My struggle with Zwave.. HEEEEELP

Setup, Configuration and Use Runtime
,
46

In reality, lowering the wakeup won’t really help much. If you change from never to 1 week, it might reduce life by a few days. If you change from a week to a day, again, you might loose a small amount, but it won’t be noticable - even dropping down to an hour will only make a small amount of difference to life. As you start to reduce further, you will then see significant reduction.

I use 1 hour on my devices and get 1 1/2 to 2 years life on devices like door sensors and motion sensors…

This will be interesting, thanks - can you grab a debug logfile as well please :wink:

47

Maybe you have already added this, @jeroenvdwaal but just for your convenience:
To get a separate z-wave.log in OH2 add the lines below to : runtime/karaf/etc/org.ops4j.pax.logging.cfg

log4j.logger.org.openhab.binding.zwave = DEBUG, zwave

# File appender - zwave.log
log4j.appender.zwave=org.apache.log4j.RollingFileAppender
log4j.appender.zwave.layout=org.apache.log4j.PatternLayout
log4j.appender.zwave.layout.ConversionPattern=%d{yyyy-MM-dd HH:mm:ss.SSS} [%-5.5p] [%-36.36c] - %m%n
log4j.appender.zwave.file=${openhab.logdir}/zwave.log
log4j.appender.zwave.append=true
log4j.appender.zwave.maxFileSize=10MB
log4j.appender.zwave.maxBackupIndex=3
# no zwave logging in openhab.log
log4j.additivity.org.openhab.binding.zwave = false
48

Hi,

Thanks for logging settings! I did some testing with the latest cloudbees build. I was able to add a new Thing, called called WALLC-S. So that is nice to start with.

2016-09-19 14_09_21-Paper UI.png1173×142 8.74 KB

It is recognized with the correct configuration: for four groups A, B, C and D. Three groups are sending scene numbers and the other group is associated with a fibaro dimmer.

2016-09-19 14_10_06-New notification.png1531×946 62.8 KB

The next image shows the group info. It seems that the group info is not shown? Scene sending groups are associated with the ZME_USB1 z-wave stick.

2016-09-19 14_10_26-New notification.png1185×282 5.54 KB

After the device is accepted as a Thing, the following appears:

I am not sure what to expect. I think that there should only one scene activation entry. When I push the different switches on the WALLC-S, both the scene activation numbers are changing simultaneously.
All scenes of the three groups are reported.
The dimmer configuration seems to be correct. However it is not updated when I push the buttons. In the log some error warning is shown:

2016-09-19 14_16_59-pi@raspberrypi_ _opt_openhab-2.0.0-snapshot.png999×620 16.3 KB

49

Now I am totally lost… how did you get to the point that node15 is recognized as a WALLC-S?
Did you do a manual wake-up of the WALLC?

Hope you can also provide the zwave.log files, to see where difference are…

50

@dries_dokter: Sorry for a short hijack of this thread: When I use your commands to get a separate zwave.log, I indeed will get an own log, but (nearly) all zwave commands are also still logged in the openhab.cfg. Can you confirm this?

Until now I only have made a restart of the openhab service. Maybe I need a full reboot?

51

Solution: take a look here:

2 Likes
52

That did the trick! Thanks netmask guy! :wink:

53

… :grinning:

54

[quote=“jeroenvdwaal, post:48, topic:14113, full:true”]
Hi,

Thanks for logging settings! I did some testing with the latest cloudbees build.[/quote]

Please note that Chris would want a DEBUG log, not an INFO log… (Changed that in my original reply too )

55

Hi dries_dokter,

The device was recognized after a manual wake-up. First as an unknown device. After the wake-up it became a WALLC-S. Tomorrow I will repeat the test and provide a zwave log file.

56

Thanks for that Jeroen!

57

Did it change from unknown to known immediately? At my end this only works if I remove the Thing and re-discover…

58

I hate to say it, but it’s topics like this that make me wary of trying zwave rather than just simple Wi-Fi devices that have always just worked for me.

I wonder what actual proportion of zwave installs are so problematic? I’d this just a rare case, or is it just really hit and miss? I’ve never once had problems with wemo or wireless tag devices.

59

I think it is either: me doing something wrong / something not quite as it should be in OH2 / both at the same time. Nothing to do with the principle of zwave.

