[Solved] My upmpdcli Subsonic getting timeout for no apparent reason

[Lenzelot]

Hello,

I have to ask for help since I am currently out of ideas. mpd-alsa:latest is still working fine and upmpdcli-docker:edge with Subsonic plugin used to work fine a couple of weeks ago but now it's broken and I have no idea why.

There seems to be a timeout when connecting to my subsonic server. So the OpenHome/UPnP media server does show up but nothing comes up when browsing.

upmpdcli  | :3:../src/mediaserver/cdplugins/cmdtalk.cpp:140::CmdTalk:readDataElement: fatal timeout (60 S)
upmpdcli  | :2:../src/mediaserver/cdplugins/plgwithslave.cxx:611::PlgWithSlave::browse: slave failure

But using curl inside the docker it can connect just fine docker compose exec -u 1000 upmpdcli curl https://my-navidrome.server.com/app.

Did I mess up my configuration or something? I tried

1. going back to edge-2026-04-18.1-stable-full
2. compose up -d --force-recreate
3. deleting my /opt/upmpdcli folder
4. double checking the subsonic login data still works
5. check that mpd works - it does

Weird thing: I tried checking the logs, strangely enough they are empty. IIrc there were pretty verbose logs here.

ll /opt/upmpdcli/log 
total 0

I believe last time I changed something was an attempt to customize the upmpdcli manufacturer etc. using the mound /opt/upmpdcli/config:/user/config:ro but I commented out and recreated the container, so it should be gone.

My compose.yml

services:
  mpd:
    image: giof71/mpd-alsa:latest
    container_name: mpd
    user: 1000:1000
    devices:
      - /dev/snd:/dev/snd
    # add user to audio group
    group_add: # Mandatory if socket permissions require group audio
      - 995
    ports:
      - 6600:6600/tcp
    environment:
      - USER_MODE=Y
      - PUID=1000
      - PGID=1000
      #- AUDIO_GID=29
      # This flag does not seem to do anything, group_add is required
      - AUDIO_GID=995
      - PULSE_AUDIO_OUTPUT_ENABLED=yes
      - PULSE_AUDIO_OUTPUT_CREATE=yes
#      - PULSE_SERVER=unix:/run/user/1000/pulse/native
      - PULSE_SERVER=unix:/run/pipewire/pulse/native
      #- AUDIO_BUFFER_SIZE=16MB
# because pulseaudio is not starting alsa is needed
      - ALSA_OUTPUT_CREATE=no
      - ALSA_OUTPUT_ENABLED=no
      # - MPD_ENABLE_LOGGING=yes
      # - MPD_LOG_LEVEL=verbose
    volumes:
      - /mnt/m/media/music:/music:ro
      - /opt/mpd/db:/db:rw
      - /opt/mpd/log:/log:rw
      - /opt/mpd/playlists:/playlists:rw
      #- /run/user/1000/pulse:/run/user/1000/pulse:rw
      - /run/pipewire/pulse/native:/run/pipewire/pulse/native:rw
      # - /opt/mpd/pulse-client.conf:/etc/pulse/client.conf:ro
    restart: unless-stopped

  upmpdcli:
    image: giof71/upmpdcli:edge
    container_name: upmpdcli
    network_mode: host
    environment:
      - LOG_LEVEL=3
      - LOG_ENABLE=yes
      # - PORT_OFFSET=2
      # - RENDERER_MODE=NONE
      - FRIENDLY_NAME=upmpdcli-thor
      - SUBSONIC_ENABLE=yes
      - SUBSONIC_BASE_URL=${SUBSONIC_BASE_URL}
      - SUBSONIC_PORT=${SUBSONIC_PORT}
      # - SUBSONIC_SERVER_PATH=${SUBSONIC_SERVER_PATH:-}
      - SUBSONIC_USER=${SUBSONIC_USER}
      - SUBSONIC_PASSWORD=${SUBSONIC_PASSWORD}
      - RENDERER_MODE=both
      - UPNPAV=yes
    volumes:
      - /opt/upmpdcli/cache:/cache
      - /opt/upmpdcli/log:/log
      #- /opt/upmpdcli/config:/user/config:ro
    labels:
      - com.centurylinklabs.watchtower.enable=false
    restart: unless-stopped

[Lenzelot]

Ah... using the navidrome demo credentials it works again... so it must be my navidrome or something.

[Lenzelot]

Oh my god... I forgot I changed my Dynamic DNS setup so now my Navidrome hostname resolved to some IPv6 address that went nowhere. My curl test command resolved to v4 for some reason... 🤦🏽‍♂️

[GioF71]

I'm glad it's ok now. Happy listening!