This seems like a common question yet I searched and didn’t find anything? Is there a way to set the max number of backups it saves? Seems like a option that should be easily accessible. As if I back up weekly or daily I imagine pretty quickly the space used would get huge, especially backing up Plex. Do I have to manually delete older backups?
Why would it become huge, let alone backing up plex? If by plex you mean media files – then space won’t change at all, because media files don’t ever change. So every backup after the first one will occupy zero additional space. If you backup plex database – same deal. it will only keep backing up difference from the last state, so it will also grow imperceptibly slowly.
In a way, it is, its’s just “number of backups” is not as useful reference as time interval based backup cadence. People don’t think “I need to restore backup 281 backups back”. People think “I need to restore backup from first week of last month in the evening on Friday”. Whether this is backup number 313 or 45 is not as important.
Yet, if you backup daily, it’s trivial to limit number of days. For example: prune -a -keep 0:200 will only keep 200 days worth of backups (“Keep 0 backups older than 200 days” is that that parameter means")
My Plex metadata folder is over 500GB so therefore if I did once a week Id have 2tb occupied in a month. Are there incremental backups? Is that a possibility to just update daily any changes instead of making a whole new backup? Where do I add this " prune -a -keep 0:200" I’m a newb so be patient with me 
This is not true. Does your entire metadata change every week? No, it is fetched/generated, and never changes. So your first backup will he 500gb, and the next 20k, 25k and maybe 10k. Depending on what changed.
This is the only way how it works. Each duplicacy backup behaves as a full independent backup but it’s actually created incrementally.
Read about “content addressable storage” — each piece of data is assigned a name derived from its content. If the piece already exists on the storage it gets referenced in subsequent backups instead of being uploaded again, so every unique piece only occupies space once.
(A separate question is why do you backup plex metadata in the first place - it’s derivative useless data… but maybe if you customize everying it may make sense. Myself I don’t backup plex at all. Neither media, nor metadata.)
Ah then that basically is why number of backups doesn’t matter. And as for your question yes I have done a ton a modifications and have had a boo boo that caused me to loose it all and have to start over in the past. What exactly do you use duplicacy for then? Im using it for appdata and plex.
Precisely! Think about it this way – duplicacy combines all of your files into one long sausage and then shreds them at some “convenient” locations. Those location are usually quite stable so even if you change some files, add some fies, and remove some files – teh majority of cuts is still done in the same places, resulting in the same chunks, that are alredy present on the storage. The snapshot file is pretty much just a list of files with reference to which chunks need to be used in what order to reconstruct them. (In reality the snapshoit fles are also shredded to chunks because they too can get quite large)
yeah, then it does make sense to back it up! I never had patiencce to orgianize my plex library. I just pile stuff over and hope for the best the crawlerst and parsers to do the right thing.
My documents mostly. And very selective subset of files from ~/Library: such as some application settings, Mail, and Messages,(analogous to windows’s APPDATA on macOS). On a mac there is a mechanism for application developers to tell whdich data does not need to be backed up – and some well witten apps actually use that. On Windows apps are less orgianized and it’s much harder to tell what to backup and what to skip – so perhaps a good approach on windows is to backup the whole thing by default, then watch few backups, and manually exclude obviously transient trash – like browser caches, game shaders (some games I remmember had a brilliant idea (or rather, were reckless enough) to dump machine specific shaders into My Document folder that syncs to OneDrive)
I dont’ use duplciacy to backup media itself – instead I copy them to amazon glacieri deep archive storage. media files don’t change (no need for versioning), theuy are not compressible (no need for compression) and they are unique (no need for deduplication) – so just copy without overwrite works fine.
that way duplicacy backup stays very small – essentially only hot document and some application data
To me cloud storage has just gotten way to expensive, I used to use dropbox’s unlimited business account.They ended that and basically said sucks to be you pay this ridiculous amount or well delete your stuff. so Id rather backup essential data in a backup to make catastrophes not so much. Media can always be redownloaded the metadata is much more personal time and work. Thanks for all your help.
Sure. That’s why I specifically use Glacier Deep Archive - it’s cheap to store, but very expensive to restore – which is counteracted by the probability of ever needing to do so being close to 0%.
My Duplicacy backup goes to STORJ and replicated to Glacier (Duplicacy does not support Glacier due to thawing requirements – but some folks here had good success with Google Archive tier. Beware minimum retention charge of 365 days) – it’s a small subset so paying for hot storage is OK, it’s a few bucks a month.
Main backup though is on a ZFS array at home, that is replicated to another ZFS array at my brother’s house 1000 miles away. So yeah. Data is priceless, and “someone else’s computer” is usually more expensive.
If a series of events will lead to data loss at my house, my brother’s house, decentralized storage (STORJ), and Amazon tapes – we are probably going to have a much bigger problem to deal with as a humanity than restoring my cat photos.
There is no such thing as unlimted anything. My prior rant on the topic: (Newbie) Backup entire machine? - #6 by saspus
Hey sorry to restart this thread but I was just wondering since there are multiple options. When you set up a schedule and you “add a job” it has multiple options “backup, copy, check, prune” the default is “backup” which I left it on. This option does the incremental backups correct? It’s about half done backing up Plex. Or do I choose copy?
Duplicacy web ui is a simple front end to duplicacy CLI. You can find descriptions of the commands and what they do here: Duplicacy User Guide (Updated)
TLDR: backup creates versioned history of repository on the target storage.
Copy replicates the versioned history between storages.