-
|
I've never been able to actually get analysis or any other tasks started -- My tasks just stay forever pending even if i cancel/retry i've triple checked my database and navidrome user/passwords and all ports are correct. I'm running this on unraid. 
`[INFO]-[04-02-2026 00-32-21]-Starting AudioMuse-AI Backend version v0.8.8
|
Beta Was this translation helpful? Give feedback.
Replies: 5 comments 14 replies
-
|
Hi and thanks for the feedback. Do you mean that the enqueued task with cron don't start OR they don't start at all even manually clicking on start analysis button? |
Beta Was this translation helpful? Give feedback.
-
|
Hi, It does not start at all even manually clicking on start analysis. re: worker containers -- I think this is the issue... I'm using the unraid template and perhaps it never downloaded/started any worker containers? |
Beta Was this translation helpful? Give feedback.
-
|
@ptichalouf suggestion? |
Beta Was this translation helpful? Give feedback.
-
|
Try deploy a second "main" container that has the |
Beta Was this translation helpful? Give feedback.
-
|
The worker need to see PostgreSQL for save the analysis result, Redis to get the task from the queue and the Music Server because it need to download the song. check one of the docker-compose example for your Music Server to read more: And here you can read more details about the architecture: https://github.com/NeptuneHub/AudioMuse-AI/blob/main/docs/ARCHITECTURE.md Who already deployed this on Unraid please if he can integrate/fix this example and raise a PR: if it doesn’t raise the worker there is an issue https://github.com/NeptuneHub/AudioMuse-AI/tree/main/deployment/unraid-templates |
Beta Was this translation helpful? Give feedback.
-
|
I was quick to say the unraid templates worked because I just checked that the webui popped up. I never actually tried to analyze! Sorry about that I'm trying to wrap my head around running the same container twice, with one as a worker, and how to get the two containers communicating. I'll read the posted examples. Thanks! |
Beta Was this translation helpful? Give feedback.
-
|
Look it's all very simple:
Now I don't have Unraid to learn how it work and fix. Bue please when you fix raise a PR on the documentation. |
Beta Was this translation helpful? Give feedback.
-
|
@s0ybean you can just deploy another version of the main audiomuse but instead of attaching to Then make all the other settings the same but at the end add the new env var I mentioned to get everything working |
Beta Was this translation helpful? Give feedback.
-
|
That fixed it. Thanks! I can now start and complete tasks. The second hurdle is now that the containers don't seem to be using my navidrome credentials (according to the logs the credentials are undefined). I made sure to put them in both the worker and main flask containers as variables. I wonder if this is another unraid quirk again or user error? There is a way to use environmental files with unraid containers, so I'll put my credentials there and give it a shot later today. |
Beta Was this translation helpful? Give feedback.
-
|
Small update: it is most definitely an unraid configuration issue. When I use docker-compose, I am able to analyze my library without issue. @darrynten solved the "no worker" issue, but I can't seem to figure out why my credentials aren't being set for navidrome. Maybe I will just stick with compose for now 😂 |
Beta Was this translation helpful? Give feedback.
-
|
I've just managed to make it work after reading this discussion (the documentation is a bit confusing for Unraid 😅).
|
Beta Was this translation helpful? Give feedback.
-
|
If you have the solution, please raise a PR to the deployment example / documentation so we can help also the other user. Thanks. |
Beta Was this translation helpful? Give feedback.
-
|
Is here someone who successfully deployed AudioMuse-AI on Unraid ? and che raise a PR and fix the example in deployment ? Otherwise I’m going to delete the deployment example. @ptichalouf you raised the PR for Unraid can you give a look ? Thanks. |
Beta Was this translation helpful? Give feedback.
-
|
Hey sorry for the late reply, I have little free time those days but should be able to take a look over the weekend. |
Beta Was this translation helpful? Give feedback.
-
|
People is continuing opening bug related to Unraid creating me extra work in something that I’m not able to support. |
Beta Was this translation helpful? Give feedback.
-
So sorry, but I do not know how to do a PR (never did it before). The good news is, I tested the unraid setup and have been running without issues for 36hr. My initial tests gave me memory issues where I would run out of RAM and the server becomes unresponsive. I limited memory to 4gb and it resolved the problem. This was probably user specific (I only have 16gb). Two things that need to be changed or noted for the unraid setup
**I did not come up with any of these fixes myself, but analysis has been functioning perfectly. |
Beta Was this translation helpful? Give feedback.
-
|
If you have the corrected file ready, that you already tested and they work, doing a PR is very simple and you can do all from web browser if you want:
|
Beta Was this translation helpful? Give feedback.
-
|
Hi and thanks all for the feedback, If someone would like to come back with an Unraid working deployment, please raise a new PR where at least another 2 Unraid user need to test that everything work and confirm in the PR. Meanwhile AudioMuse-AI remain a contanerized application that can run on different technology, but find how to deploy it on specific technology will be up to the single user (I can only give example and support on what I know, that is Docker Compose and K3S). Thanks everyone, |
Beta Was this translation helpful? Give feedback.
@s0ybean you can just deploy another version of the main audiomuse but instead of attaching to
bridgeyou choosecontainerand choose the existing audiomuseThen make all the other settings the same but at the end add the new env var I mentioned to get everything working