backup system

Most often: crash or update corruption, broken chunks, accidental deletions during maintenance, grief that slips past protections, and player data problems like missing inventories or ender chest contents. It is also the main safety net for big mistakes, like a bad config change, as long as there is a clean snapshot from before it happened.

Usually whatever happened since the last backup. That could be minutes or hours depending on the server. Expect recent mining, crafting, trades, and building to be the first things at risk, especially on busy survival servers where a restore affects many people at once.

Sometimes. Some servers can restore individual player data or rebuild a specific area from an earlier save, which reduces collateral damage. Other times the only clean fix is a wider restore. Ask what they do in grief cases and data-loss tickets, not just whether they have backups.

No. Claims, logging, and moderation are prevention. Backups are what lets a server undo damage after prevention fails, whether that means restoring a region or reverting to a point before the grief happened.

How often they back up, how long they keep old snapshots, whether player data is backed up separately from the world, and what their restore policy is when a rollback would impact other players. Clear answers beat vague reassurance.