This page last changed on Aug 24, 2013 by garfield.arbuckle.

Hey guys,
I bought OR Pro and I'm migrating everything from the free designer to the pro designer.
Everything seems to be working again except for the Web console.
When I go to the webconsole via http://192.168.1.X:8688/webconsole/ it starts loading and then throws the following error:
Console Error: Panel Definition is not correct.
I've set the web console to the same resolution as the panel of my Touchpad. The same error occurs when I manually select the panel in the web console instead of automatically loading it via the controller settings in the web console.
It worked perfectly fine on the free controller.
Anyone any ideas?
Thanks
Stijn

In addition to this, the pro controller is extremely slow on my raspi.
Turning on a ligth now takes at least 5 times the amount of time to turn it on on the free controller.
CPU of the raspi runs at an average of 70% but spikes to 100%.
Is the raspi not suited for the pro controller?
Thanks
Stijn

Posted by garfield.arbuckle at Aug 27, 2013 22:17

I don't have access to Pro accounts for the moment, so will need to wait until next week to be able to test the webconsole against your design.

In the meantime, have you tried the usual cleanup operations, clearing your browser cache, etc. In case it is attempting to load previous settings from the local storage, or something else that might be confusing it?

Posted by juha at Aug 28, 2013 11:29

There shouldn't be major performance differences as far as we are aware between the free and pro versions. This is the first such report that I recall.

Which protocol integration are you using with the Pro version?

Posted by juha at Aug 28, 2013 11:31

Hey Juha,
As far as protocol integration goes I'm using KNX and some HTTP commands to control my squeezebox.
The design itself is not much, two screens with some buttons and sliders just to test all functionality.
I think I have tried with several browsers, but I'll do a cleanup tonight and maybe also test from a different device just to be sure.
For the slowness the only thing I can maybe add is that I'm running hardfloat on the pi. I don't know if that would make any difference...?
Thanks a lot
Stijn

Posted by garfield.arbuckle at Aug 28, 2013 11:40

In general I would expect the hardfloat to perform better. Did you use the same JVM (for hard floats I assume you're on Oracle Embedded Java 1.8 Early Access) with both free and pro versions?

Which version of free did you use previously?

Posted by juha at Aug 28, 2013 11:52

Yes I didn't change anything to the JVM and it's the same device I'm running it on.
I have a folder for the free version and a folder for the pro version and yes I do think I am on Oracle Embedded Java 1.8 Early access although I would have to verify (don't have access now).
The free version I used is the one you get when you go to the download page that's on your raspi tutorial, I can get you the exact version this evening.

Posted by garfield.arbuckle at Aug 28, 2013 12:03

Ok so some extra details.
I've tested on OSX with Opera and Safari, both give the same error.
Tried on a winXP VM with FF and also gives the same error. IE wouldn't load it properly but well yeah IE...
java version "1.6.0_27"
OpenJDK Runtime Environment (IcedTea6 1.12.5) (6b27-1.12.5-1+rpi1)
OpenJDK Zero VM (build 20.0-b12, mixed mode)

Was running: OpenRemote-Controller-2.1.0_SNAPSHOT-2013-06-17 (works fine)
Now running: OpenRemote-Controller-Pro1.0.10 (works slow)

Running Hardfloat on raspi

Hope this gives you some extra insight. Seems not to be Java1.8 but 1.6

Posted by garfield.arbuckle at Aug 28, 2013 19:49

Can you check if the Pro Controller has the following property set to "true".

file: webapps/controller/WEB-INF/classes/config.properties
property: controller.performBeehiveSyncing

If that is true can you try and set it to false to see if the performance if the Pro controller would be better.

