see on GitHub

Backup & Restore: Frequently Asked Questions


FAQ

  • Q: Is there a one-time backup?
  • A: No. Backup is an on-going task, and is meant to back your data up continuously.
    • If you wish, you can trigger it for immediate execution and then disable the task.
    • You can also use Smuggler as an equivalent of a full backup for a single export operation.

  • Q: How do I create a backup of the whole cluster?
  • A: Backup and restore your database group's data.
    The cluster and its information regarding the database and nodes can be easily re-created, there's no need for a backup for this.

  • Q: How should I set nodes' time?
  • A: The Backup task always uses the server's local time.
    It is recommended that you set all nodes to the same time. This way, backup files' time-signatures are consistent even when the backups are created by different nodes.


  • Q: Can't I simply copy the databases directory whenever I need to create a backup?
  • A: Simply copying the databases folder is not a good substitute for RavenDB's backup procedures.
    Creating a backup routine is a one-time operation. There really is no reason to do it manually again and again.
    RavenDB's own backup also has a few advantages. Among them:
    • RavenDB creates a reliable point-in-time freeze of backed-up data.
    • RavenDB ensures backed ACIDity, preventing any dependencies upon files or connections during restoration.

  • Q: Does RavenDB automatically delete old backups?
  • A: No, RavenDB does not automatically remove backup files, you need to take care of it yourself.
    You can use services like crontab (a linux scheduling procedure) to create an old-backup-files-removal routine.

  • Q: Are there any locations in which I better not store backup files?
  • A: It is recommended not to store your backups on the same drive as your database's data files, to avoid self inflicted scenario of a database storage distress because of backups piling on.