Looking at how things work in OH1.8 (without any problems, repeatable) I gladly recommend zwave devices to anyone.
Looking at how things could work in OH2: it looks even better than OH1.8… Great thing about zwave: you have all these devices from different manufacturers that all use the same protocol…

But then again, the great thing about OH is: you can pick whatever you want and integrate it to your system… :slight_smile:

1 Like
60

So, I think the problem is still what I said earlier - when the NIF is received, the binding treats this as a wakeup and sends the request to set the parameter. I don’t think that the device is really awake at this stage, so it isn’t received. In OH1, when the WAKEUP is received, the binding retries the message immediately and we get a response, but in OH2, it doesn’t retry and it times out.

It’s the only difference I can see. In both cases the binding sends exactly the same data, so I don’t think it’s related to what is sent, but more likely, when it is sent.

61

@chris, I understand your analysis. Looking at the log, that indeed appears to be happening…
But how, in your opinion, would I (we) deal with this? What is your proposal on that (if any)?

62

I knew you would ask this, but at the moment I don’t have one. I didn’t add anything in my initial response as I need to think about it and look at the code.

We could remove the use of the NIF as a wakeup, but this was added for a reason as I think some devices didn’t send the wakeup! I might do this though as a proof of concept at least so that we can see if this is really the problem - I’ll let you know when I get a chance to look at the code.

63

I tried to ask the question as neutral as possible making sure not to sound demanding :smiley:

I have come-up with another work-around:
added 2 WALLC-S switches using openhab1.8 and set the wake-up time to 1 hour. Then switched back to OH2 hoping that when I come home tonight both devices are working fine, so I can set the parameters I want, wait at most one hour and then continue adapting my rules to make use of those switches…

If you have anything for me to test, to create a log from etc.: just let me know and I will…

64

No problem ;).

A log of this would be good - if we can prove that this works when the NIF is not sent first that would be good as it will help to decide how to proceed.

Clearly this needs to be fixed properly - its a question of how is best without causing a problem with a different device…

65

I am back with a another test on the WALLC-S. First of all, I like to thank Chris for its contribution to the OpenHAB software! Your work is contributing every day, especially when the lights are going on and off :slight_smile:

So I am trying to help here on the WALLC-S challenge. Yesterday I was able to get my WALLC-S recognized. Today the test is repeated, using a clean install. I was not able to get my device recognized:

2.0.0-SNAPSHOT
Build #487 (from cloudbees)

Started with offline version, nothing installed yet.
Adding log4j, separate zwave logging file
Extensions->Bindings->Z-Wave binding
Things->"+"->Z-Wave binding->Z-Wave Serial Controller-> Serial Port:(/dev/ttyACM0)

Controller shows ‘Online’

This is my configuration:
Z_Wave Node 1: ZME USB stick
Z-Wave Node 6: FGS221 Double Relay Switch 2x1.5kW
Z-Wave Node 7: FGD211 Universal Dimmer 500W
Z-Wave Node 8: FGD211 Universal Dimmer 500W
Z-Wave Node 9: WALLC-S Controller*)
Z-Wave Node 15: WALLC-S Controller*)
Z-Wave Node 17: FGRM222 Roller Shutter*)
Z-Wave Node 18: FGD212 Dimmer 2
Z-Wave Node 21: WALLC-S Controller*)
Z-Wave Node 24: NS210 Smart Powernode*)
Z-Wave Node 22: PAN04 In Wall Dual Relay(1 Way) Switch Module 2x 1.5kW with power meter
Z-Wave Node 28: FGBS001 Universal Binary Sensor

*) not recognized by OH2, node 17 identified by (010F:0301:1001:22.22)

After refresh and search again
4 devices are reported as ‘Unknown zwave device’
Node 9, 15, 21, 24
1 device is reported without proper name:
Z-Wave Node 17 (010F:0301:1001:22.22)

Removed all unknown devices and searched again (14:53 in the log file):
Did not change the outcome.

I tried to get the WALLC-S recognized again. But that failed. Woke up the
device several times and it did not work.

Created two zwave log files. Changed the extension, otherwise the file was rejected by the forum system.

zwave.log.1.7z.pdf (434.8 KB)
zwave.7z.pdf (312.9 KB)