Does every PeerTube instance store all other instances’ content metadata (title, description, comments)? Would federating virtually with YouTube (through a YouTube frontend like Piped) give a too high ammount of data to store on other instances?
You must log in or register to comment.
You don’t need to do that. PeerTube already has the ability to mirror YouTube channels, and yes, it would be far too much data. But you can do a personal backup of YT channels on your storage if ya really want to.
It would have been great to view and comment YouTube content with a PeerTube account. But it would needs a different approach on storing other instances metadata.
As of June 2022, more than 500 hours of video were uploaded to YouTube every minute. If that video is stored at 6Mb/s, a conservative estimate, that equates to about 10TB a minute. Obviously you wouldn’t want everything, but it’s still way more than the peertube network could handle in it’s current form I’m sure. I cannot fathom the amount of data that Big Datatm handles, the fact that most of the services are free makes me highly suspicious.
