Placeshifter Causing Windows Server Crash

[Taddeusz]

I updated the Sage.jar file on my Windows server. Since then I cannot connect with Placeshifter to the server. It causes an immediate crash. Extenders & Windows Clients work fine. Attached is the log file but I didn't see any indication of a reason. Also, no crash is logged by Windows and the service doesn't restart itself. It just stops running.

[stuckless]

The only issues that I'm seeing is this one...

tmiranda.cd.api.IsComskipActive(api.java:107)
Sat 9/5 13:58:10.807 [MiniUIServerConnection@18b6919] sun.reflect.GeneratedMethodAccessor8.invoke(Unknown Source)
Sat 9/5 13:58:10.807 [MiniUIServerConnection@18b6919] sun.reflect.DelegatingMethodAccessorImpl.invoke(Unknown Source)
Sat 9/5 13:58:10.808 [MiniUIServerConnection@18b6919] java.lang.reflect.Method.invoke(Unknown Source)
Sat 9/5 13:58:10.808 [MiniUIServerConnection@18b6919] sage.Catbert$ReflectedJEPFunction.run(Catbert.java:2136)
Sat 9/5 13:58:10.808 [MiniUIServerConnection@18b6919] sage.jep.JEP$CommandEvaluator.evaluate(JEP.java:558)
Sat 9/5 13:58:10.808 [MiniUIServerConnection@18b6919] sage.jep.JEP.getValueAsObject(JEP.java:428)
Sat 9/5 13:58:10.808 [MiniUIServerConnection@18b6919] sage.Catbert.evaluateExpression(Catbert.java:641)
Sat 9/5 13:58:10.808 [MiniUIServerConnection@18b6919] sage.ZPseudoComp.processChain(ZPseudoComp.java:4703)
Sat 9/5 13:58:10.809 [MiniUIServerConnection@18b6919] sage.ZPseudoComp.processParentActions(ZPseudoComp.java:5212)
Sat 9/5 13:58:10.809 [MiniUIServerConnection@18b6919] sage.ZPseudoComp.evaluate(ZPseudoComp.java:4928)
Sat 9/5 13:58:10.809 [MiniUIServerConnection@18b6919] sage.ZPseudoComp.evaluateTree(ZPseudoComp.java:1154)
Sat 9/5 13:58:10.809 [MiniUIServerConnection@18b6919] sage.ZPseudoComp.evaluate(ZPseudoComp.java:5127)
Sat 9/5 13:58:10.809 [MiniUIServerConnection@18b6919] sage.ZPseudoComp.evaluateTree(ZPseudoComp.java:1154)
Sat 9/5 13:58:10.809 [MiniUIServerConnection@18b6919] sage.ZPseudoComp.evaluate(ZPseudoComp.java:5127)
Sat 9/5 13:58:10.810 [MiniUIServerConnection@18b6919] sage.ZPseudoComp.evaluateTree(ZPseudoComp.java:1154)
Sat 9/5 13:58:10.810 [MiniUIServerConnection@18b6919] sage.ZPseudoComp.evaluate(ZPseudoComp.java:5127)
Sat 9/5 13:58:10.810 [MiniUIServerConnection@18b6919] sage.ZPseudoComp.evaluateTree(ZPseudoComp.java:1154)
Sat 9/5 13:58:10.810 [MiniUIServerConnection@18b6919] sage.PseudoMenu.activate(PseudoMenu.java:100)
Sat 9/5 13:58:10.810 [MiniUIServerConnection@18b6919] sage.UIManager.setUI(UIManager.java:1932)
Sat 9/5 13:58:10.810 [MiniUIServerConnection@18b6919] sage.UIManager.advanceUI(UIManager.java:1558)
Sat 9/5 13:58:10.811 [MiniUIServerConnection@18b6919] sage.UIManager.run(UIManager.java:1255)
Sat 9/5 13:58:10.811 [MiniUIServerConnection@18b6919] sage.UIManager.createAltUI(UIManager.java:326)
Sat 9/5 13:58:10.811 [MiniUIServerConnection@18b6919] sage.MiniClientSageRenderer.handleNewConnection(MiniClientSageRenderer.java:537)
Sat 9/5 13:58:10.811 [MiniUIServerConnection@18b6919] sage.MiniClientSageRenderer.access$700(MiniClientSageRenderer.java:18)
Sat 9/5 13:58:10.811 [MiniUIServerConnection@18b6919] sage.MiniClientSageRenderer$1$1.run(MiniClientSageRenderer.java:401)
Sat 9/5 13:58:10.811 [MiniUIServerConnection@18b6919] sage.Pooler$PooledThread.run(Pooler.java:252)
Sat 9/5 13:58:10.812 [MiniUIServerConnection@18b6919] EXCEPTION in getValueAsObject:sage.jep.ParseException: Error in method reflection of tmiranda_cd_api_IsComskipActive of java.lang.reflect.InvocationTargetException for:tmiranda_cd_api_IsComskipActive() || (NTEState != null) || (AlertLevel > 0) || IsDoingLibraryImportScan() || (Size(GetCurrentlyRecordingMediaFiles()) > 0) || (Size(GetAiringsThatWontBeRecorded(false)) > 0) || (GetProperty("show_mouse_icons", true) && IsDesktopUI())

But I don't know if that is causing your issue or not... It looks like it's coming from comskip plugin, but again, I can't say if that is the reason for the failure or not.

[Taddeusz]

Placeshifter works fine with both my 32-bit & 64-bit Linux test boxes. Not sure why there's a difference with the Windows version? I have the comskip plugin installed on both.

[stuckless]

Did you update the Jar on the windows placeshifter as well? You might need a new jar if you've updated the server jar.

