General question about the current configuration of my backups

Support
#1

Hello everyone,

I became aware of duplicacy 2 days ago and would like to replace my existing duplicati backup with it.

However, I have a few questions before I start the replacement.

  1. currently VMs on my server are backed up every night and stored as a packed image in a folder. Only the last 5 zip files are always kept (all older ones are automatically deleted). This backup is performed automatically by my server.

  2. the same applies to my docker systems

  3. I have a folder with private files, which is not available as a .zip but as an entire folder structure.

What about the duplicacy function? Are these folders mirrored? does this mean that if the folder is deleted in the source, it is also deleted in the target or are all files backed up and this again daily? i would prefer a mirroring. is this possible with duplicacy? so that even if my server deletes folder a, it is also deleted in backup xyz?

many thanks!

#2

What issue are you experiencing that you are trying to fix by an entire different backup solution, let alone the one you’ve just learned about?

  1. It’s best to backup content of the VM, not the whole image. And if you do want to backup the whole image — don’t compress it beforehand, to take advantage of deduplication, and save space. You may also want to adjust/fix chunk size, or even consider Vertical Backup solution from the same author, that is optimized for VM backup.
  2. Docker containers shall not be backed up at all. They are disposable entities. You want to backup the data containers operate on.
  3. Great.

What about it?

No. Mirroring is not a backup.

Duplicacy creates space efficient history of snapshots of the state of your data in time. If file is deleted on the source, it will not be present in the future snapshots, but it will remain in the previous ones, so you can restore it. That’s the point of backup, and pretty much any backup tool is designed to do this.

Frequency or backup and retention is up to you.

Mirroring provides just one version. If your data is corrupted on the source, it’s going to be also corrupted on the target. What’s the point of such “backup”? If you want to mirror data you can use rsync or rclone.

Correct. Every backup revision will reflect state of your data at the time of backup.

1 Like