@LoudLemur speaking generally (and not in the context of selfhosting), the best way apps can and should interact with each other is via API.

For example, if you drop a file somewhere in nextcloud data directory, it won't pick it up (and for good reason). Unless the app is built around the concept of picking up files from filesystem, this system (around volumes) doesn't work. This is why while we have made it possible to share files via volumes in apps, we haven't invested much energy on this. Most apps have their filesystem layout (say lychee, jellyfin, castopod) and even though they store files in filesystem, it's really like a database (i.e opaque) and it's best not to move things around.

@rmdes

There is this option too, but I think it is probably necessary to know about this in advance and plan accordingly. I don't know if it could be implemented retrospectively:

https://forum.cloudron.io/post/86912

@Kubernetes thanks for reporting. Fixed.

Their demo has 512MB atleast - https://www.navidrome.org/demo/ . So, I will bump up the memory to atleast that.

This is now fixed in https://github.com/navidrome/navidrome/issues/202

@jdaviescoates jealous ! but good to know 👍
I will crack on. Has to be resolvable.

@jdaviescoates Sorry, I meant to say "Does music fail to play in subsonic apps?" 🙂 .

Not sure what the logs mean. Maybe you can ask the navidrome project what the error means if everything is working. If you can link here, I can follow up accordingly.

Usually, the HTTP code 500 implies some sort of docker error. I think what you did was pretty much what one can do, reboot the server or restart docker.

@privsec
No, this app has its own user management.

@ruihildt just trying that now, thanks

@nebulon thanks. Seems like you need to keep the browser web terminal window open until the process finishes too? I guess perhaps add something about that too?

@jdaviescoates Yes, looks like there is some fd leak and they are nearing a fix.

Resolution for the "issue" if someone els has it:

Stop Navidrome, remove the cache folder (found in the configured DataFolder) and start it again

@girish said in After update to 1.0.0, I have to create the admin user again:

@necrevistonnezr It's only a browser caching issue. The admin user is still there, it just shows the wrong UI because of some service worker issue.

Thanks, indeed! Please mark as solved (can I do that on my own on mobile? Can’t find a button...)