Zwave : Status.Level vs ZWaveNode.Basic : Bug ?

[nolio]

Hi,
I have 2 Aeon lab aeotec multisensor. And i got a problem on the first one some weeks later. Tomorrow, i have the same problem on the second one.

The problem, my alarm trigger with no real reason (no move).

With the 2 multisensor, i got something like this :

Code: [Select]

20150324 xx:09:36.4604800 ; Detectxx ; ZWaveNode.Basic ; 255
20150324 xx:13:35.6964290 ; Detectxx ; ZWaveNode.Basic ; 0
20150324 xx:13:36.1620110 ; Detectxx ; Status.Level ; 0Instead of usually :

Code: [Select]

20150324 xx:33:57.5923260 ; Detectxx ; ZWaveNode.Basic ; 255
20150324 xx:33:57.6414700 ; Detectxx ; Status.Level ; 1
20150324 xx:37:57.5569650 ; Detectxx ; ZWaveNode.Basic ; 0
20150324 xx:37:57.5853450 ; Detectxx ; Status.Level ; 02 weird things :

• So the Status.Level ; 1 seems lost ....
• None of the first 3 line have a real move in the room

But i see in the "90 Security Alarm System" code, this parameter :

Code: [Select]

"Status.Level"
"Sensor.Generic"
"Sensor.Tamper"
"Sensor.MotionDetect"
"Sensor.DoorWindow"
"Sensor.Flood"
"Sensor.Smoke"
"Sensor.Heat"
"Sensor.CarbonMonoxide"
"Sensor.CarbonDioxide" But not the ZWaveNode.Basic ...

So a lot of questions and not a lot of answer ... And very difficult to reproduce this bug.

I write a little bash script to survey if it happen again, and this appear one with one month between first multi and second one ...

the script :

Code: [Select]

#!/bin/bash
echo -n `date +"%x-%X"`// >> /home/pi/test
echo -n `grep "DetectSalon " /mnt/xxx/xx/xx-LastEvent.log | grep -E 'ZWaveNode.Basic|Status.Level' | grep "ZWaveNode.Basic ; 255" | wc -l`/ >> /home/pi/test
echo -n `grep "DetectSalon " /mnt/xxx/xx/xx-LastEvent.log | grep -E 'ZWaveNode.Basic|Status.Level' | grep "Status.Level ; 1" | wc -l`// >> /home/pi/test
echo -n `grep "DetectGarage " /mnt/xxx/xx/xx-LastEvent.log | grep -E 'ZWaveNode.Basic|Status.Level' | grep "ZWaveNode.Basic ; 255" | wc -l`/ >> /home/pi/test
echo -n `grep "DetectGarage " /mnt/xxx/xx/xx-LastEvent.log | grep -E 'ZWaveNode.Basic|Status.Level' | grep "Status.Level ; 1" | wc -l`// >> /home/pi/test

I unplug a zwave device, that i suppose to interfere the zwave signal ...

If someone as any answer, i would appreciate a lot to understand.

Bye

[nolio]

Hi,
An update on the subject.
My raspberry pi B was a little bit overload, i suppose it can be part of the problem. But after my update to raspberry pi 2 B, the problem is still there (less often) ...

I can see that there is the same "lost" in the HG log :

Code: [Select]

[07:xx:xx.575672] SPI > xx xx xx xx xx xx 03 20 01 FF 30
[07:xx:xx.007646] SPI > xx xx xx xx xx xx 03 20 01 00 CF
[08:xx:xx.034937] SPI > xx xx xx xx xx xx 03 20 01 00 CF (lost of a "FF 30" ??)
[08:xx:xx.097726] SPI > xx xx xx xx xx xx 03 20 01 FF 30

I am not sure where is the lost :

• Zwave network ...
• HomeGenie

I will probably try to do a fresh install from a new raspbian image ...

Bye

[nolio]

I found another good option, i have sometimes error to write my log file "Sharing violation on path /xx/xx/xxx.log". So it's probably just because the raspberry pi 2 is too fast and have to write the log file too quickly.

The good news, the raspberry pi 2 solve probably my problem !

[nolio]

But often, i need to do a "reset" of "Remote Modules Routing" in "Maintenance" menu ...

[nolio]

Does someone see something like this ? Perhaps a weird behavior ?

i think about 2 options :

• Do a new install of HG
• Something of my neighbor interfere with my zwave signal ...

[nolio]

Hi,
Usually, this problem didn't appear by regularly "restart the deported modules" , but since someday it isn't enough ....
I have now a lot a "status.level" lost (perhaps some other zwave but it's harder to take a look to other).

@Gene, did you think homegenie use acknowledgement (like : http://www.domoticz.com/forum/viewtopic.php?f=6&t=5034) ?

About Zwave, we usually say that is it safe (therefore X10 or other) by using confirmation, but in my configuration, it doesn't seems to be true ....

I try to take a look at health of my zwave network (with HG and other software), but it seems to be ok ... So i don't know what i can do to solve that ...
In the HG log, the status lost doesn't appear so it can be lost by :

• HG ?
• Zwave Network ?
• ?

Bye

[dani]

Hi Nolio,

I asked Gene to add a status node error in the node parameter now it's in rrom in the last revision.
We receive a Status.Error timeout message for each node.
I have adjust all current widgets to care of that status and show the error state in HG UI screen.
For that I test the date of last event and compare with the date of Status.Error event, and I have vreate an error image.
If you are interresting I can join all the widgets and image.

But yes the ack in Gene ZWave is test after the status is changed. But by that we know the node is on error.
 
Cheers
Dani

[nolio]

Hi Dani,

I understand that your "error" message can only be detected, in case the request is initialize by the zwave controller.
But in my case, the request is initialize by several zwave equipment ...

A history capture from HomeGenie Plus App in attachement. This appear several times during one day ...

Bye

[nolio]

Hi,
Since 5 days, no more lost... Probably my z-wave network go healthy. :-)
Bye

[bkenobi]

This is fairly common with X10 unfortunately.  I guess it's good to see that ZWave users expect to have perfect command reception so that is basically the standard.  For us X10 users, we have to bolster our system with filters/repeaters/etc in order to get 99%.