---|---|---|
| One isolated line, no noticeable lag | A short spike exceeded 50 ms, but the loop recovered. | Record what was happening and monitor; do not change settings solely because of one line. |
| Repeated lines during exploration or teleportation | Chunk loading or generation is creating long ticks. | Profile the slow ticks while reproducing movement or teleportation. |
| Repeated lines near farms or technical builds | Entities, mob AI, redstone or scheduled block work may be saturating the tick. | Profile in that area, then reduce or reconfigure the measured workload. |
| Warnings during saves or scheduled tasks | Saving or a plugin task is producing periodic spikes. | Correlate timestamps with saves and plugin activity before changing autosave or task settings. |
| Low TPS and high MSPT continuously | The workload is larger than the server’s available processing capacity. | Use a long-tick profile, isolate the largest contributor and re-measure after each change. |
Averages can hide a problem. A server may report a healthy average MSPT while occasional very long ticks cause visible freezes. Always inspect maximum or percentile values as well as the average, and correlate them with what players felt.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Measure before changing anything
Check TPS and MSPT with spark
- Run
/spark tpsduring the period when players see lag. - Record the reported TPS, average MSPT and the available maximum or percentile values.
- Repeat the measurement while reproducing the symptom: explore new terrain, teleport, visit a farm, trigger the redstone circuit or wait for the scheduled task.
- Write down the time and activity for each spike. This gives the later profile a concrete event to match.
Do not judge the server from a single quiet reading. A measurement taken after the spike may look normal even though the warning was real. Conversely, a low TPS reading without a corresponding player-visible problem may be a short-lived recovery period; gather several readings under the target player load.
Separate server-loop lag from network symptoms
The warning concerns the server’s tick loop. It is not, by itself, proof of a bad internet connection or a player’s high network latency. If the server’s TPS and MSPT remain healthy while only one player experiences delay, investigate that player’s connection separately. When TPS drops or MSPT rises at the same time that everyone sees delayed actions, continue with server profiling.
Find the work inside slow ticks
Use the tick monitor to catch spikes
Run /spark tickmonitor while the issue is reproducible. The monitor establishes an average tick rate, watches subsequent ticks and reports those exceeding its threshold. Its default threshold is a 100% increase over the established average. You can use an absolute threshold such as 50 ms when you want to flag ticks at the point where the server begins missing its normal budget.
Rank #2
- ULTIMATE COLLECTION — Includes Minecraft: Java & Bedrock Edition, five Bedrock add-ons, five exclusive Character Creator items, and 1000 Minecoins.
- CREATE YOUR WORLD — Build whatever you imagine in an infinite world that’s unique in every playthrough.
- EXPLORE AND CRAFT — Discover biomes, resources, and mobs, then craft your way through a sandbox world filled with surprises.
- SURVIVE THE ADVENTURE — Face mysterious foes, travel across exciting landscapes, and venture into perilous dimensions.
- PLAY TOGETHER — Play cross-platform with friends in Bedrock Edition on console, mobile, and PC, or join community servers in Java Edition on PC, Mac, and Linux. Online console multiplayer requires a platform-specific subscription (sold separately).
Leave the monitor running long enough to capture the activity that causes the warning, rather than stopping immediately after a quiet minute. Note the threshold you used so the profile can be interpreted correctly.
The Tool Desk
Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Profile only the long ticks
After the monitor shows the size of the problematic ticks, run a focused profile such as /spark profiler --only-ticks-over 150, replacing 150 with a threshold below the slow ticks you observed. Restricting the report to long ticks keeps normal, uninteresting work from overwhelming the result.
Open the generated viewer report and inspect the largest stack frames or categories. Look for correlations with:
- chunk generation, loading or sending;
- entity ticking, mob spawning and pathfinding;
- redstone and block-entity updates;
- plugin commands, scheduled tasks or event handlers;
- world saves and other storage activity.
A profile is evidence of what consumed time in the captured window, not a permanent ranking. Capture another profile after a change to verify that the same workload has actually fallen and that a different bottleneck has not replaced it.
Isolate plugins without making the server less safe
Back up and stop first
Paper’s troubleshooting guidance recommends stopping the server and taking a backup before changing plugin files or configuration. A live copy can be writing data while you rename directories or edit settings, increasing the chance of corruption and making rollback difficult.
