LMS very unresponsive after upgrade from 7.9.1 to 7.9.2

[GoodVibrations]

Been running a rock stable setup on Debian for more than 12 years now.
I almost never "fiddle" with it, so I'm more than a little rusty with my own system. That's the price you pay when it just works. :-)

But the other day I got tired of it bugging me with a message about an upgrade being available.
So I upgraded from 7.9.1 to 7.9.2. (dpkg -i logitechmediaserver_7.9.2_all.deb).
I'm running it on 4.9.0-12-686-pae (perl is v5.24.1).

After the upgrade it became almost completely unresponsive.
I cannot access it from a browser or any of the players.

Top shows:

 PID USER      PR  NI    VIRT    RES    SHR S  %CPU %MEM     TIME+ COMMAND
  424 squeeze+  20   0  186788 178296  12868 R 100,0  4,3   1300:10 squeezebox+

I have these processes:

squeeze+   405  0.0  0.0   5508  3228 ?        S    jun09   0:00 /bin/bash /usr/sbin/squeezeboxserver_safe /usr/sbin/squeezeboxserver --prefsdir /var/lib/squeezeboxserver/prefs --logdir /var/log/squeezeboxserver/ --cachedir /var/lib/squeezeboxserver/cache --charset=utf8
squeeze+   424 63.9  4.3 186788 178296 ?       R    jun09 1300:59 /usr/bin/perl /usr/sbin/squeezeboxserver --prefsdir /var/lib/squeezeboxserver/prefs --logdir /var/log/squeezeboxserver/ --cachedir /var/lib/squeezeboxserver/cache --charset=utf8
squeeze+   556  0.1  0.0 163280  3696 ?        Sl   jun09   2:05 /var/lib/squeezeboxserver/cache/InstalledPlugins/Plugins/CastBridge/Bin/squeeze2cast-x86-static -Z -I -b 192.168.86.221 -x /var/lib/squeezeboxserver/prefs/castbridge.xml

It stays this way for around 9 hours.... And then it suddenly comes alive again and I can access it via a browser and the different players.

I then rebooted my box, and LMS goes crazy again for the next 9 hours!?:

2020-06-09 10:42:03 squeezeboxserver_safe stopped.
2020-06-09 10:44:00 squeezeboxserver_safe started.
[20-06-09 10:44:05.7496] main::init (387) Starting Logitech Media Server (v7.9.3, 1591161343, Thu Jun  4 04:16:22 CEST 2020) perl 5.024001 - i686-linux-gnu-thread-multi-64int
[20-06-09 19:25:59.9784] Plugins::LazySearch2::Plugin::scanDoneCallback (2332) Lazifying database items not already done
[20-06-09 19:26:00.8171] Plugins::LazySearch2::Plugin::lazifyDatabase (2389) No database items require lazification
[20-06-09 19:26:01.4761] Slim::Networking::SqueezeNetwork::_error (500) Unable to login to SN: Ugyldigt brugernavn eller adgangskode til mysqueezebox.com.
[20-06-09 19:26:01.4770] Slim::Networking::SqueezeNetwork::_init_error (180) Unable to login to mysqueezebox.com, sync is disabled: Ugyldigt brugernavn eller adgangskode til mysqueezebox.com. (https://www.mysqueezebox.com)
[20-06-09 19:26:01.4782] Slim::Networking::SqueezeNetwork::_init_error (196) mysqueezebox.com sync init failed: Ugyldigt brugernavn eller adgangskode til mysqueezebox.com., will retry in 185100 (https://www.mysqueezebox.com)
[20-06-09 19:26:29.5620] Slim::Networking::Async::connect (114) Failed to connect to ayera.dl.sourceforge.net:443, because
Net::HTTP: connect: Connection refused
[20-06-09 19:26:29.5630] Slim::Utils::PluginDownloader::_downloadError (260) unable to download CastBridge from http://downloads.sourceforge.net/pro...dge-1.40.3.zip - Connect timed out: Connection refused

This morning i added a new album and then started a scan:

[20-06-10 07:41:30.5650] Plugins::TrackStat::Storage::refreshTracks (1242) TrackStat: Synchronizing TrackStat data, please wait...
[20-06-10 16:25:56.3386] Plugins::TrackStat::Storage::refreshTracks (1891) TrackStat: Synchronizing TrackStat data finished
[20-06-10 16:25:56.3406] Plugins::LazySearch2::Plugin::scanDoneCallback (2332) Lazifying database items not already done
[20-06-10 16:26:02.4100] Plugins::TrackStat::Storage::refreshTracks (1242) TrackStat: Synchronizing TrackStat data, please wait...

It's been unresponsive since and top show 100% cpu usage by squeezeboxserver.

Yesterday I upgraded to the latest nightly in the hope that it would help. It didn't.

I haven't changed any settings at all.
I have just upgraded the server-package.

I have no idea what the server is working so hard on? :-)

Any kind of help will be much appreciated.
Regards.

[GoodVibrations]

I just accepted an update to the CastBridge plugin.
The server then restarts, and has now gone into the "woods" to do god knows what?!

Oh well. I guess It'll be back in another 9-10 hours...

[20-06-11 16:52:59.3242] main::init (387) Starting Logitech Media Server (v7.9.3, 1591161343, Thu Jun  4 04:16:22 CEST 2020) perl 5.024001 - i686-linux-gnu-thread-multi-64int

Could it be a plugin problem?
Some plugin working on 7.9.1 but not on 7.9.2?

I think the time it takes Trackstat to do its thing, looks fishy:

[20-06-10 07:41:30.5650] Plugins::TrackStat::Storage::refreshTracks (1242) TrackStat: Synchronizing TrackStat data, please wait...
[20-06-10 16:25:56.3386] Plugins::TrackStat::Storage::refreshTracks (1891) TrackStat: Synchronizing TrackStat data finished
[20-06-10 16:25:56.3406] Plugins::LazySearch2::Plugin::scanDoneCallback (2332) Lazifying database items not already done
[20-06-10 16:26:02.4100] Plugins::TrackStat::Storage::refreshTracks (1242) TrackStat: Synchronizing TrackStat data, please wait...
[20-06-11 01:13:55.8223] Plugins::TrackStat::Storage::refreshTracks (1891) TrackStat: Synchronizing TrackStat data finished

[Apesbrain]

I'd like to help, but I'm on Windows. If I had your issue I'd terminate all Squeezebox-related processes: squeezesvr/squeezeboxserver, scanner, squeeze*. I'd also clear all Squeezebox-related cache:

http://wiki.slimdevices.com/index.ph...file_locations

Restart Squeezebox in safe mode. I'm not sure how this is done in linux. Try adding --failsafe or -failsafe to command line.

If not a plugin, your problem may be related to the version of Perl in use. On my install of LMS 7.9.3 beta, the Perl version is 5.14.1.

[bpa]

I just accepted an update to the CastBridge plugin.
The server then restarts, and has now gone into the "woods" to do god knows what?!

Oh well. I guess It'll be back in another 9-10 hours...

All the "bridge" plugins are very large about 30Mb and they come from sourceforge. Sourceforge has been behaving very badly recently sometimes taking over 20 minutes to download. LMS in 7.9.2 is setup with a 20 sec download timeout, some fixes were put in to 7.9.3 & 8.0.* to help with such long downloads.

My advice, forget about plugins for the moment and get your system running and clear of other issues before dealign with plugins.

[GoodVibrations]

So I kill'ed the processes, cleared cache.db and started in failsafe.
And that works.
So my guess is that it's a plugin-problem.

Seeing that Trackstat took a long very long time to synchronize I'm a bit suspicious of this.
So I removed failsafe and rebooted and waited for it to respond (9+ hours).

I then cleared the Trackstat DB and restored a Trackstat backup from the night before I upgraded LMS.
I then deleted unused Trackstat statistics (which took 9 hours. The same time LMS is unresponsive after a reboot....).
And then rebooted my box.
No luck. Same response, LMS will be unresponsive for the next 9 hours.

I'm considering clearing the Trackstat DB, reboot, and then see if it's unresponsive.
I have 12 years of statistics in the DB and will be very tired of losing these...
If it's Trackstat that's the problem, I have no idea why it should be?
I haven't moved any of my music or changed any settings.
Only upgraded the LMS package.

But supposing that it's another plugin that's the problem, can I disable them all (not uninstall), and then enable them one by one, without losing any of their settings?
If I remove their tick-marks, will they then be uninstalled or disabled?



Sorry for the Danish. :-)

[GoodVibrations]

My advice, forget about plugins for the moment and get your system running and clear of other issues before dealign with plugins.

Sound advice.
But my problem is I have absolutely no idea what the issues are...
I can't find anything in any logs. Just a squeeze process that runs 100% for 9-10 hours after every reboot/scan.

I did a minor upgrade from 7.9.1 to 7.9.2. Nothing else.
After that, my system is literally bricked.
I had absolutely no problems or issues on 7.9.1.
So why did I upgrade then? Well... :-)

I hate to say this after close to 13 years with a very stable system...
But i have one legacy player left, sb3, the rest are CCA's and squeezelites.
And feeling very unknowledgeable about my own system, I've bought a single Ikea Symfonisk.
So instead of keeping on banging my head against my own LMS-ignorance, I'll be using the Ikea to begin looking in to that "other system". :-)
I have absolutely always loved my Slim/Squeeze/Logitech system. I just don't know enough anymore to keep it running.

[bpa]

Sound advice.
But my problem is I have absolutely no idea what the issues are...
I can't find anything in any logs. Just a squeeze process that runs 100% for 9-10 hours after every reboot/scan.

I did a minor upgrade from 7.9.1 to 7.9.2. Nothing else.
After that, my system is literally bricked.
I had absolutely no problems or issues on 7.9.1.
So why did I upgrade then? Well... :-)

I hate to say this after close to 13 years with a very stable system...
But i have one legacy player left, sb3, the rest are CCA's and squeezelites.
And feeling very unknowledgeable about my own system, I've bought a single Ikea Symfonisk.
So instead of keeping on banging my head against my own LMS-ignorance, I'll be using the Ikea to begin looking in to that "other system". :-)
I have absolutely always loved my Slim/Squeeze/Logitech system. I just don't know enough anymore to keep it running.

Start slimserver with -d_startup command line option to see whether install is OK.

[GoodVibrations]

Start slimserver with -d_startup command line option to see whether install is OK.

Like this?
service logitechmediaserver start -d

[bpa]

Like this?
service logitechmediaserver start -d

No . Offhand I don't know how to add command line option to a service, I think they are called startup parameters and the parameter will be "-d_startup"

[atrocity]

Seeing that Trackstat took a long very long time to synchronize I'm a bit suspicious of this.

In my experience, TrackStat will bring the server to its knees after a scan, at least on my 100,000 track library. I've just gotten used to it and only scan when I don't need to access that server for a while.

Though I only see the problem for maybe 2 hours, not 9.