peterathans

@hart2hart I've got some follow up questions - you mentioned "I had to run it many times to get mine sorted", can you expand on this? What made you discover that you had to do this multiple times? What were your metrics for success? Did you have some specific performance issues that you were trying to correct? How did you measure or check that each Heal process was making an improvement or change? I followed my process by watching the Level 3 event viewer, but you can also see each device's link table in the X-Ray viewer (although keeping track of these tables for all devices can get very cumbersome). Did you have cases where the Heal process just didn't work for specific devices, and if so, how did you fix those? Sorry to pester with so many questions, thanks for any help you can provide!

Remember, the time officially changes at 2am, so any sunrise/sunset calculations made at midnight (if this is how the controller works) will be off by an hour that day. Not trying to excuse this behavior, just pointing out one possibility. I'm also in the camp of "this has been going on for years, why can't the controller take it into account?".

Here are the nodes that show up when the device is added:

Yes it pairs no problem. I have hardwired and interconnected First Alert smoke and CO detectors in my house, and the ZEN55 outputs signals for both types of alarm.

Yes I have one, it works very well. Any specific questions?

I've got three switches acting as a 4-way switch, linked in a scene using basic association 2. I'm noticing some strange responses in the event viewer when using Polisy to turn the scene on or off. When I turn the scene off, what I'm seeing below is that three Turn Off commands go out to the switches (first three lines transmit, next three lines transmit done), then two of the switches respond with a D2D EVENT status=0 (switches turning off) and then the second switch also responds with D2D-CMP responses. The third switch never responds, which is what led me to look at the logs (probably a different issue). I'm wondering why the responses differ, these three switches are all the same Zooz Zen71 On/Off switches and are all running the latest 3.70 firmware. I have also confirmed their configuration parameters match, with the only non-default parameter being parameter 7 set to 1. Wed 02/11/2026 10:31:55 AM : [ZW-TX-A 00017 33.0 ] [ZY033_1] Binary Switch Set val=0 duration=255 Wed 02/11/2026 10:31:55 AM : [ZW-TX-A 00018 34.0 ] [ZY034_1] Binary Switch Set val=0 duration=255 Wed 02/11/2026 10:31:55 AM : [ZW-TX-A 00019 19.0 ] [ZY019_1] Binary Switch Set val=0 duration=255 Wed 02/11/2026 10:31:55 AM : [ZW-DONE 00017 33.0 ] [ZY033_1] Binary Switch Set val=0 duration=255 Wed 02/11/2026 10:31:55 AM : [ZW-DONE 00018 34.0 ] [ZY034_1] Binary Switch Set val=0 duration=255 Wed 02/11/2026 10:31:55 AM : [ZW-DONE 00019 19.0 ] [ZY019_1] Binary Switch Set val=0 duration=255 Wed 02/11/2026 10:31:55 AM : [D2D EVENT ] Event [ZY019_1] [ST] [0] uom=78 prec=0 Wed 02/11/2026 10:31:55 AM : [ZY019_1 ] ST 0 (uom=78 prec=0) Wed 02/11/2026 10:31:55 AM : [D2D EVENT ] Event [ZY033_1] [ST] [0] uom=78 prec=0 Wed 02/11/2026 10:31:55 AM : [ZY033_1 ] ST 0 (uom=78 prec=0) Wed 02/11/2026 10:31:55 AM : [D2D-CMP 0030] STS [ZY033_1] ST op=1 Event(val=0 uom=78 prec=0) is Condition(val=100 uom=78 prec=0) --> false Wed 02/11/2026 10:31:55 AM : [D2D-CMP 0006] STS [ZY033_1] ST op=1 Event(val=0 uom=78 prec=0) is Condition(val=100 uom=78 prec=0) --> false

I know IOS 6 came out a month or two ago and was waiting for others to be the guinea pigs before upgrading, but since then haven't heard much about it. I'm running Polisy 5.9.1 and am wondering if it's worth it (and safe) to upgrade to 6.0?

Thanks, that really worked! I first chose the overall Heal Z-Wave Network option and watched it work through the event viewer. Looks like it did a thorough job of stepping through each device, building a connection list, and then deleting the old device neighbor lists and rewriting them, via several iterations. I did see there were some devices that were not reachable on the first iteration or two, but those got better with further attempts. I do think there were a couple of devices that showed as "not awake" even though they are standard Zooz hardwired light switches (and which otherwise work normally). Later I found that each device has a Heal Z-Wave Network option so tried those on individual devices (namely the ones above that didn't get updated with the system-wide procedure), but strangely it looks like the Neighbor list was appended rather than replaced. In other words, those devices used to show Neighbors that I think are incorrect (i.e. neighbors that are very far away from the device). After the Heal procedure, those incorrect neighbors still exist, but there are new entries in the table with (correct-looking) neighbors too. The total number of entries in the table grew. I'm wondering if those old, incorrect entries could cause problems by asking the device to talk with neighbors who are too far away? @Michel Kohanim can you confirm details about how the Heal Z-Wave Network process works differently on a per-device bases vs. the overall network?

Does anyone know if there's a way to force a ZWave device to rebuild its "neighbors" table? I installed a couple of devices a few weeks back, out in the back yard, close to each other and farther from the house (but still within range of a few other devices). Before that I set them up (added them to Polisy) at my desk. I looked at the data.neighbours table today and it shows a lot of neighbors within the house, but not neighbors of each other, which I assume they should be (since they're the closest two devices to each other). They will respond to Polisy commands but sometimes are slow, so I know they are somewhat within range but believe they can be better. I've heard that the ZWave devices are "self-healing" but it's been a few weeks now and am wondering if I should do something to make them rebuild the neighbor tables.

Way back when I got my ISY-99, there was a default program that existed before I added my own logic, called Query All. It still exists through my upgrade to Polisy, which I'm running now. The program seems to (currently) consist of a single instruction, to "Set 'ISY' Query" at 3am. Does anyone know what this is supposed to accomplish? I'm asking because I have a Zooz ZEN16 with one input monitoring a N/C microswitch rain sensor, and I occasionally find that the module will report that the rain sensor detected rain at precisely this time, perhaps a couple of times a year (and no it's not raining when this happens). It almost appears that this Query program forces the input module to report input status and for some reason the module claims to not see voltage when it actually is present. I know the input state changes at this time because I send an email when the rain sensor enters and exits rain detection. I had never connected the dots and suspected that this Query action might be related to the bogus input update, until going through the programs today and noticing this 3am Query being the same time as the 3am rain sensor trigger. Anyone have thoughts?

I use one of these to monitor a rain switch, you should be able to use an In as a status in your program.

Thanks for the feedback. Zwave network is great - the holiday modules aren't installed yet (new features being added this season) and the network has been great for years. So these new/temporary modules will only make it better (at least until January when they're removed).

Unfortunately the logic to turn the modules on and off is combined with logic to control devices that exist year-round. And I'd prefer to not make two separate programs. Hence the original question. I don't want to leave the modules installed (Zooz Zen04) because, when the holiday lights aren't up, they don't pass the WAF test.

Not sure if this a general Polisy question or specific to Zwave, so asking it here. I've got some Zwave outlet modules that are only used for holiday decor and are unplugged and stored away the rest of the year. During this time, I'll have some programs that will send on/off commands as usual, to both the missing modules as well as to other modules that remain plugged in year-round. Anyone know if there is a system-wide impact caused by Polisy sending commands to a Zwave module that doesn't exist? I'm not worried about reported errors for the missing module, but am worried about impacts on the rest of the system, like hanging programs or Zwave command delays to other modules.

I've done half of the solution you're talking about - basically I have the same Zooz ZEN16 with one input tied to a magnetic switch (Used this one from Amazon) to monitor whether the door is closed or not. I don't feel comfortable yet giving control of the door to my home controller, just because I don't see it as 100% reliable. So I'm not using a relay to "press the button".