forum

I've been using the razberry for a 2 months now. Ever since this week, i'm having problems. These problems being that the lights are either still on at moments they should be off or off at the moments they should be on.

When this occurs, the following situation exists:

* i can still connect to the web-interface (expert UI).
* my domotica software (domoticz) is still working fine. Controlling CoCo stuff works fine, controlling Z-wave stuff does not work.
* Restarting (sudo service Z-way) Z-way does not help
* The logfiles in /var/log don't tell me anythis usefull (Z-way.log latest entry is from 2013-11-09 10:58:07.147 and z-way-server.log is empty.

I'm running 1.4.1-rc1 dated 2013-10-21 on a raspberry firmware 44dc4f093c9ec075e6704965665311a94f6371aa with domoticz 1.1175

Update: I just found out that rebooting the z-wave chip (from within the expert UI) works. When i do that, i regain all functionality.

I now have 2 questions:

1. Is there a way to fix this?
2. A a temporary fix, how can i reboot the z-wave chip from the command line so i can schedule is with cron a couple times a day? I don't want to reboot the system every six hours or so

Update2: i found the soft reset functionality in the documentation, i think. Can anyone confirm that i can reboot the chip with http://IP-OF-RAZBERRY:8083/ZWaveAPI/Run ... ftReset(1)

It seems that it can support only a few commands before crashing. I had a cron every minutes to get a binary switch state.

And after only 30-40 minutes z-way see the device as dead, which is not.

I have a cron running to update the state too, but indeed it crashes after some time. Maybe i do too much queries indeed.

In the query script i do this for every device or instance on the device):

wget --quiet "http://10.0.0.29:8083/ZWaveAPI/Run/devi ... .Meter.Get()" -O /dev/null
wget --quiet "http://10.0.0.29:8083/ZWaveAPI/Run/devi ... Binary.Get()" -O /dev/null

I have 3 single plugs and 4 smart plug 6 devices, so that makes 54 queries in total. Every 2 min was the initial idea, to have the 'current' state. For now, i reduced the interval; i now query every 30 minutes.

See also http://www.domoticz.com/forum/viewtopic ... wave#p2958

Can you rerpoduce a crash with some minimal number of devices involved? We have never seen any crash for more than 6 months and thousands of users! Please reproduce it and we will try to do same following your instructions.