ffmpeg not restarting after crash

[paddatrapper]

Describe the bug

After ffmpeg exists 0, it is restarted, but exits 1. In this case, there is no attempt to restart again and earhorn continues to run without actually running ffmpeg. Logs are:

2023-03-12 14:06:38,703 | INFO     | earhorn.stream:listen:123 - ffmpeg command exited with 0
2023-03-12 14:06:46,945 | INFO     | earhorn.stream:listen:128 - stream listener stopped
2023-03-12 14:06:47,747 | ERROR    | earhorn.stream:check_stream:83 - could not stream from 'https://edge.iono.fm/xice/uctradio_live_medium.aac'
2023-03-12 14:06:53,229 | ERROR    | earhorn.stream:check_stream:83 - could not stream from 'https://edge.iono.fm/xice/uctradio_live_medium.aac'
2023-03-12 14:06:59,205 | INFO     | earhorn.stream:listen:96 - starting stream listener
2023-03-12 14:06:59,205 | INFO     | earhorn.stream:listen:101 - preparing ArchiveHandler 
2023-03-12 14:06:59,207 | INFO     | earhorn.stream_silence:parse_process_output:66 - starting to parse stdout
2023-03-12 14:07:00,377 | INFO     | earhorn.stream:listen:123 - ffmpeg command exited with 1

Environment

• Docker image: ghcr.io/jooola/earhorn:v0.17.0

[jooola]

Weird, the child threads handling the output of the parent 'listening thread' should be joined and automatically stop after the process exited.

I'll check if I can reproduce.

[paddatrapper]

Even exiting completely is a valid solution here - for me, it runs in Kubernetes, so gets restarted anyway. The issue is having the parent continuing, but not actually listening to the stream.

[jooola]

Could you check if the new release fixes your issue ?