Understanding Duplicacy

Support
#1

Hello I’ve have spend a lot of time the last couple of days trying to understand how Duplicacy works, so that I don’t mess up my backups by mistake. I have read the “Design and Information” how-to’s.

From what I understand the storage contains:

  • Chunks: The actual files split into pieces.
  • Snapshots: A snapshop represents an actual backup. The snapshot contains references to the chunks
  • Config-file: The config-file is needed for duplicacy to read the storage.

At my computer I need only the storage info and encryption keys I used when I configured the storage to restore a snapshot. For example: If I delete the c:\Users\username.duplicay-web folder, I can just recreate the folder.

I can delete individual shapshot files manually from the storage if some of them contains missing chunks. I can then delete the unreferenced chunks with prune with the - exhaustive option (as long as there is no backup running).

Is all this correct?

Two further questions

  1. What happens if I delete the config file from the storage?
  2. What happens when I restore af snapshot with missing chunks? How much data will I lose? Is it only one file per chunk or could it be several files for a single chunk.

Thank you
Benjamin

#2

This is correct.

You won’t be able to access the storage if the storage is encrypted. Otherwise you may be able to recreate the config file if you use all default settings.

It depends. Small files are usually packed into one chunk but a large file is split into multiple chunks.

1 Like
#3

A naming correction: “A revision represents an actual backup”.

A brief explanation about the nomenclature used by Duplicacy:

  • repository is the location/folder of the source/original files;
  • snapshot is associated with the repository;
  • storage is where backups will be saved;
  • each time a backup is performed, a revision is generated;
2 Likes
#5

It depends. Small files are usually packed into one chunk but a large file is split into multiple chunks.

Ahh okay, so if a chunk is missing I will still be able to retrieve all files except the ones packed into the chunk?

#6

And the more elaborate explanation is now here:

1 Like