Direct answer

Protect the data first. Stop repeated failed saves, retain the original world and logs, and determine whether the failure is disk space, write access, a storage I/O error or a particular world/region operation. Do not delete region files as an initial troubleshooting step.

What it means

A save failure indicates some part of Minecraft, a plugin or a mod could not persist data. The first IOException, No space left on device, AccessDeniedException or I/O error in the log is more useful than a later generic shutdown or corruption message.

If you are a player

  • Tell staff the time, world/dimension and action when the issue occurred; avoid building or modifying items while persistence is uncertain.
  • Do not repeatedly log in to “test” whether progress is saving. Ask the admin to confirm a successful save and restart first.

If you run the server

  1. Preserve logs/latest.log and any storage/panel alerts. If Minecraft is still responsive, announce maintenance and stop it cleanly before file-level work. Do not force-kill unless absolutely necessary.
  2. Check available space, inodes, storage mounts and filesystem error logs on the server host. The df commands below are Linux-only and show different exhaustion conditions.
  3. Check the exact world directory and file listed in the first save exception; verify the service account can write there. Fix a full filesystem or failing disk before restarting.
  4. Create a verified backup/snapshot of all world dimensions, player data and configs. If the error identifies a damaged region/datapack/plugin, recover on a copy using a known-good backup; do not indiscriminately delete .mca files.
  5. On the test copy, run the server, save normally, stop cleanly, restart and confirm recent changes persisted. Only then return to production.
df -h
 df -i
# Replace this path with the actual world folder:
ls -ld /srv/minecraft/world
# On systems where journalctl is available, inspect storage errors:
journalctl -k --since "1 hour ago"
Data-loss warning

A new save may overwrite recoverable data. If a disk is failing, stop writing to it and recover from a trusted backup or snapshot; seek storage assistance before further repairs.

References

Use the documentation for the exact software and release you run. These sources support the settings or diagnostic behavior used in this fix.

Was this article helpful?
Help us improve this article

Tell us what is missing, unclear, outdated, or incorrect. You can also suggest something we should add.

What could we improve?
Do not include passwords, server IPs, or other private information.0 / 1000