Any way to tell whether UMS is streaming?

[ishareit]

I am building a little utility that will prevent system from entering sleep mode only when UMS is streaming any content to a renderer. I was wondering if there is anything that I can monitor on the server machine to determine whether UMS is actively streaming or not? Thanks in advance!

[squadjot]

You know UMS has a "Prevent Windows from sleeping" option on the "General configuration tab"?

[ishareit]

You know UMS has a "Prevent Windows from sleeping" option on the "General configuration tab"?

Yes, but that option prevents the computer from going to sleep irrespective of whether it is streaming or not... essentially keep it awake all the time which is not what I want for energy conservation reasons. Please correct me if that is not how this option is supposed to work.

In any case, I ended up creating a little utility that sends sleep prevention signals if the computer's outgoing network utilization is beyond a certain threshold. Seems to be working well so far.

[Wolfgan]

You know UMS has a "Prevent Windows from sleeping" option on the "General configuration tab"?

Yes, but that option prevents the computer from going to sleep irrespective of whether it is streaming or not... essentially keep it awake all the time which is not what I want for energy conservation reasons. Please correct me if that is not how this option is supposed to work.

In any case, I ended up creating a little utility that sends sleep prevention signals if the computer's outgoing network utilization is beyond a certain threshold. Seems to be working well so far.

I agree some kind of "intelligent sleep" option would be welcome (ie Sleep after X' of no streaming activity). UMS prevents the server to go to sleep even if not streaming, so when I disabled that and played with windows sleep setting, the machine went to sleep right after the streaming ended (so no chance for me to catch the remote and start another file!

[squadjot]

Hi!, If i test by

1. Set sleepmode to 2 minutes
2. Have no active renderers/players on the network.
3. Start UMS but without doing any streaming/browsing.

Result: The computer sleeps after a couple of minutes, atleast it does on my win7.
If instead use UMS right away, the computer does not sleep.
If UMS has a minimum time before sleep, or if UMS keeps computer live aslong it can see the render/player.. i dont' know, but the times i've fell asleep with UMS running, the computer was sleeping when i woke up

If youre interested, there's also a "shutdown" plugin (manual/remote)
Its made for earlier version, but will still work with UMS:
http://www.ps3mediaserver.org/forum/vie ... =12&t=6117

Edit: Yea i guess it could be useful to have more detailed options.

[ishareit]

2. Have no active renderers/players on the network.

That is an interesting prerequisite. I have several devices on my network that are considered renderers like my sound receiver, tv, amazon tv stick running kodi, etc. For me, having no active renderers/players is almost impossible.

The inbuilt option needs to do this:
- If UMS is streaming actively any content, then ask Windows to keep prevent sleeping.
- When UMS stops streaming the content, the inbuilt Windows sleep timer starts. For example, if the sleep setting is 30 minutes, and I stopped playing a movie at 9:00 am, then the computer should sleep at 9:30 am.
- Should be independent of whether there are any active renderers/player on the network.

The above is what my little outbound n/w traffic monitor utility does essentially.

PS: If I check the "Prevent Windows from Sleeping" option and restart UMS, I can see that javaw.exe is requesting system to stay awake. If I uncheck that option, javaw.exe is no longer making that request. All without streaming anything from UMS (but with active renderers on the network). You can check this info by running "powercfg -requests" in Windows using admin command prompt.

[squadjot]

Thanks for your input, i widely agree.

[Wolfgan]

I may be wrong, but looking at the code at https://github.com/UniversalMediaServer ... Utils.java it seems the Prevent Sleep is implemented as a timeout of 40 seconds regardless the system is streaming or not.
Maybe some logic can be added for an "Sleep if no activity for X seconds/minutes"?

[squadjot]

I'm no java programmer, but I think the 40 secs is to avoid too many calls to the function.
So if its more than 40 secs cince the function was last called...then the function is allowed to continue.

It seems UMS does not set any "sleeptime", it just uses whatever settings you've set in windows.

Looking at the code:

In RequestHandlerV2.java

Code: Select all

public void messageReceived(ChannelHandlerContext ctx, MessageEvent e) 
79 		throws Exception { 
80 		RequestV2 request = null; 
81 		RendererConfiguration renderer = null; 
82 		String userAgentString = null; 
83 		ArrayList<String> identifiers = new ArrayList<>(); 
84 
 
85 		HttpRequest nettyRequest = this.nettyRequest = (HttpRequest) e.getMessage(); 
86 
 
87 		InetSocketAddress remoteAddress = (InetSocketAddress) e.getChannel().getRemoteAddress(); 
88 		InetAddress ia = remoteAddress.getAddress(); 
89 
 
90 		// Is the request from our own Cling service, i.e. self-originating? 
91 		boolean isSelf = ia.getHostAddress().equals(PMS.get().getServer().getHost()) && 
92 			nettyRequest.headers().get(HttpHeaders.Names.USER_AGENT) != null && 
93 			nettyRequest.headers().get(HttpHeaders.Names.USER_AGENT).contains("UMS/"); 
94 
 
95 		// Filter if required 
96 		if (isSelf || filterIp(ia)) { 
97 			e.getChannel().close(); 
98 			LOGGER.trace(isSelf ? 
99 				("Ignoring self-originating request from " + ia + ":" + remoteAddress.getPort()) : 
100 				("Access denied for address " + ia + " based on IP filter")); 
101 			return; 
102 		} 
103 
 
104 		LOGGER.trace("Opened request handler on socket " + remoteAddress); 
105 		PMS.get().getRegistry().disableGoToSleep(); 

This tells me that PMS.get().getRegistry().disableGoToSleep(); is called every time a message is received.