Quick wins for a faster PC:
Clear out junk files and repair common Windows errorsFree Scan →Scan for outdated or missing drivers - takes under a minuteDriver Scan →Rank #3
- CREATE YOUR WORLD — Build whatever you imagine in an infinite world that is unique in every playthrough.
- EXPLORE AND CRAFT — Discover biomes, resources, and mobs, then craft your way through a sandbox world full of surprises.
- SURVIVAL ADVENTURES — Face mysterious foes, travel across varied landscapes, and venture into perilous dimensions.
- PLAY TOGETHER — Play solo or join friends in local split-screen and cross-platform play across console, mobile, and PC. Online multiplayer supports up to 8 players.
- COMMUNITY PLAY — Connect with players on community servers, or subscribe to Realms Plus (sold separately) to play with up to 10 friends on a private server.
- Announce maintenance and stop the server cleanly.
- Copy the world, configuration and plugin files to a separate backup location.
- Review the latest log and any stack traces for a plugin name, repeating task or event handler.
- If a plugin is implicated, update it from its official distribution source and review its configuration and documented compatibility before testing again.
Use a binary search when the log is inconclusive
To test whether plugins are responsible, temporarily rename the plugins directory (for example, to plugins-disabled) or disable all plugins using your server’s normal administration method. Restart, reproduce the same activity and compare TPS/MSPT and the profile.
- If the warning remains with every plugin disabled, the cause is more likely world workload, chunk activity, saving or the server environment; restore the plugins and investigate those paths.
- If the warning disappears, stop the server and divide the plugin set into two groups. Test one group, then the other.
- Keep required library or dependency plugins with the plugins that use them; otherwise a test may fail for missing dependencies rather than reveal the real performance behavior.
- Continue splitting the suspected group in half until one plugin or a small interaction set reproduces the spike.
Test with the same player count, location and action each time. Changing several plugins, world areas and settings together prevents you from knowing which change mattered.
Match common causes to the evidence
Chunk generation and exploration
Generating previously unseen terrain can produce large, brief ticks, especially when several players travel or teleport in different directions. A profile should show chunk-generation or loading work if that is the cause. If the report does not show it, do not assume that pre-generating terrain or changing chunk limits will solve the warning.
Entities, mobs and pathfinding
Dense farms, item piles, villagers and other entities are updated every tick. Mob spawning and pathfinding can become expensive in a concentrated area. Profile while standing in the affected location, then reduce the measured source—such as an excessive entity population or an unusually complex farm—and verify the result with another /spark tps reading.
Redstone and block entities
Clocks, large contraptions and repeatedly updating block entities can consume time even when players are standing still. Capture a profile with the circuit running and another with it stopped. That controlled comparison is more reliable than guessing from the build’s size.
Rank #4
- Skins! We have biome settlers, city folk, town folk, and more!
- The Nether and all its inhabitants. Fight Ghasts and make friends with Pigmen
- Cross platform play for up to five players between Pocket Edition and Windows 10
- Revamped touch controls, controller support, and a controller mapping screen
- Enhanced Weather effects! Accumulating snow and more
Saves and scheduled work
Periodic spikes that line up with world saves, backups or plugin jobs point to scheduled work. Check timestamps in the server log and profile a window that includes the spike. A task that runs infrequently can create a noticeable freeze without making the long-term average look unhealthy.
Plugin event handlers and commands
A plugin may perform database, file, scan or world operations on the main server thread. The long-tick report can identify the plugin’s stack frames even when the console warning itself cannot. Update or reconfigure the plugin, then repeat the same profile; disabling unrelated plugins does not prove that the implicated one is harmless.
Recommended Free Tools
Paper controls to review after profiling
Paper’s global configuration documents several controls that shape workload. They are workload management tools, not universal fixes. Change one setting, retain your backup, reproduce the same scenario and measure again.
