yep me too..
Only a sniffer can tell you for sure.
It would seem to be hugely beneficial to have a bit more diagnostic information available (but, possibly, "hidden") directly from the hub itself. Having to buy a stick, hack it into becoming a mesh sniffer, then decoding and understanding it--just to figure out issues is a lot.
Plus, apparently, getting two sticks is almost necessary--one for sniffing, one for doing "dirty work" as a secondary controller when the Hub gets cranky??
2 Likes
Well the problem is that if you're trying to troubleshoot the traffic coming in and out of the hub, you will never know if it is a hub problem or an external device problem unless you can see the raw traffic...
Because even if the hub were reporting one thing, you still wouldn't know for sure if the hub itself is the problem.
2 Likes
For everything else, I agree that you should not need an external tool to troubleshoot.
3 Likes
True, to a degree.
However, having access to the deeper levels of raw data on the hub would be very helpful and could resolve a lot of the issues--with a separate device only needed if that wasn't adequate.
The hub currently has the "Z-Wave Logs" -- but those show SO little information right now. If there was a way to crank them up to show ALL the traffic (albeit, only for short time periods), that would seem very useful.
And, if you had such a tool that you knew DID show ALL the traffic the hub saw, then you'd know a lot, even without the external sniffer.
1 Like
Agreed. The new zwave log was intended to show all traffic, as I understood it. It just doesn't work. Lol
1 Like
I had an identical issue. Bobby from Hubitat asked me to find the devices that were spamming the busy messages. Not surprisingly, the culprits were two switches in the fringes of the basement. I have 65 devices in the house, so should have a strong mesh, but that area was not being served adequately by any device. I excluded the two culprits from the network, for a ring range extender which I out in between the culprits and known good nodes, shut the hub down, restarted, did a repair, and all seems to be working well now. I recommend searching out those bad nodes and excluding the devices from the network, rebooting and trying without them first before attempting to reinclude Fingers crossed.
From comments made by the staff, it sounds like waiting for 2.2.4 before doing anything too drastic might be a good thing.
Obviously, if you are able to correct mesh issues, that's always a good thing.
But, they are hinting that there will be major improvements in z-wave.
2 Likes
Yup. He told me the same. I think we just need a bit of patience. I will hold of on any major automations until it all settled down. It's an awesome platform, coming off the disastrous new Smartthings platform.
1 Like
I'm hitting the check for updates every hour or so now. Lol
I don't know if it's in beta yet or not, so I may be a bit premature. But, if I wanted patients, I'd have become a doctor!!!
I am using a couple of these to monitor temperature for fans and window units. I don't know exactly how often they update, but I do know it isn't fast enough for this purpose. To get around this I created a WebCoRE piston that sends a refresh command to them every 10 minutes.
Be careful about battery life, then.
1 Like
A quick update here.
I have not had to reboot my hub, or reset the z-wave radio in the last week. There are two notable changes since then:
- A number of my battery powered sensors have dead batteries; this is likely due to the longer wake windows that have been required in order to send status updates to the main hub.
- I have disabled most of my rule machine 4 logic roles that turn lights on or off as a result of PIR trips throughout my house.
I suspect that between the combination of the two, z-wave storms have been kept to a minimum; I also suspect that logic that has been activated from these sensors is partially to blame for the z-wave storms, where the hub may be attempting to send a large number of changes as a result of these sensors tripping or reporting their status.
The problem likely self exacerbates; a z-wave storm increases overall number of retry attempts for in-bound z-wave events from devices, and repeated messages are likely all acted on from the hub, sending out additional traffic.
Because I do not have a z-wave sniffer, this is anecdotal and should be taken with a grain of salt. For others who have been having z-wave stability issues, I recommend temporarily disabling any rule machine logic that is keyed off of sensor updates, and then examining in-bound z-wave sensor events to detect duplicate messages.
I usually see those duplicate messages on peoples systems with mesh issues or ghost devices in settings, zwave details.
2 Likes
Also consider updating a device to the newest firmware if there is any available.
As I mentioned in another thread I upgraded a few troublesome Zooz Zen23 switches and Zooz Zen 24 dimmers to the latest and greatest this seemed to reduce some of the issues - still need to verify this is working over the next few days.
The switches in question are on the edges of my mesh but are definitely not out of range of an extender and several other working switches so am speculating the device update has calmed things down a bit.
@BorrisTheCat, how many usb MS6's do you have?
I've had lots of issues with my zwave mesh on my C5 hub but I had presumed it was my other zwave devices that were the problem and so started transferring some USB powered MS6's to my new C7 yesterday and although the first 15 MS6s were excluded, factory reset, and then included to the c7 hub fine, the mesh on the new C7 hub seems s mesh and it won't even complete a zwave repair (it's says the nodes aren't reachable).
If you've got more than 15 and it's working fine for you then I'll need to do some more trouble shooting. I nearly cried when I saw the comments that the USB powered MS6s were bad repeaters - but you've given me some hope!
I was one of those complaining about the usb MS6's on my C-4. I had 6 at one time and after Zniffing (watching the Z-Wave packets being transmitted using a UZB-3 stick) I could see how things were routing and slowing devices routing through the sensors from 100/40kbits down to 9.6kbits. It was pretty bad back then..
I still have 3 MS6's and they are all on the "edge" of my network so are not repeating other devices. They are working for me on the C-7 so far.
I have 6 all powered. I have never zniffed them so I don't know, they don't seem to be causing any issues. Not for me anyway, they are all on the latest firmware.
I finally found those devices. I know it took me a while as we are doing major renovation at the moment and I haven't touched the hub at all until a week ago. I found a post on here (Hubitat forum) where people were complaining about Zooz 4-in-1 sensors spamming their networks on C-7 and I had two of them. Removed them and all is well now.
3 Likes
