Since migrating from C5 to C7 I have had a number of weird issues - not sure why, but here is one.
I am having trouble doing ZWave Repair on multiple devices. Looking at Hubitat Z-Wave Mesh Details I have found a lot of devices that have zero neighbors (which doesn't make sense). I notice a bunch have unknown for RTT Avg and LWR RSSI and zero or null in neighbor count. And some of those with no neighbors are not direct from hub? Weird. I tried (don't yell at me now) doing a ZWave repair - but it logs "starting" then never does anything - which is likely one problem. But today I tried doing repairs on a couple of devices and the repair fails. I can control the devices - so that's a plus. Here is a shot of the logs from two devices I tried.
Here is a shot from the mesh details - just guessing, but if a device is 100% red across it must be bad?
This is not unheard of.. It won’t have a lot of these details available on devices that don’t communicate often.. That gets refreshed when a packet is sent from the hub, the return ack sends this data.
The timing out on repair updating node neighbors is probably not a concern here either as I see the nodes have a lot of neighbors.. More than I have seen before, I may need to increase this timeout.. But that timeout is only affecting the UI presentation as the SDK is doing the work here and will continue even with that timeout..
I'm concerned I cannot run a full ZWave repair at all - I know, it shouldn't be needed, but in the past that was the first answer given.
Also, even doing one device repair - it seems to nearly stop all processing. I just ran one against node 84 (failed) and when I walked through my house and tripped two motion sensors there was ~10 second delay in response. I feel like my ZWave network is just screaming and choking on its own responses. This is MUCH worse after C5 to C7 migration - not sure why.
Yea.. I would avoid doing repairs, at least for now.. The massive amount of neighbors will cause a massive amount of traffic for each node repair.. Unless you are having problems before running the repair..
The main reason I'm chasing ZWave is because I am getting activity reports of devices not reporting for > 24 hours - even with refreshes sent every 6 hours. Something is definitely wrong and all started with C7 migration.
Great questions. I am using bptworld Device Watchdog (and have been for a very long time). I have the Activity Report triggering once every 6 hours and reports on devices inactive for > 24 hours. When it runs, it sends a refresh to any device that has had no activity in > 6 ours and waits to see if the device responds or not then reports on those who have not responded. This is a great tool to identify when devices actually go offline (or used to).
Since my migration I have had a ton of devices start showing up on the report. If I manually trigger it again once or twice (as the list > 6 hours shrinks) the remaining few DO report back and are then good another 24 hours.
Since this started after the migration - I have been looking at commonalities (all zwave, zigbee are fine) and then have been going through troubleshooting zwave things. Interesting the migration instructions for C5 to C7 explicitly call for full ZWave Repair before starting (preferably a day before) and it worked fine C5. Now ZWave Repair just won't work on single devices and never even moves if I start a full repair (have waited an hour with no log entries and had to power down and back up to make it stop).
As to "working" or not - I believe every device DOES respond to on/off and then properly reports new latest activity info. And as I mentioned above, running the report a second/third time DOES clear the stragglers and get their activity back. But this was never a problem before - it seemed every device (that reports) always had something to say within a 24 hour period - my feeling is traffic is falling off and not getting back to the hub.
Also of note (I think) I had 6 devices report at 3pm (PDT) today. I'm trying zwave repair on them individually and the repairs do fail. They are also far from the hub (though you see I have a TON of repeaters).
I had to chop off the bottom of that screenshot - was taller than my 4k monitor. I have 122 ZW devices and 51 Zigbee devices (like no wonder nothing answers, right?)
I thought the max was 200+ or so. That said, is the C7 more fragile than the C5 was? This all started when I migrated.
So what you're really saying is I should have 3 hubs dedicated to ZWave? Assuming I add one or two more - is there any "easy" way to move ZW devices from one to the other? HP worked great to move them "all" - but manually removing, adding, then putting back in dashboards and rules would be - tough.
232 is spec maximum. But spec and real world are two different things. This has nothing to do with the hub -- only with Z-Wave itself. A larger mesh is vulnerable, and the fragility goes up with size.
Zigbee is very different. One guy had 400 devices working, but then broke it up into multiple hubs.
oh well i have 99 zwave devices.. is it really better to get a 2nd hub.. aren't they also going to be issues with crosstalk/interference. with two hubs in the same location.
your zwve devices must be a lot closer to each other than mine.. as my topology table looks nothing like yours.
On the other hand, some people enjoy obsessing about things, and Device Watchdog is the perfect accessory for that. Lot's of reported problems are from Device Watchdog, that turn out actually not to be problems at all. Hmmm, "watch dog", "crying wolf"... write to support, etc.
i am using device activity check not device watchdog.. but never had any false alarms only actual battery devices that dont check in very often and it found dead batteries for me more than once.. but i have the parameters quite wide for expected activity