| Control area | What it affects | Trade-off to consider |
|---|---|---|
| Per-player chunk generation | How much new terrain generation one player can cause. | Lowering the allowance can reduce concentrated generation work but may make exploration or teleportation fill in more slowly. |
| Chunk load and send-rate limits | How quickly chunks are loaded for the server and sent to players. | Restricting rates can smooth spikes while increasing the time before a player sees terrain. |
| Worker-thread selection for parallel chunk generation | How chunk-generation work is distributed across worker threads. | More parallelism is not automatically faster; it must match available CPU capacity and the rest of the server workload. |
| Catchup ticks after a lag spike | How the server attempts to process missed ticks after falling behind. | Aggressive catch-up can create another burst of work instead of restoring a stable tick rate. |
| Player autosave limits | How player saving is distributed or limited. | Changing the limit can alter the size and frequency of save spikes; verify that data is still saved as expected. |
| Watchdog early-warning timing | When the watchdog warns that a tick is taking too long. | This changes diagnostic timing, not the underlying workload, and should not be treated as a performance cure. |
Do not disable safety mechanisms simply to hide warnings. If a setting changes player behavior, document the old value, the new value and the measured effect so you can roll back cleanly.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.A repeatable recovery procedure
- Confirm impact: ask whether multiple players see delayed actions and run
/spark tpsduring the incident. - Capture context: note player count, locations and the exact action occurring when the warning appears.
- Catch the spike: run
/spark tickmonitorand choose a threshold that identifies the problematic ticks without recording every normal tick. - Profile: run
/spark profiler --only-ticks-over 150or an equivalent threshold below the observed slow ticks. - Form one hypothesis: select the largest matching category—plugin, chunk work, entities, redstone or saves.
- Make one controlled change: update or isolate a plugin, reduce the measured workload, or adjust one Paper control.
- Reproduce the original scenario: use the same route, farm, teleport or scheduled event and compare TPS, MSPT and the profile.
- Keep or roll back: retain the change only if it improves the measured symptom without unacceptable gameplay side effects.
If the profile points to several contributors, address the largest one first. A small improvement in many unrelated settings is harder to validate than a clear before-and-after result for one dominant source.
Best Value
- Play and share with friends on console, mobile and Windows 10
- discover community creations in the new in-game store
- access new mini games and game modes through servers
Choosing server software or hosting for a lag-prone world
Switching software or providers can change available controls and hardware, but it does not remove the need to measure. Compare the options on the following evidence-based axes:
PC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Crashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minute| Decision axis | Questions to ask |
|---|---|
| Measured headroom | What TPS, average MSPT and worst or percentile MSPT does the option sustain with your actual player count and world workload? |
| Chunk performance | How does it handle generation, loading and sending when players explore or teleport? |
| Compatibility | Will the Paper plugins or mods your world requires run correctly on the target software? |
| Profiling access | Can you run spark or an equivalent profiler while the issue is occurring? |
| Backups and rollback | Are world backups available, restorable and independent of the live server? |
| Support and limits | Are CPU scheduling, storage performance, memory limits and player-resource policies clearly stated and appropriate for your measured workload? |
There is no evidence-based universal player count per CPU, memory amount or guaranteed TPS for every world. Adding memory, moving to faster storage or changing a network cable may be irrelevant when a plugin or main-thread workload is the measured bottleneck. Choose based on repeatable measurements and the controls you actually need.
Common mistakes that prolong the problem
- Treating every warning as an emergency: an isolated spike without player-visible lag may not justify configuration changes.
- Looking only at average MSPT: maximum and percentile values reveal spikes that averages hide.
- Changing five settings at once: you lose the ability to identify the effective fix and make rollback ambiguous.
- Blaming RAM automatically: the warning identifies missed tick time, not a specific memory shortage.
- Disabling the watchdog or catch-up behavior as a “fix”: suppressing a safeguard does not reduce the work in a slow tick.
- Testing plugins without dependencies: a missing library can create a misleading failure.
- Editing files while the server runs: stop cleanly and back up first, as Paper’s troubleshooting guidance advises.
- Assuming a host upgrade guarantees 20 TPS: hardware can help only when it addresses the measured bottleneck.
What a successful fix looks like
After a change, the original reproduction should produce fewer or shorter slow ticks, not merely a quieter console. During the same scenario, TPS should remain closer to 20, average MSPT should stay at or below the 50 ms budget, and the maximum or percentile values should show whether spikes still occur. Players should also be able to confirm that the specific delay—exploration, farm activity, redstone, saving or a command—has improved.
Keep the profile and measurements from before and after the change with the date, player count and scenario. That record makes future regressions easier to identify and prevents a temporary quiet period from being mistaken for a permanent solution.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
