Replies: 6 comments 9 replies
|
Hi and thanks for the feedback. About your point I can clearly say that:
About the setup config and credential lost I need to check. Do you have the possibility to share the full log from flask and worker container ? Which version of AudioMuse-AI do you have ? Edit: meanwhile I think that the new credential was empty because in the provider migration you don’t specify them. The functionality is usually needed to migrate from one provider to a different one.
could be this the case ? |
|
For all the procedure, as alredy wrote a while ago is to do a provider migration Navidrome to Navidrome.
Doing this you will have ID of AudioMuse-AI that doesn’t match the ID of Navidrome. At this point you had to do a provider migration on AudoMuse-AI form Navidrome to Navidrome. BE sure that on Navidrome you have enabled the real path (so in Navidrome, in the player list, there is the possibility to show to the player the real path of the song, this is needed). After the provider migration do another analysis it will be needed to recreate all the index and in case some track where not migrated successfully. Important: during the provider migration put the credential of Navidrome again. This because the functionality was initially designed to migrate to a NEW provider. So you need to put even if you’re migrating versus the same server. |
|
I did the migration it matched every single track based on path, But now I'm doing a rescan and it's reanalysing lots of tracks. It looks like the migration did some sort of dedupe and now it's rescanning all of the tracks that had dupes. Like I check my duplicates list and it's way tinier than it was before with the max song dupe count of 3, when I know I had some in the teens. In my backup I had 76,000 rows in track_server_map and now I only have 68,000, so it's redownloading and scanning the missing 8000, which all seem to be duplicates |
|
I updated Navidrome to .64 over the weekend but I did not update AudioMuse from 3.4.0 because I’m dumb and forgot. I also have AudioMuse set to do an analysis over night. I woke up this morning to AudioMuse reanalyzing my library, which is because of the Navidrome ID change. And that’s cool, I’m fine just letting it run through again. My question, will I need to do anything after the new analysis is done or will Audiomuse remove the old file analysis when it’s done? Hopefully that makes sense and thank you for a super cool program! |
|
This went very smoothly for me. Updated Audiomuse-AI plugin and server to the latest versions, backed up both databases, updated Navidrome, enabled Real Paths and then began the migration. I ran a full scan and then a database cleanup. Very nice. I had a handful of albums not migrate, no big deal. Thank you for your work! |

Uh oh!
There was an error while loading. Please reload this page.
I just completed the Navidrome 0.64 ID migration with AudioMuse-AI. Overall the provider migration worked, but I ran into a few issues worth reporting:
artist_server_mapwas empty and noserver_sweeptask had run. As a result,/api/similar_artistsreturned 404. Running a server alignment sweep rebuilt artist mappings, after which similar artists returned 200 and worked normally.So the migration itself seems solid, but I think the post-migration setup/config preservation and automatic server sweep need some attention.
All reactions