[Taddeusz]

Originally Posted by stuckless

Did you update the Jar on the windows placeshifter as well? You might need a new jar if you've updated the server jar.

Yes, I updated it also.

[sacrament055]

I just tested my windows placeshifter and noticed it seems to work fine. This is the original 7.1.9 placeshifter without any updates.

the sage.jar I'm running is the one Fuzzy posted in the thread I created looking for a new version.

[Taddeusz]

I've now determined that something is actively calling a shutdown. As far as I can tell It's these lines in the log that seem to be calling a shutdown.

Wed 9/9 9:22:24.441 [Thread-88@e90b51] [[SageTCPServer-Player:9279]]: EXCEPTION java.net.BindException: Address already in use: JVM_Bind AT java.net.DualStackPlainSocketImpl.bind0(Native Method)
Wed 9/9 9:22:24.442 [Thread-88@e90b51] java.net.DualStackPlainSocketImpl.socketBind(Unknown Source)
Wed 9/9 9:22:24.442 [Thread-88@e90b51] java.net.AbstractPlainSocketImpl.bind(Unknown Source)
Wed 9/9 9:22:24.444 [SageTV Shutdown@5508ab] SageTV SHUTDOWN is activating!

Not sure how to continue troubleshooting this.

[Narflex]

Originally Posted by Taddeusz

I've now determined that something is actively calling a shutdown. As far as I can tell It's these lines in the log that seem to be calling a shutdown.

Wed 9/9 9:22:24.441 [Thread-88@e90b51] [[SageTCPServer-Player:9279]]: EXCEPTION java.net.BindException: Address already in use: JVM_Bind AT java.net.DualStackPlainSocketImpl.bind0(Native Method)
Wed 9/9 9:22:24.442 [Thread-88@e90b51] java.net.DualStackPlainSocketImpl.socketBind(Unknown Source)
Wed 9/9 9:22:24.442 [Thread-88@e90b51] java.net.AbstractPlainSocketImpl.bind(Unknown Source)
Wed 9/9 9:22:24.444 [SageTV Shutdown@5508ab] SageTV SHUTDOWN is activating!

Not sure how to continue troubleshooting this.

Modify the source code in the sage.SageTV.java file where it prints out "SageTV SHUTDOWN is activating!" and right at that point have it call:

sage.AWTThreadWatcher.dumpThreadStates();

Then it'll print out the stacks of all the Java threads and you can find the one that's currently calling System.exit. I *think* that should work (because I'm pretty sure the thread that calls System.exit stays within that method until the VM has terminated so the stack trace should still show it while executing shutdown hooks).

[Taddeusz]

Ok, here is the log file that includes the dump. I'll look over it but posting it here for others to look at.

[wnjj]

There are two of these:

"Starting a player on socket 9279"

The second one is crashing due to "Address already in use".

The first one was opened with 192.168.22.121. The second one is on 157.142.237.13 if that helps any.

Is there some kind of resource conflict between local extenders and the placeshifter client?

[Narflex]

Here's the offender:

Wed 9/9 14:08:29.749 [SageTV Shutdown@1c0a1ce] Thread[Thread-91,1,main]
Wed 9/9 14:08:29.749 [SageTV Shutdown@1c0a1ce] java.lang.Thread.start0(Native Method)
Wed 9/9 14:08:29.749 [SageTV Shutdown@1c0a1ce] java.lang.Thread.start(Unknown Source)
Wed 9/9 14:08:29.749 [SageTV Shutdown@1c0a1ce] java.lang.ApplicationShutdownHooks.runHooks(Unknown Source)
Wed 9/9 14:08:29.749 [SageTV Shutdown@1c0a1ce] java.lang.ApplicationShutdownHooks$1.run(Unknown Source)
Wed 9/9 14:08:29.750 [SageTV Shutdown@1c0a1ce] java.lang.Shutdown.runHooks(Unknown Source)
Wed 9/9 14:08:29.750 [SageTV Shutdown@1c0a1ce] java.lang.Shutdown.sequence(Unknown Source)
Wed 9/9 14:08:29.750 [SageTV Shutdown@1c0a1ce] java.lang.Shutdown.exit(Unknown Source)
Wed 9/9 14:08:29.750 [SageTV Shutdown@1c0a1ce] java.lang.Runtime.exit(Unknown Source)
Wed 9/9 14:08:29.750 [SageTV Shutdown@1c0a1ce] java.lang.System.exit(Unknown Source)
Wed 9/9 14:08:29.750 [SageTV Shutdown@1c0a1ce] sagetcpserver.TCPServerPlayer.run(TCPServerPlayer.java:91)
Wed 9/9 14:08:29.750 [SageTV Shutdown@1c0a1ce] java.lang.Thread.run(Unknown Source)

Error message from plugin before this:

Wed 9/9 14:08:29.656 [Thread-91@6777a] [[SageTCPServer-Player:9279]]: Starting a player on socket 9279
Wed 9/9 14:08:29.657 [Thread-91@6777a] [[SageTCPServer-Player:9279]]: EXCEPTION java.net.BindException: Address already in use: JVM_Bind AT java.net.DualStackPlainSocketImpl.bind0(Native Method)

So it's TCPServerPlayer's code that's doing it. Naughty plugin. It looks like it'll call System.exit() if it has a failure binding a socket to a specific port. You may have something else running on port 9279...killing that should solve the problem. Ideally though; the plugin would be fixed so that it doesn't do things like this...plugins should NEVER be calling System.exit().

[Taddeusz]

Yeah, we can chalk this up to a misbehaving plugin. Looks like the SageTCPServer plugin is holding open a bunch of ports. Disabling that plugin has fixed the issue. I guess I can live without the remote control app on my iDevices. I rarely use it anyway.