01-24-2020, 06:03 PM
@Maitresinh
I've reread this thread and I'm unclear about some of your statements.
1. Is it the case that you eventually reach this crash condition even when you run moOde as was previously suggested by @CallMeMike. That is to say, with MPD audio output -> Local (and, preferably with Bluetooth Renderer setting "OFF")? Or is the case that you eventually reach this crash condition only when you run with MPD audio output -> Bluetooth? You've also mentioned using Bluetooth as the source. Do you eventually reach the crash condition with Bluetooth as source and Local as output?
2. You've talked about moOde eventually crashing, after 12h - 24h, say, but in your latest post you said it's crashed and the system log shows it's only been up for an hour. Did it crash right away?
3. Of course, the underlying system hasn't crashed: you can open an SSH session. It's the UI that isn't responding because parts of the moOde application aren't running. There's several possibilities.
To try to gain a toehold, let's look for several possible hints in the logs.
egrep -i '(warn|error|fail)' /var/log/moode.log
egrep -i '(warn|error|fail)' /var/log/syslog
So, for example, here's the output from a moOde player where I purposely killed the MPD process (and the watchdog script helpfully started it again.)
The FIPS failure reports are expected and can be ignored for this exercise.
It's probably most enlightening to do this test as soon after your system crashes as possible since it's possible the logs get rotated or truncated after some time (I really don't recall when this happens.)
And, yes, Tim was suggesting you re-image your uSD card.
Regards,
Kent
I've reread this thread and I'm unclear about some of your statements.
1. Is it the case that you eventually reach this crash condition even when you run moOde as was previously suggested by @CallMeMike. That is to say, with MPD audio output -> Local (and, preferably with Bluetooth Renderer setting "OFF")? Or is the case that you eventually reach this crash condition only when you run with MPD audio output -> Bluetooth? You've also mentioned using Bluetooth as the source. Do you eventually reach the crash condition with Bluetooth as source and Local as output?
2. You've talked about moOde eventually crashing, after 12h - 24h, say, but in your latest post you said it's crashed and the system log shows it's only been up for an hour. Did it crash right away?
3. Of course, the underlying system hasn't crashed: you can open an SSH session. It's the UI that isn't responding because parts of the moOde application aren't running. There's several possibilities.
To try to gain a toehold, let's look for several possible hints in the logs.
egrep -i '(warn|error|fail)' /var/log/moode.log
egrep -i '(warn|error|fail)' /var/log/syslog
So, for example, here's the output from a moOde player where I purposely killed the MPD process (and the watchdog script helpfully started it again.)
Code:
pi@moode3a:/etc $ egrep -i '(fail|error|warn)' /var/log/moode.log
20200124 121516 watchdog: MPD restarted (check syslog for errors)
pi@moode3a:/etc $ egrep -i '(fail|error|warn)' /var/log/syslog
Jan 24 10:42:15 moode3a rngd[307]: stats: FIPS 140-2 failures: 0
Jan 24 11:42:15 moode3a rngd[307]: stats: FIPS 140-2 failures: 0
Jan 24 12:15:14 moode3a systemd[1]: mpd.service: Failed with result 'signal'.
Jan 24 12:42:15 moode3a rngd[307]: stats: FIPS 140-2 failures: 0The FIPS failure reports are expected and can be ignored for this exercise.
It's probably most enlightening to do this test as soon after your system crashes as possible since it's possible the logs get rotated or truncated after some time (I really don't recall when this happens.)
And, yes, Tim was suggesting you re-image your uSD card.
Regards,
Kent
