I have 3 servers at 3 disparate locations. I know syncing metadata is not a resonable thing, but it would be great if when I edited one server, if those manual changes could be sync’d across all my servers if the servers were identical? It’s a little tiring editing posters and sort names in 3 different locations, especially when I do sweeping changes to things like file structure. They all pull the metadata back in fine after the folder structure/filename changes proliferate to the servers, but this would help manage the servers a lot.
While you’re coming at it from a different angle, your suggestion seems to attempt achieving very much the same already described in this existing feature suggestion. While I get your point about not syncing metadata in the first place… without that… how should Plex even know you have the same stuff on all those servers?
Probably a silly question… what’s the exact use case here and why not simply stream your media remotely from one server (avoiding all the hassle of cloning manual changes)?
I guess initially it would be on the server owner to know that their media servers were the same. When I envision it, I picture one being the master server where all the editing would take place. Then the other servers in the same account would have the media directories mapped 1:1 Like how you have to map directories in Sonarr/Radarr:
Server 1:/path/to/Movies=Server 2:/another/path/to/Movies
and
Server 1:/path/to/Movies=Server 3:/even/another/path/to/Movies
Mapped directories and filenames would have to match exactly, or no change to the metadata should take place. Directories and filenames are all handled in the background by the backup app running sync between the servers. So everything does always match. I haven’t put a lot of thought into it because I know a lot of feature request are white noise.
Syncing the DB seems like a good idea on the surface, but really how many changes are you making that often where you need a full database refresh? Syncing even a full day of just media edits seems trivial in comparison; as I stated, right now, backup processes handle the heavy lifting for filenames/directory sync. I just need (would like) my clicks and manual editing sent to other servers at the same time.
That said, if I read the article you linked right and I can simply backup the DB from one location to another, that seems like a possible way to do it using active sync backups, but I wonder how an active PMS media streaming session (not editing) would handle a full DB replacement on the fly? Maybe I could just make that happen one time a day instead of real time. I am the only one that does media editing so no one would be doing editing when DB sync occurs. It would be realively simple it seems to snapshot the DB then write it off to another location. If this is a valid path of thought let me know and I’ll persue it. I have plenty of DB backups available and unless I decided to blow up all 3 locations for some stupid reason, there’s always one live one I can always replicate.
To answer your last question, I originally used the alternate servers as backup locations for my data as my friends got fiber. So they are the download points (my WAN is much slower), then it automatically proliferates to my home. Then I started thinking I could just run PMS at the fastest location. Then I thought why not run it at all locations, allowing everyone access to all servers, then if one is offline or busy with too many transcodes or something, they could just start media from another server. So I guess the short answer is it’s just grown over time into what it is now. It would really help all this if there was just a central client server that offloaded requests to multiple servers based on current average server CPU/network load too, but that’s for a seperate thread. :\
2023 clean-up: considered to be a duplicate