Posted by mredeker at Aug 29, 2013 08:37

Hey Marcus,
I tried that but doesn't change anything.
Although I did find out the issue only exists when turning a light on and off the first time.
Afterwards the speed is the same as with the free controller. I don't know if this is an issue that only occurs at bootup of the controller or if the turn on/off speed also deteriorates throughout time.
I will leave the controller on for some time now to see if the speed stays the same or changes again.
Cheers
Stijn

Posted by garfield.arbuckle at Aug 30, 2013 12:12

I noticed I also get this in the "console" output
Seems like the controller is looking for an ISY-99 device, which I don't have and certainly not at that IP.
Maybe I should disable this... anyone any idea how?

ERROR pool\-1\-thread\-1: IOException while reading data from ISY-99
org.apache.http.conn.HttpHostConnectException: Connection to http://192.168.41.12 refused
at org.apache.http.impl.conn.DefaultClientConnectionOperator.openConnection(DefaultClientConnectionOperator.java:127)
at org.apache.http.impl.conn.AbstractPoolEntry.open(AbstractPoolEntry.java:147)
at org.apache.http.impl.conn.AbstractPooledConnAdapter.open(AbstractPooledConnAdapter.java:108)
at org.apache.http.impl.client.DefaultRequestDirector.execute(DefaultRequestDirector.java:415)
at org.apache.http.impl.client.AbstractHttpClient.execute(AbstractHttpClient.java:641)
at org.apache.http.impl.client.AbstractHttpClient.execute(AbstractHttpClient.java:576)
at org.apache.http.impl.client.AbstractHttpClient.execute(AbstractHttpClient.java:554)
at org.openremote.controller.protocol.isy99.Isy99StatusReader$QueryTask.run(Unknown Source)
at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:471)
at java.util.concurrent.FutureTask$Sync.innerRunAndReset(FutureTask.java:351)
at java.util.concurrent.FutureTask.runAndReset(FutureTask.java:178)
at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$201(ScheduledThreadPoolExecutor.java:165)
at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:267)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1146)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:615)
at java.lang.Thread.run(Thread.java:679)
Caused by: java.net.ConnectException: Connection timed out
at java.net.PlainSocketImpl.socketConnect(Native Method)
at java.net.AbstractPlainSocketImpl.doConnect(AbstractPlainSocketImpl.java:327)
at java.net.AbstractPlainSocketImpl.connectToAddress(AbstractPlainSocketImpl.java:193)
at java.net.AbstractPlainSocketImpl.connect(AbstractPlainSocketImpl.java:180)
at java.net.SocksSocketImpl.connect(SocksSocketImpl.java:384)
at java.net.Socket.connect(Socket.java:546)
at org.apache.http.conn.scheme.PlainSocketFactory.connectSocket(PlainSocketFactory.java:123)
at org.apache.http.impl.conn.DefaultClientConnectionOperator.openConnection(DefaultClientConnectionOperator.java:123)
... 15 more

besides that I also constantly get
Aug 30, 2013 1:14:26 PM org.restlet.engine.http.connector.HttpClientHelper start
INFO: Starting the default HTTP client
Aug 30, 2013 1:14:30 PM org.restlet.engine.http.connector.HttpClientHelper stop
INFO: Stopping the default HTTP client
Aug 30, 2013 1:14:35 PM org.restlet.engine.http.connector.HttpClientHelper start
INFO: Starting the default HTTP client
Aug 30, 2013 1:14:37 PM org.restlet.engine.http.connector.HttpClientHelper stop
INFO: Stopping the default HTTP client
Aug 30, 2013 1:14:43 PM org.restlet.engine.http.connector.HttpClientHelper start
INFO: Starting the default HTTP client

But I read on the forum that's just a cosmetic error

Posted by garfield.arbuckle at Aug 30, 2013 12:18

Should not make a problem.
The ClientHelper start/stop should go away if you turn the above property to false.

