Mesh Weirdness

I have a C7 (currently running v2.3.7.140) which I installed about 18 months ago. I built out my Z-Wave network as suggested in the Hubitat documentation, building out from the hub using powered devices, with the core of my network being Zooz switches and dimmers. I currently have 59 AC powered Z-Wave devices (54 of which are Zooz) all but a couple of which use 700 series chips with S2 Authentication and another 30 battery-powered devices, 10 of which are Bali/Somfy shades or remotes, 12 are Zooz devices, and the remainder are from a variety of other vendors. Most of my network devices are S2 Authenticated, but 2 use S0 and 22 use no security. I did have 3 ghost devices, but I recently removed them, so I am ghost-free.

My issue is that the routing shown Z-Wave details doesn't seem to be as stable as I would expect, especially for some devices. The thing I am actually having trouble with are a couple of Zooz tilt sensors which are battery powered and attached to garage doors. I have them mounted on the end of wooden extensions to keep them away from the metal door. The closest AC-powered devices (Zooz switches) are in the same garage about 10 and 30 feet away from one, and maybe 30 and 40 feet away from the other, through the garage (no walls in the way). I checked the routing on both of them last week (Jan 2) and they both were connected with 1 hop through one of those garage switches. But today, I had some activity that wasn't captured and I looked again and they both have a non-sensical route with 3 hops and the final hop being to different devices that are actually in the basement, one of which is all the way across the house, maybe 60-70 feet away. I also looked at a wall switch which is in a plastic box about 30 feet from the hub but on a different floor. Last week it was directly connected to the hub but today it has a 3 hop route.

What causes the mesh to reroute if I don't manually initiate a repair? Is here any way to force a route and/or keep it from changing for a particular device? Any suggestions on how to avoid problems like this from popping up?

Suggest you ignore the routing graphics and add another mains powered (routing) device somewhere near the devices where you have experienced the drops and see if that improves your experience.

I have one less than 10 feet away from one of the tilt sensors with clear line-of-sight (no obstructions. Couldn't get one closer than that if I tried.

Try a Zooz 800 Series Z-Wave Plus Range Extender ZAC38 - The Smartest House. They have a strong signal and have worked wonders for me solving routing issues with a section of my house which is a long distance from the hub.

But....do the rules work as they should?

Not sure what your question has to do with things, but yes, if the event is received by the hub, my rules, a combination of basic rules, button controller, and WebCore, work fine.

I may have to give the 800 series repeater a try if those garage sensors continue to be flakey. But, the device that is actually closest to one tilt sensor is a series 800 ZEN30 from Zooz and I have never seen a routing that connects to that device.

Sorry.

Periodically, when I have nothing else to do, I go through the routing and do manual repairs and try to get them to route direct to hub, if possible.

I just gave it a quick look, and most are still paired direct to the hub.

An interesting thing I noticed is the Aeotec home energy monitor, which sends a power reading every 5 seconds, running at 9.6 with a hop, even though it's about 6 feet away in the basement. I think I ran into something like this before, and someone noted the position of the device in relation to the hub antennas could be a factor. This sensor is almost directly below the hub, which has its antennae pointed straight up. Anyway, just an idea.

My Z-wave Ecolink garage tilt sensors are paired direct at 100, and that's including the one on the farthest (steel) door on the three car garage, and there are two walls, a metal shelf unit, and an enclosed breezeway in the way..

I'm thinking also since there's no car in the first bay anymore, I think there's less interference. I've noticed that with my ST presence sensor (Zigbee) as well.

I think it means a weak connection.

Well today, both of the tilt sensors are connected through a single device (2 hops). The devices are not the closest devices to the sensors, nor are they "half-way" between the sensors and the hub, but they are in reasonable locations between the sensors and the hub, so I guess they are probably OK.
Number of route changes is high and signal strength is so-so.


I reconnected my automation just so that I will notice if things start to go south again. But I wish there was a way to do a manual override of the mesh routing in certain cases.

Unfortunately that is not something Hubitat can control. We can all send cards and letters to Silicon Labs . . .