Regarding the best use cases for tardigrade + duplicati

In the documentation for setting up tardigrade with duplicati it says

"To restore a small file, Duplicati has to download the entire block it is contained in. Therefore, the best fit for Duplicati are the following two backup use cases:

  1. when none of the files contained in the same block need to be ever restored again in the future.
  2. when all files in a block need to be restored at the same time."

Would someone be so nice as to explain this to me in laymans terms? :slight_smile:

My need is your normal personal backup situation. Backup images, documents, videos and so forth. I will mainly use it as “cold storage”, meaning mostly uploading, but also downloading from time to time if I for example have alot of stuff on tardigrade which doesnt fit on my computer and need to download to work on it.

Is duplicati for me then? :slight_smile:

1 Like

Rclone or Filezilla might be a better choice for you. How many files? How big is your backup in total? How much of that is in text files vs images and videos? How frequently do you want to download some files? Which files are you going to download? Best case would be the video files. Worst case you download small text files.

Duplicati will reduce the costs of the backup with compression, deduplication and packing small files into bigger blocks. The downside is when you download a smal file that will create 64MB download traffic. You could simply upload 5MB blocks. That will increase the number of files and you might pay maybe a few cent extra for that but it would reduce the download overhead. I hope this explains the downside of Duplicati.

Thanks for the quick reply! My scenario would e.g. be 20-50GB of images mostly for cold storage. Other than that I dont need to backup too much other stuff than small encrypted documents and lets say a few GB’s of videos which will also mostly be in cold storage. Yes and also 10-30GB of music will also be in cold storage.

So I guess long story short, mostly cold storage :wink:

If duplicati is the worst option for downloading small text files, but best for the other purposes, this doesnt necessarily mean that I wouldn’t go for duplicati since I wont be downloading text files very often (like, not even once a week/month).

In my case my backup folder was 350GB with 70K files in it. I had to pay 10 cent for object count per month. Duplicati also found some duplicates and the remote backup was 70 GB smaller at the end. But lets ignore that part for a moment because I could also clean up my source folder to get nearly the same effect.

So Duplicati is saving me at least $ 1.2 per year. Download a small file will cost me $0.004. In my case I could download 300 small files per year and Duplicati would still be cheaper then other tools. If I download more then 300 small files per year I should switch to some other tool.

Great, this makes sense to me!

I have read countless explanations on the tardigrade site etc. about up/down, objects, different uses and so forth. Still I havent found one that explains it reeeally simple for users who would like to use it for personal backup instead of dropbox. I don’t know if you are affiliated with StorJ but I think that it would be a good idea to make this as easy to understand as possible. Maybe without even mentioning technical terms, even though you will meet them when you sign up for your account etc. Maybe even a video explaining the different ways to go would be a good idea.

I’ll go for duplicati then I guess, looks alot nicer graphically than FileZilla anyways.

It would be nice if you could help us with that. What is your opinion on our documentation? https://documentation.tardigrade.io/how-tos/backup-with-duplicati

We are working on a video. Ideally the video is just a copy of the documentation. If the video contains important information we should also note them down in the documentation. Let us know what is missing in the documentation and we will improve it.

I haven’t gone through the steps yet, but I think it looks clear and easy to follow. The thing I meant when I wrote “as easy as possible” was more how non-programmers could get an easier explanation as to which option to go for, I wasn’t necessarily critiquing the documentations as to how to get it up and running, if that makes sense?

A “troubleshooting” section at the end might be a good addition though.