Thank you for your donation!


Cloudsmith graciously provides open-source package management and distribution for our project.


Problem: Auto-shuffle not working on Raspberry Zero with music library >33.000 flac files
#1
I recently found out that Random play is not possible (does not happen anything) on a Raspberry Zero and Raspberry B+ with my music collection of 33.000 flac files. This behaviour occurs with flac files from the Nas as well as with a harddisk directly connected. The library shows all the music.

On a Raspberry 3 random play works without a problem.

This behaviour occurs in version 4.2 and 4.3.

Any ideas?
Reply
#2
Hi,
I have the same problem (30K plus mp3's) although Tim Curtis refers to the how and what to set it up: IT JUST doesnt work on the Zero
Reply
#3
On a 3B and 3B+ it does work, believe me, at least if you followed the correct procedure...
Just give it some time. It can take up more than a minute before a song is selected and played...
The bigger your library, the longer it takes.

Johan
 :@
Reply
#4
Run the command

Code:
./sysmon.php

and see how much RAM is being consumed.
Enjoy the Music!
moodeaudio.org | Mastodon Feed | GitHub
Reply
#5
(12-10-2018, 02:01 PM)Tim Curtis Wrote: Run the command

Code:
./sysmon.php

and see how much RAM is being consumed.

@JST1963 I have got several Raspberry 3B's and auto shuffle does work here without a problem. Doing the same with a Raspberry Zero after 10 minutes still no music.

@Tim Curtis This is the outcome of the command: Ram_Used 36%
Reply
#6
It works fine on my Pi-ZeroW. The collection I use for testing is ~ 4000 tracks on a NAS.

Assuming the ashuffle process has not crashed, and given that RAM usage looks OK in your case then troubleshooting is a matter of trying to find the bottleneck.

First make sure that ashuffle is running

Code:
pgrep -l ashuffle


Then try the command below to see if a process or thread is hogging the CPU


Code:
top -H
Enjoy the Music!
moodeaudio.org | Mastodon Feed | GitHub
Reply
#7
(12-11-2018, 05:25 PM)Tim Curtis Wrote: It works fine on my Pi-ZeroW. The collection I use for testing is ~ 4000 tracks on a NAS.

Assuming the ashuffle process has not crashed, and given that RAM usage looks OK in your case then troubleshooting is a matter of trying to find the bottleneck.

First make sure that ashuffle is running

Code:
pgrep -l ashuffle


Then try the command below to see if a process or thread is hogging the CPU


Code:
top -H

Hello Tim,

My Pi-ZeroW also works fine with the number of tracks you use for testing. The problem occurs with the large number of tracks mentioned.

I tried your commando and see no problems but atteched a screenshot.

Thank you for all your good work.

Kind regards,

Frank


Attached Files Thumbnail(s)
   
Reply
#8
Hi Frank,

If I could reproduce the issue I could help troubleshoot but it looks like it requires a large collection.

In the absence of any resource hogging, I-O bottleneck or crashing of ashuffle the symptom would suggests a bug in ashuffle itself.
Enjoy the Music!
moodeaudio.org | Mastodon Feed | GitHub
Reply
#9
(12-11-2018, 07:48 PM)Tim Curtis Wrote: Hi Frank,

If I could reproduce the issue I could help troubleshoot but it looks like it requires a large collection.

In the absence of any resource hogging, I-O bottleneck or crashing of ashuffle the symptom would suggests a bug in ashuffle itself.

Hi Tim,

Strange thing is that it doesnot happen with the PI3 models with the same music collection.
Reply
#10
(12-11-2018, 07:59 PM)rylfas Wrote:
(12-11-2018, 07:48 PM)Tim Curtis Wrote: Hi Frank,

If I could reproduce the issue I could help troubleshoot but it looks like it requires a large collection.

In the absence of any resource hogging, I-O bottleneck or crashing of ashuffle the symptom would suggests a bug in ashuffle itself.

Hi Tim,

Strange thing is that it doesnot happen with the PI3 models with the same music collection.

Well, the Pi Zero and the Pi 3 do have different CPUs supporting different instruction sets. I wonder if the behavior you observe is the result of some difference between the code and/or libraries compiled for them.

Regards,
Kent
Reply


Forum Jump: