Yes, normal. It logs everything, even if it’s not used.
few times with 2 seconds pause:
openhab> threads | grep "Squeeze Server Listener"
"pipe-grep "Squeeze Server Listener"" Id=26053 in TIMED_WAITING on lock=java.io.PipedInputStream@1d59569
~ openhab> threads | grep "Squeeze Server Listener"
"pipe-grep "Squeeze Server Listener"" Id=26050 in RUNNABLE
~ openhab> threads | grep "Squeeze Server Listener"
"pipe-grep "Squeeze Server Listener"" Id=26053 in RUNNABLE
~ openhab> threads | grep "Squeeze Server Listener"
"pipe-grep "Squeeze Server Listener"" Id=26050 in TIMED_WAITING on lock=java.io.PipedInputStream@13de78b
~ openhab> threads | grep "Squeeze Server Listener"
"pipe-grep "Squeeze Server Listener"" Id=26052 in RUNNABLE
~ openhab> threads | grep "Squeeze Server Listener"
"pipe-grep "Squeeze Server Listener"" Id=26050 in RUNNABLE
~ openhab>
This looks normal. Did you see this while the “no connection” problem was occurring?
I think I found the cause of this issue.
If you are still in TRACE mode, when the problem occurs, check the log file for a line that contains subscribe playlist
Hi
Logging in TRACE mode is still on. Just don’t have time to look into it.
If you are still in TRACE mode, when the problem occurs, check the log file for a line that contains
subscribe playlist
will do.
By any chance are you using a Windows app called Squeeze 11?
Nope. Only piCorePlayers used together with LMS + addons UPnP and Chromecast (disabled at the moment). Only piCorePlayers are used with the binding.
If you are still in TRACE mode, when the problem occurs, check the log file for a line that contains
subscribe playlist
I didn’t find this.
I attachedl log (had to change name to .xml)
oh.log.xml (837.1 KB)
Thanks for the log.
It’s all rather strange. The connection seems to just drop for no apparent reason. I also see some socket timeout exceptions in the Telegram action. Not sure if that’s a symptom of the same issue that caused the squeezebox binding to lose its connection to the LMS.
Did you happen notice the status of the SqueezeBox things? Were they still showing ONLINE? I’m guessing they were.
How about CPU utilization on the OH box? Normal?
And what about the Squeeze Server Listener thread?
Did you happen notice the status of the SqueezeBox things? Were they still showing ONLINE? I’m guessing they were.
They seems to be ON-LINE, players and the server, at least this is how it looks in Paper UI.
How about CPU utilization on the OH box? Normal?
And what about the Squeeze Server Listener thread?
Will check it when I go back home.
I also see some socket timeout exceptions in the Telegram action.
Yes, this happens when telegram is idle for some bigger amount of time. When my door lock change to OPEN/CLOSE then one of my rules send postupdate to telegram bots in the same way as i play sounds. Anyway telegram messages are always delivered even if this error occurs. Do you think there may be connection between this telegram error and squeezebox issue? It occurs kind of at the same time as i trigger playsound.
It’s hard to say. Maybe, maybe not. It was just something I noticed in the log.
I understand why the reconnection is not being attempted. The reconnect only triggers when the listener thread has an error, and that doesn’t appear to be happening here (hence the reason I’m asking about the listener thread).
Just checked remotely listener with the command you specified before and the result is the same as before.
Looking at top command, the java CPU% balances between 60 - 300 both ways. Memory ~20%.
Thanks. I need to give some thought to how this can be fixed.
Since I can’t reproduce this, if I make a version of the binding, would you be able to test it for me?
Sure I can do any testing you need, depending on available time of course.
@zacofunny I posted a new version of the binding here, along with installation instructions. I don’t recall what version of OH you’re on. The binding was built for 2.4, so I’m not sure how it will work if you are still running 2.3.
When the binding fails to send a command to the LMS due to an exception, it should activate the logic to attempt to reconnect to the LMS every 60 seconds.
You should only need debug level logging, since I added quite a bit of debug statements around the impacted code. If my change doesn’t fix the problem, at least we’ll get more information about what’s happening.
Cool. I’m on OH 2.4 latest milestone running on OpenHABian Pine64. I will try the new binding version once I back home today.
1 Like
OK I’ve set it up. Hopefully we will see the results shortly.
BTW Why do I need to set
feature:install openhab-transport-upnp
?
Good question. This command installs the uPnP feature.
The squeezebox binding depends on uPnP for discovery. When you install from Paper UI or addons.cfg, dependencies are installed automatically. When you drop a jar in the addons folder, dependencies don’t get automatically installed, so it must be done manually.
Thanks. I see my fix didn’t work, but there are a few more clues in the log file.
Let me work on another version that might fix it.
1 Like