Posted by mredeker at Aug 30, 2013 12:32

Ok I will give that a go, but what about the ISY99 device timeout?
Stijn

Posted by garfield.arbuckle at Aug 30, 2013 12:34

Ok so I disabled the beehivesynch again and indeed the ClientHelper start/stop errors are gone.
However still getting this

ERROR pool\-1\-thread\-1: IOException while reading data from ISY-99
org.apache.http.conn.HttpHostConnectException: Connection to http://192.168.41.12 refused
at org.apache.http.impl.conn.DefaultClientConnectionOperator.openConnection(DefaultClientConnectionOperator.java:127)
at org.apache.http.impl.conn.AbstractPoolEntry.open(AbstractPoolEntry.java:147)
at org.apache.http.impl.conn.AbstractPooledConnAdapter.open(AbstractPooledConnAdapter.java:108)
at org.apache.http.impl.client.DefaultRequestDirector.execute(DefaultRequestDirector.java:415)
at org.apache.http.impl.client.AbstractHttpClient.execute(AbstractHttpClient.java:641)
at org.apache.http.impl.client.AbstractHttpClient.execute(AbstractHttpClient.java:576)
at org.apache.http.impl.client.AbstractHttpClient.execute(AbstractHttpClient.java:554)
at org.openremote.controller.protocol.isy99.Isy99StatusReader$QueryTask.run(Unknown Source)
at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:471)
at java.util.concurrent.FutureTask$Sync.innerRunAndReset(FutureTask.java:351)
at java.util.concurrent.FutureTask.runAndReset(FutureTask.java:178)
at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$201(ScheduledThreadPoolExecutor.java:165)
at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:267)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1146)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:615)
at java.lang.Thread.run(Thread.java:679)
Caused by: java.net.ConnectException: Connection timed out
at java.net.PlainSocketImpl.socketConnect(Native Method)
at java.net.AbstractPlainSocketImpl.doConnect(AbstractPlainSocketImpl.java:327)
at java.net.AbstractPlainSocketImpl.connectToAddress(AbstractPlainSocketImpl.java:193)
at java.net.AbstractPlainSocketImpl.connect(AbstractPlainSocketImpl.java:180)
at java.net.SocksSocketImpl.connect(SocksSocketImpl.java:384)
at java.net.Socket.connect(Socket.java:546)
at org.apache.http.conn.scheme.PlainSocketFactory.connectSocket(PlainSocketFactory.java:123)
at org.apache.http.impl.conn.DefaultClientConnectionOperator.openConnection(DefaultClientConnectionOperator.java:123)
... 15 more

When trying to load the webconsole I get this

ERROR HTTP\-Thread\-4: failed to extract panel.xml for panel : No such Panel :NAME =
org.openremote.controller.exception.NoSuchPanelException: No such Panel :NAME =
at org.openremote.controller.service.impl.ProfileServiceImpl.getProfileDocumentByPanelName(Unknown Source)
at org.openremote.controller.service.impl.ProfileServiceImpl.getProfileByPanelName(Unknown Source)
at org.openremote.controller.rest.FindPanelByID.handleRequest(Unknown Source)
at org.openremote.controller.rest.RESTAPI.doPost(Unknown Source)
at org.openremote.controller.rest.RESTAPI.doGet(Unknown Source)
at javax.servlet.http.HttpServlet.service(HttpServlet.java:617)
at javax.servlet.http.HttpServlet.service(HttpServlet.java:717)
at org.apache.catalina.core.ApplicationFilterChain.internalDoFilter(ApplicationFilterChain.java:290)
at org.apache.catalina.core.ApplicationFilterChain.doFilter(ApplicationFilterChain.java:206)
at org.openremote.controller.rest.support.json.JSONCallbackFilter.doFilter(Unknown Source)
at org.apache.catalina.core.ApplicationFilterChain.internalDoFilter(ApplicationFilterChain.java:235)
at org.apache.catalina.core.ApplicationFilterChain.doFilter(ApplicationFilterChain.java:206)
at org.springframework.web.filter.CharacterEncodingFilter.doFilterInternal(CharacterEncodingFilter.java:78)
at org.springframework.web.filter.OncePerRequestFilter.doFilter(OncePerRequestFilter.java:77)
at org.apache.catalina.core.ApplicationFilterChain.internalDoFilter(ApplicationFilterChain.java:235)
at org.apache.catalina.core.ApplicationFilterChain.doFilter(ApplicationFilterChain.java:206)
at org.apache.catalina.core.StandardWrapperValve.invoke(StandardWrapperValve.java:233)
at org.apache.catalina.core.StandardContextValve.invoke(StandardContextValve.java:191)
at org.apache.catalina.core.StandardHostValve.invoke(StandardHostValve.java:128)
at org.apache.catalina.valves.ErrorReportValve.invoke(ErrorReportValve.java:102)
at org.apache.catalina.core.StandardEngineValve.invoke(StandardEngineValve.java:109)
at org.apache.catalina.connector.CoyoteAdapter.service(CoyoteAdapter.java:286)
at org.apache.coyote.http11.Http11Processor.process(Http11Processor.java:845)
at org.apache.coyote.http11.Http11Protocol$Http11ConnectionHandler.process(Http11Protocol.java:583)
at org.apache.tomcat.util.net.JIoEndpoint$SocketProcessor.run(JIoEndpoint.java:354)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1146)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:615)
at java.lang.Thread.run(Thread.java:679)

Posted by garfield.arbuckle at Aug 30, 2013 12:54

The "No such panel" is ok. The webconsole is doing a request without panel name to see if the controller is there.

Posted by mredeker at Aug 30, 2013 13:39

Ok,
is the ERROR pool-1-thread-1: IOException while reading data from ISY-99 normal?
Stijn

Posted by garfield.arbuckle at Aug 30, 2013 13:41

http://www.openremote.org/display/forums/Getting+strange+ISY-99+errors

Posted by pz1 at Aug 30, 2013 14:09

Thanks Pe Zanstra,
So you're suggesting I revert to 1.0.8?
Any idea where I can find that build? Only 1.0.10 is available on the resource download in the designer.
Cheers
Stijn

Posted by garfield.arbuckle at Aug 30, 2013 14:18

I only wanted to raise awareness that others did report this problem quite a while ago.
Sorry, I have no idea where you can get this old version. I do not know either if the error hurts. It only messes up the logs like a couple of other useless messages like the Panel thing and the Zwave & Vera messages.

Posted by pz1 at Aug 30, 2013 15:19

Ok thanks and no problem

Posted by garfield.arbuckle at Aug 30, 2013 15:21

At this stage, we're not keeping an archive of the older Pro versions, so only the current one is available for download.

We'll look into fixing the ISY error for the next Pro release.
Although the error does not cause any real harm to the controller, it might indeed degrade performance on low power hardware like the Raspberry.

Posted by ebariaux at Sep 02, 2013 12:22

I will try to run OR from my Ubuntu server, that has more than enough power.
Anyone any idea why the webconsole isn't working?

Posted by garfield.arbuckle at Sep 02, 2013 15:33

Ok so started up OR on my Ubuntu server which has a dual core atom and it's blistering fast.
In comparison to free OR on raspi it's also noticeably faster.
However I didn't get any feedback on the Android app, meaning I was able to turn the lights on but the status on the switch would not show as on so I couldn't turn it off. After a while I was also not able to turn the light on anymore.
Tried to revert back to the raspi however it is now also unable to connect to the KNX bus, spitting out the same error as the ubuntu controller.
So it seems both are now unable to connect to the KNX gateway and ETS is not connected either.
I've now turned the raspi off again and rebooted everything else (including the bus) and I'll let it sit overnight to see if they come to some sort of an agreement on connectivity during the night .
Stijn

Posted by garfield.arbuckle at Sep 02, 2013 23:05

KNX gateways always have a timeout until they release the connections.
If I play around with OpenRemote I have to unpower my gateway if I want to test quickly in a row.
I think OpenRemote is missing some sort of message to tell the IP gateway to free up the connection.
Maybe this is in already but is not called all the time.

Posted by mredeker at Sep 02, 2013 23:51

Indeed Marcus I have noticed the same. When I was using OR on the Raspi I couldn't commission the KNX devices anymore with ETS because the gateway only allows one connection. The strange thing was when I killed the OR service the KNX gateway still wouldn't release the session like you mention. However when I hard rebooted the Raspi itself the KNX gateway released the connection and I could commission again with ETS.
So to get back to the issue at hand. I left OR and the KNX gateway overnight but it seems they couldn't come to terms hence it still doesn't work.
Rebooted OR several times and rebooted the gateway yesterday evening. OR on the pi is off.
Ow yeah and just to be sure I tried if I still could commission via ETS and that was not a problem. So the KNX gateway was not being occupied by any other service...
This is getting complicated.
Ow yeah and the web console also doesn't work on the unbuntu server

PS: thanks all for the support.

Posted by garfield.arbuckle at Sep 03, 2013 10:27

I have now reproduced this same error and will check what the issue might be with Richard.

For me, the issue comes when I have a switch that is linked to a switch object. It seems that the sensor reference in the panel.xml is what makes the WebConsole fail.

Posted by ebariaux at Sep 03, 2013 13:00

Thanks a lot Eric.
I'm happy you were able to reproduce it. But linking a switch to a switch object is quite a common function isn't it?
If anyone has any more info on the KNX connection failing, please let me know.

Posted by garfield.arbuckle at Sep 03, 2013 15:24

Hi Stijn,

I will look at your webconsole issue tomorrow; the problem does not occur with the latest build of the webconsole but I need to confirm what the problem is and that the problem was deliberately fixed and not an oversight.

Rich

Posted by kurrazyman at Sep 03, 2013 22:18

Thanks a lot Richard,
much appreciated.
Stijn

Posted by garfield.arbuckle at Sep 03, 2013 22:20

Hi Stijn,

Right after some digging the problem you are experiencing was fixed in r7061; Eric is going to update the controller downloads with the latest webconsole build r9080, this can be found at the following link if you want it: -

https://www.dropbox.com/s/0xghxxl2cq2twkk/webconsole.war

Rich

Posted by kurrazyman at Sep 04, 2013 21:59

Hey Rich,
Thank a lot, the webconsole works now. The sliders look a bit strange though, not like on the android app, can't really remember if it was the same on the free controller or not.
They are basically fat and ugly , looking a bit windows 3.1 style. I'm using Safari on OSX. Is this normal?
In addition to this my KNX gateway and openremote stopped talking to eachother since I moved the OR pro version of the raspi to the ubuntu server. Now none of my OR versions (pro ubuntu, pro raspi, free raspi) can talk to the KNX bus.
I've tried rebooting everything. The log only spits out it can't connect to KNX. However the KNX gateway seems fine because ETS can still commission devices on the bus. I'm aware ETS and OR can't connect to the bus at the same time, but ETS has been off for a while and the gateway has been rebooted in the mean time so very unlikely ETS is still hogging the gateway.
Should I create another post for this?
Thanks
Stijn

Edit: won't load at all with FF23.01 on OSX: Controller error: Webconsole can't be reached. Make sure your controller is running and if it is secured check your browser supports HTML5 CORS. The controller is running as it loads fine on Safari, apart from the strange sliders .

Posted by garfield.arbuckle at Sep 04, 2013 23:12

Hmm, no one likes fat and ugly sliders. The latest styles should give you sliders that look the same as those on its. Will look at this to see what's going on and report back.

Regarding KNX issue, generally create a new thread but don't think your problem is with config or OR itself so can discuss here. Usually the KNX log is pretty good at describing what the problem is. Have you tried rebooting your IP interface? Sometimes ETS doesn't close the connection properly but you should at least see this in the KNX log. Also make sure your Ubuntu hosts.allow file is configured to allow access to anything on your LAN.

Rich

Posted by kurrazyman at Sep 05, 2013 06:17

Hey Rich,
I have rebooted the KNX gateway multiple times.
This is what I get in the logs.

INFO 2013-09-05 08:10:37,943 (KNX): KNX connection not available.
INFO 2013-09-05 08:10:37,944 (KNX): KNX connection not available.
INFO 2013-09-05 08:10:38,181 (KNX): KNX connection not available.
INFO 2013-09-05 08:10:38,243 (KNX): KNX connection not available.
INFO 2013-09-05 08:10:38,248 (KNX): KNX connection not available.
INFO 2013-09-05 08:10:38,254 (KNX): KNX connection not available.
INFO 2013-09-05 08:10:38,319 (KNX): KNX connection not available.
INFO 2013-09-05 08:10:38,322 (KNX): KNX connection not available.

in knx-debugging.log

INFO 2013-09-05 08:12:26,092 (KNX): KNX connection not available.
DEBUG 2013-09-05 08:12:26,093 (KNX): Polling device status for FRAME L_Data.req 0.0.0 -> 5/0/7 Data: 0x00
INFO 2013-09-05 08:12:26,093 (KNX): KNX connection not available.
DEBUG 2013-09-05 08:12:26,095 (KNX): Polling device status for FRAME L_Data.req 0.0.0 -> 1/5/4 Data: 0x00
INFO 2013-09-05 08:12:26,095 (KNX): KNX connection not available.

I never adjusted the hosts.allow file and I can ping the knx gateway, but I'll look into it closer.
Stijn

Posted by garfield.arbuckle at Sep 05, 2013 07:13

The Pro Controller v1.0.10 distribution that can be downloaded from your Pro account has now been updated to include the correct version of the WebConsole (but you can keep using the one from Rich's link if you want, nearly identical).

The eBox image has NOT been updated to include the correct version of the WebConsole. This will be done as part of Pro 1.0.11 release.

Posted by ebariaux at Sep 05, 2013 13:07

There should be a section earlier in the log where the KNX attempts to figure out which network interface to use. Do you have more than one NIC in your Ubuntu server? Can you find those log statements to see which IP interface the KNX attempts to use?

If it automatically would pick the wrong one, it may be necessary to fix the NIC IP address in the config.properties file.

Posted by juha at Sep 06, 2013 13:33

Hey Juha,
Yes I have multiple interfaces on the motherboard however only 1 is in use and the same issue occurs on the raspi which only has one interface.
Anyway I checked what you suggested and here is the output

WARN 2013-09-08 21:42:54,474 (KNX): Could not connect
org.openremote.controller.protocol.knx.ConnectionException: KNX-IP interface not found
at org.openremote.controller.protocol.knx.KNXIpConnectionManager.waitForConnect$
at org.openremote.controller.protocol.knx.KNXIpConnectionManager.getConnection($
at org.openremote.controller.protocol.knx.KNXIpConnectionManager$ConnectionTask$
at java.util.TimerThread.mainLoop(Timer.java:534)
at java.util.TimerThread.run(Timer.java:484)
INFO 2013-09-08 21:42:54,475 (KNX): Trying to create connection
INFO 2013-09-08 21:42:54,475 (KNX): Removing connection
INFO 2013-09-08 21:42:54,477 (KNX): KNX-IP socket listener IOException
java.net.SocketException: Socket closed
at java.net.PlainDatagramSocketImpl.receive0(Native Method)
at java.net.AbstractPlainDatagramSocketImpl.receive(AbstractPlainDatagramSocket$
at java.net.DatagramSocket.receive(DatagramSocket.java:729)
at org.openremote.controller.protocol.port.DatagramSocketPort.receive(Unknown S$
at org.openremote.controller.protocol.knx.ip.IpProcessor$PhysicalBusListener.ru$
WARN 2013-09-08 21:42:54,477 (KNX): KNX-IP socket listener stopped
INFO 2013-09-08 21:42:54,478 (KNX): KNX Connection manager resolving local host IP addr$
INFO 2013-09-08 21:42:54,479 (KNX): Found candidate NIC: name:eth0 (eth0)
INFO 2013-09-08 21:42:54,480 (KNX): Skipped IPv6 address (not supported by KNX) /fe80:0$
INFO 2013-09-08 21:42:54,480 (KNX): Added candidate IP address to set - /192.168.1.100
INFO 2013-09-08 21:42:54,480 (KNX): Skipping loopback interface: name:lo (lo)

I've tried adding knx.ip.interface.hostname=192.168.1.31 (IP of KNX gateway) to the config.properties but it still doesn't work.
When I add this I get more or less the same.

INFO 2013-09-08 23:10:18,377 (KNX): KNX IP interface hostname set to '192.168.1$
INFO 2013-09-08 23:10:18,387 (KNX): KNX IP interface port set to '3671'
INFO 2013-09-08 23:10:18,388 (KNX): KNX PhysicalBus clazz set to 'org.openremot$
INFO 2013-09-08 23:10:18,409 (KNX): Scheduled reconnection task
WARN 2013-09-08 23:10:18,410 (KNX): Unknown KNX property '<property name = "nam$
INFO 2013-09-08 23:10:18,429 (KNX): Trying to create connection
INFO 2013-09-08 23:10:18,442 (KNX): Removing connection
INFO 2013-09-08 23:10:18,479 (KNX): KNX Connection manager resolving local host$
INFO 2013-09-08 23:10:18,497 (KNX): Found candidate NIC: name:eth0 (eth0)
INFO 2013-09-08 23:10:18,498 (KNX): Skipped IPv6 address (not supported by KNX)$
INFO 2013-09-08 23:10:18,499 (KNX): Added candidate IP address to set - /192.16$
INFO 2013-09-08 23:10:18,500 (KNX): Skipping loopback interface: name:lo (lo)
INFO 2013-09-08 23:10:18,865 (KNX): Created KNX Command FRAME L_Data.req 0.0.$
WARN 2013-09-08 23:10:19,152 (KNX): Unknown KNX property '<property name = "nam$
... repetitive output omitted...
INFO 2013-09-08 23:10:20,892 (KNX): KNX connection not available.
INFO 2013-09-08 23:10:20,892 (KNX): KNX connection not available.
.
.
.
.

Posted by garfield.arbuckle at Sep 08, 2013 22:12

Hey Rich,
Just wanted to let you know the sliders look normal again and unfortunately I didn't change anything so I don't know why they were different first. If I would be able to reproduce the issue I will let you know.
Thanks for the support on the webconsole. Works fine now.
Now I just need to get OR to connect to the KNX gateway again... .
Stijn

Posted by garfield.arbuckle at Sep 08, 2013 22:25

Hey Everyone,
Quick update, eventually my ETS was unable to connect to the bus too.
Hence I turned my attention to my knx ip gateway, this kept refusing OR and ETS connections.
After some investigation together with the good people of ABB it turned out my KNX IP gateway was not functioning as it should.
They quickly provided me with a new unit and I hoped to be on my way again in the wonder world of OR.
However OR still won't connect to the ip gateway. ETS again has no issues connecting to the bus but OR can't.
I've started decoding the ip traffic with Wireshark.
I can see OR sending out traffic to the mcast address however the IP gateway doesn't reply.
I have noticed ETS also doesn't discover the gateway anymore however by putting in the ip address directly it can connect to it so it works fine.
I've tried putting the ip in the config.properties file as knx.ip.interface.hostname=192.168.1.101
Unfortunately I can only see OR send out mcast packets and no unicast packets.
The mcast on the network seems to be working fine as I can see other mcast packets from other devices.
The bizarrest thing is that all the misery started from the moment I moved OR Pro from the raspi to my ubuntu server.
On the raspi everything worked fine until I moved OR Pro to the ubuntu server for better performance.
From then on OR Pro and OR Free are unable to connect to the old or new KNX gateway from my raspi, my ubuntu server, my mac and my winxp.
Anyone any idea? Does anyone know how I can force OR to only use unicast to the specific ip of my knx gateway and not use mcast to search.
Thanks
Stijn

Posted by garfield.arbuckle at Sep 24, 2013 15:50

Hi Stijn,

Without looking at the source code for the KNX protocol; as far as I am aware if you specify a specific hostname in the config then the autodiscovery shouldn't occur. What messages are appearing in your KNX log file?

