Server Version#: 4.76.1 / 1.27.1.5916
Player Version#: N/A
I’m having a problem with movies added to my Plex server. It started at about the time I updated the database (something the Plex Management Console had been bugging me about for a few months).
There’s an example of the problem. I added a new movie that was properly matched. After I saw it, I thought that it was something that I might have under a different name, so I tried checking by cast. NONE of the cast members were shown as being in any other movies. In desperation, I even checked John Rhys-Davies. Strangely enough, in a movie I’ve had longer, he shows as being in 20 movies I have.
I’ve found this problem with several actors in other movies I’ve added recently. I believe that the old names and new names are being handled differently, causing this issue.
My current best guess is that I need to rebuild the DB. I’ve seen that this can be done by removing and readding the library, but I’ve put far too many hours into filling in the empty data in my DB to be willing to repeat it all.
I need a way to use the data I have to reconstruct the DB.
I do not think so, but I do not have any movies that needed manual matching.
I refreshed the metadata for my movie library. It did not change any custom posters I had selected. The number of duplicates / unmatched items stayed the same.
Edit: It also did not change the custom Sort Title I use so certain movie series display in a preferred order (ex: Star Wars listed by Episode 1, 2, etc, not release date).
Will that “rematch” any of the movies? I think not, just checking.
I don’t know if it was a rematch, but during a test it merged the Extended and non-extended versions of “The Lord of the Rings: The Fellowship of the Ring”. I tried to split them back up and it won’t do what’s needed.
Is there a better choice available? Spending days or weeks fixing what gets messed up isn’t high on my “wish list”.
I’m not aware of another way to resolve the problem.
The problem is the actor is listed with two different IDs in the Plex database. The only way I know to correct things is to refresh the metadata. There is probably some way to fix things via SQL queries & edits. However, that is a) well out of my skill set, and b) seems very labor intensive.
Plex should let you split apart the movies.
Curious, how did you differentiate between the extended & regular release? Edit the Title field for the extended release? See this post, discussing different versions of a movie. That should survive a metadata refresh.
Your correct. The new modern plex agent & scanner stores GUIDs for each movie and these are unique to each movie. So different versions - director cut, colour, bw, extended can no longer be split apart and renamed like in the past.
This is because once they are split plex will rematch them again re-adding the GUID upon a library scan or refresh.
The only solution I’m aware of if you want separate entries is to have them in different libraries. Not ideal but would work.
On the actor issue you reported, it’s somewhat similar. Plex changed the ids they use for actors a while ago so a full refresh and optimise of db is needed to update these across your libraries. Once all are updated and using the same agent/scanner then plex will search correctly. Until that’s is done you could get mixed results.
I not 100% sure what you mean but reading your op it appears to be cast members not found in different movies. If so this is because of the actor id charge I outlined above and a full refresh of metadata of all your libraries with db optimise is needed to update and fix.
If it’s something else then I’m sorry for not understanding it and maybe detailed example will help.
So, let me get this right. The developers made a change to the DB schema (the recent DB upgrade) which made the old data incompatible with the new data, but they chose to not create a tool to seamlessly convert the old data to the new format? If this were a commercial application, that would be completely unacceptable. Wait a minute. This IS a commercial application.
This change, the lack of detailed information on it, and the lackluster response of the developers, is a clear indication that they have absolutely no concern for their customers.
Just to be clear, I did perform the above suggested Metadata Refresh, but I did it a few movies at a time. I encountered three of four sets of movies that got mangled by this refresh and spent about four hours getting them straightened out (I’m ignoring the time I spent updating the metadata in groups). Overall, I expect that customers spent a total of not less than 10,000 hours correcting their movie DB after this update rolled over their setup. Considering that the code for a complete solution would have taken under 100 hours to create, I find it reprehensible that the developers wouldn’t have created and released such a tool.
If I had been paying by the month or year, that DB update would have been the last of my money they received.
Incorrect, its fairly trivial to fix it, all you need to do is a metadata refresh. Ive done it, and instructed others how to do it with fairly little issue.
As stated above, a full metadata refresh can cause damage to a user’s library that can require considerable time to repair. This can be contrasted to a dedicated single-use tool that would avoid damage.
As for “fairly little issue”, if it requires an hour of time to fix what the refresh broke (it took me considerably longer), and 1,000 users are impacted by this (I expect there were many more), what is the total customer time that was wasted fixing what a dedicated tool would have fixed? This is a question of Customer Service. Based on this DB upgrade, Plex doesn’t care about Customer Service.
In my experience this is only true when media does not follow the plex rules for naming.
Can you give an example, of 1 movie with this problem. Provide full naming of how the file and folder is named on disk along with the library configuration (agent/scanner especially). Also, provide the xml for the movie in plex so we can review.
Nope. I’ve fixed the problems that were created by the lack of a proper DB maintenance tool to be used as part of the DB update. Any information that’s not in this thread has been long since forgotten.
