Minecraft Server Crashing? Read the Crash Report

Find the crash report and log on your Minecraft Java server, read the lines that matter, and fix the usual causes: Java, mods, memory and bad chunks.

6 min read · Updated Sep 23, 2026 · By GuavaHost Team

When a Minecraft Java server crashes, it almost always writes down why, either in a crash report or in its log. This guide shows where those files are on your GuavaHost server, which few lines actually matter, and how to fix the causes we see most: the wrong Java version, mismatched or client-only mods, running out of memory and corrupted world data.

If your server never gets past Starting at all, read Why your server says Starting or is stuck offline as well.

Where the evidence lives

Open the Files tab. These are the places to look:

Open them in the panel or download them over SFTP. Not every crash writes a crash report, so if crash-reports is empty, go to latest.log.

If the process was killed from outside, for example for using too much memory, the Console shows a message from the panel saying the server process was detected in a crashed state, with an exit code.

How to read a crash report

1. Open the newest file in crash-reports.
2. Read the Description: line near the top. Common ones are Exception in server tick loop, Ticking entity, Ticking block entity and Watching Server.
3. Read the line straight after it. That is the actual error, for example java.lang.OutOfMemoryError: Java heap space.
4. Look at the first few lines starting with at underneath. If you see a mod or plugin name in them, that is your prime suspect.
5. For Ticking entity and Ticking block entity, scroll down to the section describing the thing being ticked. It gives its type and coordinates.
6. At the bottom, System Details lists the Minecraft version, Java version and mod list. Forge and NeoForge reports often include a Suspected Mods line too.

In latest.log, scroll up from the bottom to the first line containing ERROR or Exception. Later errors are often just knock-on effects of the first one. How to read server logs and crash reports explains log levels and stack traces in more depth.

The usual causes and their fixes

The wrong Java version

The log says something like:

java.lang.UnsupportedClassVersionError: ... has been compiled by a more recent version of the Java Runtime (class file version 65.0), this version of the Java Runtime only recognizes class file versions up to 61.0

Those numbers are Java versions in disguise: 52 is Java 8, 61 is Java 17, 65 is Java 21 and 69 is Java 25. The message above means the server needs Java 21 but is running Java 17. Switch it with the Docker Image setting on the Startup tab, using the table in Choosing a Java version.

It works the other way too. Very old modded versions, such as Forge 1.12.2, crash on modern Java and need Java 8.

A mod or plugin that does not match

Mods must match the loader (Fabric, Forge or NeoForge) and the exact Minecraft version, and many need a library mod alongside them. Fabric tells you with lines saying a mod requires another mod "which is missing". Forge and NeoForge print Missing or unsupported mandatory dependencies followed by the mod IDs involved. Install the library or the correct version it names.

If the crash started right after you added something, remove the newest mod and restart. Adding mods back one or two at a time finds the culprit quickly.

Plugins on Paper and Spigot are gentler. A broken plugin usually logs a line like Error occurred while enabling SomePlugin (Is it up to date?) and the server keeps running without it. Update that plugin or remove it.

Client-only mods on the server

Some mods only make sense on a player's computer: OptiFine, shader loaders and many purely visual or interface mods. On a server they crash on startup. On Forge the error reads Attempted to load class net/minecraft/client/... for invalid dist DEDICATED_SERVER. On Fabric it mentions environment type SERVER. The report usually names the mod. Delete it from the server's mods folder; players can keep it in their own game.

Running out of memory

Look for java.lang.OutOfMemoryError in the log, or a panel message that the process was killed with exit code 137 or flagged as out of memory. Modded servers are the most common victims.

Before adding RAM, try the cheaper fixes: lower view-distance and simulation-distance (see View distance vs simulation distance), pre-generate your world, and remove heavy mods you do not need. If you genuinely need more, How much RAM does my Minecraft server need? helps you size it and Changing your server RAM shows how to upgrade.

A corrupted chunk or world file

If the server crashes when loading the world, or whenever someone walks into one area, the log often mentions a chunk position or a region file. Region files live in world/region and are named like r.2.-1.mca. Each one covers 512 by 512 blocks, so divide a block coordinate by 512 and round down to find the file.

Your options, safest first:

1. Restore a backup from before the problem, from the Backups tab. See Backups.
2. Download the world and use a desktop chunk editor such as MCA Selector to delete only the damaged chunks. They regenerate as fresh terrain.
3. Delete the whole region file. That 512 by 512 block area regenerates from scratch, and any builds in it are lost.

If level.dat itself is damaged, the world folder usually has a level.dat_old copy you can rename into place.

Ticking entity or block entity crashes

One specific mob, item or block entity (like a chest or machine) errors every tick, and the report gives its type and coordinates. Restoring a backup is the cleanest fix. On Forge, the server config file forge-server.toml has options to remove erroring entities and block entities automatically instead of crashing. Only use them with a backup, because they delete the thing that errors.

The server stopped responding

When one tick takes far too long, a watchdog shuts the server down. On Vanilla, Fabric and Forge this produces a crash report with Description: Watching Server, controlled by max-tick-time in server.properties. On Paper and Spigot the log says The server has stopped responding! and prints a thread dump. The dump shows what the main thread was stuck on, and a plugin or mod name in it is your lead. Raising the timeout only hides the problem; Find what's causing lag with spark finds the real cause.

A calm routine

1. Take a backup, then read the first real error, not the last one.
2. Change one thing, start the server and read the log again.
3. If you are stuck, copy the Description line, the error line and the first dozen at lines. That is exactly what we need to help.

Still stuck? Jump into our Discord at discord.gg/GuavaHost and the team will help you out.

• crash
• crash report
• latest.log
• troubleshooting
• minecraft
• java
• mods

Host a Minecraft Java server From $3.99/mo, with instant setup and DDoS protection.

Still need help? Our team is online 24/7 to answer questions.

Contact support