How to Backup and Restore Your Minecraft World

Why Backups Matter More Than You Think

Every experienced server owner has a horror story. A griefer with a hacked client who bypassed protections and lava-casted the spawn. A plugin update that wiped every player's inventory. A disk failure that corrupted three weeks of building progress. A mistyped WorldEdit command that flattened half the map.

Backups are the only real safety net. CoreProtect can roll back individual griefing incidents, but it cannot help you if the entire world file is corrupted or if a plugin nukes its own database. A proper backup lets you rewind your entire server to a known-good state, and it takes minutes to set up.

What to Back Up

A complete server backup includes more than just the world folder. Here is what you need:

Path	Contains	Priority
world/	Overworld terrain, player data, level.dat	Critical
world_nether/	Nether dimension	Critical
world_the_end/	End dimension	Critical
plugins/	All plugin jars and their config/data folders	Critical
server.properties	Core server settings	High
bukkit.yml, spigot.yml, paper-global.yml	Server engine configs	High
MySQL databases	Plugin data stored externally	High (if applicable)

If you run additional worlds through Multiverse, include those world folders too. They are stored in the server root alongside the default worlds.

Manual Backups

The Stop-Zip-Start Method

The safest manual backup approach is straightforward:

1. Warn players: /say Server backing up in 60 seconds, expect a brief restart.
2. Stop the server with /stop. This ensures all chunks are saved to disk and no files are being written mid-operation.
3. Compress the server directory:

# Linux, creates a timestamped archive
tar -czf backup-$(date +%Y%m%d-%H%M).tar.gz /path/to/server/

# Windows, use 7-Zip or the built-in zip from PowerShell
Compress-Archive -Path "C:\MinecraftServer\*" -DestinationPath "C:\Backups\backup-$(Get-Date -Format yyyyMMdd-HHmm).zip"

4. Download the archive to your local machine or a separate storage location.
5. Start the server again.

The downside is obvious: your server is offline during the process. For small servers this might be two minutes. For a server with 50 GB of world data, compression alone could take ten or more.

Live Backups With save-off

If you cannot afford any downtime, you can take a backup while the server is running:

/save-all
/save-off
# Now copy/compress the world folders
# When done:
/save-on

The /save-off command tells the server to stop writing to the world files, so you can safely copy them without corruption. Remember to run /save-on afterward, or you will lose all changes made after the backup point. This method is acceptable but not as foolproof as stopping the server entirely, because plugin databases and configs may still be written to during the copy.

Pterodactyl Panel Backups

If your server runs on Pterodactyl (which most managed hosts use these days), you have built-in backup functionality. Go to the Backups tab in your server panel, click "Create Backup," and the panel snapshots your entire server directory. You can download these backups, restore them with one click, and even lock specific backups so they are not auto-deleted when the backup limit is reached.

Set up a backup schedule through the Schedules tab: create a new schedule, add a "Create Backup" task, and set it to run daily during off-peak hours. Pterodactyl will automatically rotate old backups once you hit your plan's backup slot limit. This is the easiest approach for most server owners and requires zero command-line knowledge.

Plugin-Based Backups

eBackup

eBackup is a lightweight backup plugin that runs directly on your server. It compresses your world and plugin files on a configurable schedule, stores the archives locally or uploads them to an FTP server, and handles rotation automatically. Configuration is minimal, set the backup interval, the number of backups to keep, and which folders to include.

DriveBackupV2

DriveBackupV2 takes things further by uploading backups to cloud storage: Google Drive, OneDrive, Dropbox, or a remote FTP/SFTP server. This is the gold standard for backup safety because your backups exist on a completely separate system. Even if your hosting provider's entire data center has a problem, your backups are safe in the cloud.

Setup requires authenticating with your cloud provider (the plugin walks you through an OAuth flow for Google Drive), then configuring which folders to back up and how often. See our automation guide for a full DriveBackupV2 walkthrough.

Restoring From a Backup

When disaster strikes, restoration is simple:

1. Stop the server completely.
2. Rename or delete the current broken world folders (or move them to a broken/ directory so you can inspect them later).
3. Extract your backup archive into the server directory, placing the world folders back where they belong.
4. If restoring plugin data or databases, replace the relevant files in plugins/ and reimport SQL dumps.
5. Start the server.

On Pterodactyl, click the restore button on the backup you want and it handles steps 2-3 for you.

Partial Restore, Specific Region Files

Sometimes you do not need to restore the entire world. If corruption or griefing affected a specific area, you can restore just the region files for that area. Minecraft stores world data in region files located at world/region/r.X.Z.mca where X and Z are region coordinates. Each region file covers a 512x512 block area (32x32 chunks).

To find the region file for a specific coordinate, divide the block coordinate by 512 and round down. For example, blocks around X=1200 Z=-800 fall in region file r.2.-2.mca. Extract just that file from your backup and replace the corrupted version. This approach limits data loss to only the affected area rather than rolling back the entire world.

Testing Your Backups

An untested backup is not a backup, it is a hope. At least once a month, download a backup and try restoring it on a local test server. Verify that the world loads, player data is intact, and plugin databases contain recent entries. There is nothing worse than discovering your backup routine has been silently failing for three months right when you actually need a restore.

Keep at least one backup copy off-site (cloud storage, your local machine, or a separate server) so that a hosting provider failure does not take out both your live server and your backups.

Looking for reliable hosting? Astroworld Hosting runs NVMe SSD, Pterodactyl panel, daily backups, and DDoS protection. See features , plans from €6.39/mo.