I have previously created a setup using exactly the same configuration as you (ABB IP Interface and OR running on Ubuntu server) without any problems.

Rich

Posted by kurrazyman at Sep 28, 2013 10:14

Hey Rich,
Thanks for confirming you had a no issues, to be honest I had no issues at all till the moment I moved my controller from the raspi to ubuntu server and now nothing works.
HOWEVER
I think I have found the problem. You might want to look in the KNX code to check if autodiscovery doesn't occur because I think it does even when the ip is set statically.
When analyzing the traffic I only saw mcast traffic going out of the OR controller and no unicast even when I set the IP interface as described above (maybe I did something wrong).
Because my ETS stopped showing the auto discovered ip interfaces I started to wonder if mcast would be the issue. But since I'm just using my dumb service provider router and I hadn't changed anything to it, I was dismissing it as unlikely.
At a point I was googling for some other info and I found out my provider had been doing upgrades to their brand new series of modems (ow yeah the one I have) and it was around the time I started having the issue.
So the mcast was back in the picture, so I fired up my ETS VM, installed OR on the VM and connected the IP interface directly to my laptop.
OR connected immediately to the bus and everything on the bus became controllable again via OR.
So shame on my for ever using service provider crap while having super equipment (Cisco) sitting in a box.
It had gotten pushed down on the way too long to do list but now has become a priority.
I will try to install it tomorrow and will let you know if my issue is solved, but I'm starting to be quite confident it will.
More to come and thanks at everyone who contributed to this long winding issue.
Stijn

Posted by garfield.arbuckle at Sep 29, 2013 22:26

Hey everyone,
VICTORY!!!
OR can communicate with the ABB IP Interface again.
I have deployed my router and am using the service provider crap as modem.
It does however take OR 5 to 10 minutes to connect with the IP interface, but once it was connected it seemed stable.

So the conclusion for anyone who started to read at the top and came to the bottom for a summary

  • "Web Console Panel definition incorrect": This was due to a bug in OR, Rich was kind enough to fix it so this shouldn't occur anymore.
  • ABB IP interface failure, after commissioning the interface with a static ip it failed and was unable to make a connection with the bus, thanks to the people of ABB for verifying the error and quickly replacing the module for me (ABB=best support ever)
  • When I had the new IP interface ETS was able to connect to the IP interface but OR wasn't. This was extremely strange since it had worked before. In the time I was troubleshooting the broken IP interface my provider had done an upgrade of the software on the modem/router model which I was using. This upgrade seems to contain a bug and because of this they broke multicast on the router which is needed for OR to find the IP interface. OR also has a setting to put the IP static however according to what I was able to find it never actually starts sending out unicast traffic to the IP interface as long as it hasn't found the IP interface via multicast. As far as I can see this setting more works as a selection in case there would be several ip interfaces discovered. Probably also why the setting is called knx.ip.interface.HOSTNAME. Anyway that's my guess .
    So I re-purposed my provider router as pure modem and unboxed my proper network equipment (which was long overdue anyway) and now everything works fine.

@Rich: Could you have a look at the code to see if there is an error in the way OR treats the statically set IP. If you want me to test anything just let me know. I will be more than willing to .

Ticket closed, thanks to everyone who contributed. Very much appreciated.
Stijn

Posted by garfield.arbuckle at Oct 01, 2013 10:21
Document generated by Confluence on Jun 05, 2016 09:42