SnapShooter Limits

SnapShooter is a cloud backup and recovery solution. Use SnapShooter to back up servers, volumes, databases, and applications from DigitalOcean and other cloud providers.


General SnapShooter Limits

  • Custom storage providers must support multi-part uploads.

  • S3 storage providers must respond to requests within 1000 milliseconds.

  • Based on DigitalOcean limits, you can have up to 25 snapshots per volume. SnapShooter makes sure you don’t schedule past this limit, and if you have snapshots we don’t know about, we notify you if a backup fails because of reaching the snapshot limit.

  • Based on DigitalOcean limits, you can take a volume backup every 10 minutes. This is not an issue unless you manually create a snapshot before a scheduled snapshot.

  • SnapShooter does not support pre- or post-backup scripts for AWS compute or serverless.

  • We do not plan to support Linode for native backups because their limit is a single backup, making it less effective for our users. We recommend using our file backups instead.

  • SnapShooter allows only 1 Agent connection per token. All other connections are dropped immediately.

Jailed Shell Hosting Limits

We support servers from different hosting providers to the best of our ability, but jailed shell hosting companies limit the terminal.

  • When you connect a server to SnapShooter, we do a scan in the background to detect if the server is running in a chroot. If so, some of our real-time logging is disabled. You can check this when looking at advanced server settings. When SnapShooter is running without real-time logs, it prints a warning with information on where to view the logs directly.

  • Some hosting providers run the SSH client in a special jailed session which limits the nohup command. In this case, when SnapShooter is running, the SSH connection times out but the backup continues to run. If we detect this, we mark the backup as started successfully and inform you of the error.

  • SnapShooter may fail to detect a backup that has stopped unexpectedly due to the hosting provider’s implementation of process IDs.