In RequestV2.java

Code: Select all

public void operationComplete(ChannelFuture future) { 
904 				chunkWriteFuture.addListener(new ChannelFutureListener() { 
905 					@Override 
906 					public void operationComplete(ChannelFuture future) { 
907 						try { 
908 							PMS.get().getRegistry().reenableGoToSleep(); 
909 							inputStream.close(); 
910 						} catch (IOException e) { 
911 							LOGGER.debug("Caught exception", e); 
912 						} 
913 
 
914 						// Always close the channel after the response is sent because of 
915 						// a freeze at the end of video when the channel is not closed. 
916 						future.getChannel().close(); 
917 						startStopListenerDelegate.stop(); 
918 					} 
919 				}); 

This tells me that; if the operation is complete, we close the stream and call PMS.get().getRegistry().reenableGoToSleep(); (Using whatever power management scheme windows is using)
That makes sense to me.

[scriptor]

I observed the same problem as ishareit: when I enable "Prevent Windows from sleeping" my computer does not go to sleep when finished streaming. The windows command

Code: Select all

powercfg -requests

(call it as administartor!) tells that UMS prevents sleeping. When I disable "Prevent Windows from sleeping" this process disappears from the list.

I am sure that there is a logical bug in the WinUtils.java:

Code: Select all

	@Override
	public void disableGoToSleep() {
		// Disable go to sleep (every 40s)
		if (configuration.isPreventsSleep() && System.currentTimeMillis() - lastDontSleepCall > 40000) {
			LOGGER.trace("Calling SetThreadExecutionState ES_SYSTEM_REQUIRED");
			Kernel32.INSTANCE.SetThreadExecutionState(Kernel32.ES_SYSTEM_REQUIRED | Kernel32.ES_CONTINUOUS);
			lastDontSleepCall = System.currentTimeMillis();
		}
	}

	/* (non-Javadoc)
	 * @see net.pms.io.SystemUtils#reenableGoToSleep()
	 */
	@Override
	public void reenableGoToSleep() {
		// Reenable go to sleep
		if (configuration.isPreventsSleep() && System.currentTimeMillis() - lastGoToSleepCall > 40000) {
			LOGGER.trace("Calling SetThreadExecutionState ES_CONTINUOUS");
			Kernel32.INSTANCE.SetThreadExecutionState(Kernel32.ES_CONTINUOUS);
			lastGoToSleepCall = System.currentTimeMillis();
		}
	}

The "if" prevents calling SetThreadExecutionState faster than every 40 seconds. That's a good idea but there are situations where it has to be called more often! For instance:

1. disableGoToSleep is called (streaming started)
2. after 1 minute reenableGoToSleep is called (streaming finished)
3. a few seconds later disableGoToSleep is called (streaming is started again)
4. a few seconds later reenableGoToSleep is called (streaming finished again)

The second disableGoToSleep will be executed but the second reenableGoToSleep will not! Even though the streaming is finished the PC will not go to sleep anymore.
You can also find examples where a needed disableGoToSleep is suppressed.

I would suggest to change the conditions in the "if" clauses:

Code: Select all

	@Override
	public void disableGoToSleep() {
		// Disable go to sleep (every 40s)
		if (configuration.isPreventsSleep() && (System.currentTimeMillis() - lastDontSleepCall > 40000) || (lastGoToSleepCall >= lastDontSleepCall)) {
			LOGGER.trace("Calling SetThreadExecutionState ES_SYSTEM_REQUIRED");
			Kernel32.INSTANCE.SetThreadExecutionState(Kernel32.ES_SYSTEM_REQUIRED | Kernel32.ES_CONTINUOUS);
			lastDontSleepCall = System.currentTimeMillis();
		}
	}

	@Override
	public void reenableGoToSleep() {
		// Reenable go to sleep
		if (configuration.isPreventsSleep() && (System.currentTimeMillis() - lastGoToSleepCall > 40000) || (lastDontSleepCall >= lastGoToSleepCall)) {
			LOGGER.trace("Calling SetThreadExecutionState ES_CONTINUOUS");
			Kernel32.INSTANCE.SetThreadExecutionState(Kernel32.ES_CONTINUOUS);
			lastGoToSleepCall = System.currentTimeMillis();
		}
	}

Unfortunately I can not test the modifications. Maybe anyone of the developer team can do that? Thank you!