Charmed Quark Systems, Ltd. - Support Forums and Community

Full Version: Official 5.3 Beta Discussion Thread
You're currently viewing a stripped down version of our content. View the full version with proper formatting.
Is anyone testing this with a different primary controller than smarthings?
Same thing. I'm starting to think that maybe the secondary controller flag just isn't going to be always enabled. I was reading some of the finer print and if there is a SUC in the network (an update controller), then it implies that the distinction between primary and secondary controllers goes away and maybe that flag isn't set. So Maybe I shouldn't require it.

And there's another call I can make that maybe is better for explicitly checking the controller type. Let me do that. I'll put up another drop here in a bit. I need to get some supper.
(03-31-2018, 02:45 PM)potts.mike Wrote: [ -> ]Is anyone testing this with a different primary controller than smarthings?

IÔÇÖve just had a quick attempt at getting it working with my Fibaro HC2, and IÔÇÖm experiencing the same result. It never shows as being in the ZWave network, but reports as having replicated successfully.
Is the HC2 a master controller?
OK, one more time with 5.2.904. I'm using a different call to decide if I'm in the network or not, which seems to be a more appropriate one that will be more accurate. The one I was calling before should work, but maybe it doesn't get set correctly until further into the process or something.

BTW, one encourage thing is that in all of the traces there's no issues with msg I/O. All of the comms are working really well, which could have been something that would only happen in my setup here or something.
Ok, on 5.2.904 I think we have made progress.

There were messages on the include window about joining the secure zwave network. After about 30-45 seconds of it completing, I see devices showing up on the driver client. The driver also shows InZWNetwork as true.

Approve new units is still grayed out but it looks like it is still trying to gather details on everything. I have 20+ units and the things are still chugging along and the trace file is growing 10 mins after starting the include and first joining the zwave network.

I am going to let it do its thing and go finish painting our kitchen. I will check back after Im done. I will let you know how it goes after the devices are all polled.

Dean, you are the man!!
If it's taking that long, then I'd imagine there are failures and it's retrying. To avoid being abusive it will typically wait 15 seconds before retrying a failed step. So it could take a while if there are failures.

I'd go ahead and just close the trace and send me what you have, so I can see what is going on. Anything that's likely to happen (other than you needing to make battery powered ones wake up potentially) will already happened. If it's big zip it up. It should zip down substantially.
I attached the trace file as it has been chugging along, and a screenshot of the units and their states.

I know I have 2 maybe 3 one way switches that are offline, 1 battery device (aeon minimote), and the VRC0P. The rest should all be switches or dimmers. Most all are Leviton, about half are ZWave Plus, the others are older gen Vizia RF ones, but they should all have the capability to report the state back. I have a few Cooper in wall dimmers, and maybe 1 or 2 dumb GE switches from when I first started.
OK, just looking at the driver capture, most of them were ready to be told what they are. For those that are in WaitDevInfo, click the left column item for them and select the query device info. Save each one to a file that indicates the device type and id, so Leviton_BR549_14 or something like that. I'll need to know the exactly model number generally since there's often not way to go from the manufacturer ids to find the model. I'll have to have the model to go look it up and see if it matches the ids we get.

A number of others seem to be stuck on getting the configured node name. If the unit supports naming, I'll ask for the name and try to use that as the initial name if it's possible. I don't think an of mine do, so I've not actually seen it work yet. So there may be something wrong with that. I should see that in the trace and can figure out what's wrong with them.
Yeh, it looks like there's an issue where if the getting of the name fails (because it supports it but no name was ever set) it gets stuck. I'll look at that.