Skip to Content
Enderdash

Plan and verify backups

Use your runtime backup tooling and test restoration before automating it.

Enderdash can run a backup script through an agent-side shell automation. It does not turn a downloaded file or successful scheduled command into a verified server backup.

Before you begin

Choose backup tooling appropriate to the game runtime, databases, and container storage. You need access to those resources and a destination outside the data being backed up.

For shell automation, the agent needs terminal support and the required access. Confirm the backup tools exist in the agent's execution environment.

Establish a restore procedure first

  1. Identify the worlds, configuration, plugin data, databases, and persistent volumes that must be restored together.
  2. Follow each application's documented procedure for a consistent snapshot. Some need a save, a database dump, or downtime.
  3. Run your backup script manually as the intended agent user.
  4. Inspect the resulting files and their destination.
  5. Restore the backup into an isolated test runtime.
  6. Start the test runtime and verify the expected world, records, and configuration.

Do not restore a test over a running production server. Keep the recovery commands and required credentials available independently of Enderdash.

Example script for a Paper server

This script archives the worlds, the plugin data, and the main configuration file. It then deletes archives that are older than 14 days. Change the paths and the world names to match your server.

backup.sh
#!/usr/bin/env bash
set -euo pipefail

SERVER_DIR=/srv/minecraft
BACKUP_DIR=/mnt/backups/minecraft
STAMP=$(date -u +%Y%m%dT%H%M%SZ)

mkdir -p "$BACKUP_DIR"
tar -czf "$BACKUP_DIR/server-$STAMP.tar.gz" -C "$SERVER_DIR" \
  world world_nether world_the_end plugins server.properties
find "$BACKUP_DIR" -name 'server-*.tar.gz' -mtime +14 -delete

An archive of a running world can be inconsistent. Before the script runs, send save-off and then save-all flush to the game console. After the script ends, send save-on. As an alternative, run the script while the server is stopped.

Keep BACKUP_DIR on a different disk or host from SERVER_DIR. The agent user must be able to read the server directory and write to the backup directory.

Schedule the tested script

Open Automations, choose New, and set Action to Shell command with the script's absolute path. Set its working directory, schedule, and timezone. Keep Retry policy at Never retry and set Offline catch-up to Skip missed jobs while offline until duplicate execution is safe.

Use Run now once. Confirm the agent's run history, then inspect the actual backup output. Monitor storage capacity, retention, and access to the destination through your backup tooling.

After an incident

Determine the intended recovery point before replacing data. Stop the affected workload where required, preserve evidence needed for diagnosis, and use the tested restore procedure.

Enderdash agent state and your application data are separate. Preserving /enderdash/data in an agent container does not back up game worlds or database volumes.

Next: scheduled jobs or file transfers.

Was this page helpful?

Send a quick note if anything is missing or unclear.

Last updated on