I am running a Seafile docker image on an Ubuntu 20.04. Cloud Server.
I have several users working with the seafile instance and have a backup routine (using borg backup) which stores the complete seafile-data directory and the seafile db to a remote server.
Now it turned out I could use a special user account for large amounts of data which I don’t want to be backupped as it would possibly clog my entire backup routine.
Is there a way of separating the data of certain users or libraries from the others so I could adjust my backup routine for ex. in a way to ignore certain directories for ex.?
You could simply exclude the respective libraries via the library ids (blocks, fs and commits directories) in your backup routine.
Or is there something against it?
Hi there, thanks for the input.
What do you mean exactly? My backup routine works with a simple shell script that does
- export Seafile DB
- backup seafile-data, db and config files via Borg
As Seafile stores all data in a cryptic manner I wouldn’t know how to decide via library IDs…
You can find out the library id using the web interface. If you open the library in the web interface, the url is something like:
From that, you can get the ID. The data is then stored in subfolders for each library. So, by excluding the respective id subfolders in the blocks, fs and commits directories (as encryptomator said) you can exclude a library from the backup. I don’t know how Borg works so I can’t tell you how to exclude directories. The mentioned directories are inside the seafile-data folder.
Hey … well yeah that looks good. Didn’t know about the library IDs!
Thanks a lot that should do it!