Clarification on Centralized management (I might have misunderstood)

Before you post: have you tried searching the forum?yes quick search, If you have found related posts that didn’t quite solve your issue, please mention (link) them in your post. did not.

Can someone please correct me if I have misunderstood the Duplicacy web-page.
To me I read it as you can install Duplicacy web GUI on one central server and then you have access to your different machine’s backup from that web GUI. In my case a Opensuse machine and a Mac.

to get the cross dedupe to functin. Do I have this wrong?

Right now I have two seperate web GUI’s one for OSTW and one for Mac, but the both share the same /nfs/storage location.

I’m a long time borg and kopia user and I want to migrate to a centralized backup solution.

Let’s separate few things.

  • Duplicacy supports cross-machine dedupe. To accomplish this, all you need to do is target the same storage location.

  • it’s entirely client side. There is no server component. It just needs storage.

  • the web ui is a a front end that runs duplicacy CLI on the system it’s installed. You can access web ui from elsewhere. So technically, you can access all of them from the same location but i would not call that “centralized” — all duplicacy instances are independent.

Yes, you can install it on third machine and have access to all other machines backups, regardless of whether they are configured to target the same or different storage by adding the same storage locations. You can check, prune, restore etc. however you cannot control the backup process - you cannot change anything configured on machine A from machine C. So I would not call it a “central server” because there is nothing central about it. It’s just another machine running duplicacy and configured to access the same storage.

Worth noting, when both machine A and B backup to the same storage they will see each others backups - because the encryption password is the same and they both will by necessity have it. If this is not desirable you can enable RSA encryption in duplicacy and have admin keep the private key. Then A and B will use public key to backup but not be able to restore and thus see each other’s data.

Thank you for the clarification, I have clearly not understood what I was reading but now I do.
I’ve been playing around with Duplicacy the last day or so, and it looks very promising.
Now i’m trying to figure out if you can search/list a single file and restore from one of the “ -r “
I see it’s been discussed in a couple of other threads a few years ago and it was said “not yet available”

Sure you can restore a single file. It would have been silly if that wasn’t possible :slight_smile:

You would not happen to have a good tutorial on how to do it through the cli? I see I can do it from the web-gui

That would be a very boring one-step tutorial. Here is the wiki page: restore · gilbertchen/duplicacy Wiki · GitHub

To restore file Documents/text.txt to revision 3 you do

 duplicacy restore -r 3 -- "Documents/text.txt"

This will restore the file in-place. You can specify multiple patterns and they can include wildcards — just like backup filters.

If you don’t want to restore it in-place, you would need to initialize a new duplicacy repository in some empty folder with the same snapshot ID and storage, and run that command there. (That’s what webui is doing)

Am I missing something, or is it the case that unlike the backup command there is no way to test filters when running restore?

For a case like this where you want to restore just 1 (or a few files) it would be a real bummer to misformulate or just mis-type the file or pattern matching, and then end up restoring the entire revision.

EDIT: I see this was suggested 7 years ago: -dry-run option in restore

1 Like

Yep, afaik that’s the case.

thank you, that was what I was wondering about, I just did not understand the documentation, that ```
[--] [pattern]
is what I wanted to achieve. I would have loved it if the docs had some actual examples like you did here.

last question for now, is there a way to brows the snapshots, to find the file or dir that you are missing (forgot the absolute path) ?
thank you for your help.

You can always use wildcards, or see list · gilbertchen/duplicacy Wiki · GitHub

e.g.

duplicacy list -r 3 -files | grep "text.txt"

Some time back I’ve asked Google Gemini to write a UI wrapper for duplicacy to provide that functionality – have a look, it support search and restore of individual files, and even diffing files from different revisions: