Roon server not completing start-up: continaully attempts restart

It is definitely not .cue files, and I am 99% sure it is not data, as the identical library that I have here can be processed without failure on a different Antipodes server.

@Noris, is it possible to set more verbose logging for Roon Server. Also, can you advise the meaning of ‘$R -eq 122’ in the script below. It appears that this is where the startup fails and the ‘Error’ message is written:

image

Hi @noris,
I have added ‘echos’ to the Roon Server start.sh:


When I kill the process it restarts, and then appears to be looping at the point after 'echo 3’in the script:


The PID is unchanged, it is not starting a new process.

Hi

@Tony_Devitt - Thanks for looking into this further, let me get your findings over to our technical team and see if they can comment on them.

@Raymond_Daniel/@Tony_Devitt - Just a thought, have we completely eliminated the library and audio analysis from being part of the equation? If either of you wish to try a test, this might be helpful:

  • Start with a fresh database
  • Under Roon Settings -> Library, set Background Audio analysis / On-Demand Audio analysis to Off
  • Try re-importing the same library
  • Verify if the issue still occurs

Hi @Noris, I did as you suggested, with mixed results. Sometimes it will start and display the ‘Running’ message but more often I will get the ‘Error’ message. After a reboot it always errors, and then if I repeatedly clear the cache I will eventually get a successful start. When this happens however loading Roon Remote is very slow, and CPU is heavily utilised. I would emphasise that the Antipodes I am working with is from mid-2015, pre-Roon, and is clearly underpowered for current requirements.

I will arrange to check @Raymond_Daniel’s Antipodes.

1 Like

Hi @Tony_Devitt,

Thanks for running that test.

Which Cache folder are you referring to, the one under RoonServer/Cache?

I have been discussing this with QA, it looks like it’s failing at the waiting for PID state because of a crash loop, but the root cause is not clear. It is possible that the crash is due to a low-level event triggering it.

Another thought while I was discussing this case today, what happens if you start with a fresh database and have no database at all on the Antipodes. It sounds like restarting Roon Server immediately causes it to go into this loop, does it still get into the loop with an empty database? What about an empty database before even logging in?

Hi @Noris,

  1. Yes, /var/roon/RoonServer/Cache.

  2. When I reinstall the Roon Server app it starts normally, I can login and set it up, and the music library is then scanned. Once the scan has completed Roon Remote can be used to browse the library and play tracks. However as soon as I restart the Roon Server app, with the music library now in place, the start fails, with the ‘Error’ message always exactly 18 seconds after the ‘Started’ message.

Unfortunately I have had to return this server to the owner so cannot test further with it. It is not a problem for him as he uses this older Antipodes as a NAS/ripper only.

Hi @Tony_Devitt,

I work form home on Mon, Tue, and Thurs. If it helps to use my Antipodes for tests we can probably arrange something, though I understand it might be clunky for you via remote (our archaic internet does not help) and more trouble than it is worth. Just a thought.