I've been on a 2+ year journey with Hubitat, and love the community and support staff. However, I've run into some fairly serious ZWave performance problems. I've had the usual 'ghost devices issue' (since fixed), and have even moved all of my ZWave Plus devices to a new C7 hub (keeping my old C5 hub for Zigbee locks, Dashboard, and Philips Hue integration).
What I've run into now is a challenge familiar to some of you - my 24 Aeotec Multisensor 6's appear to be flooding my network and causing slowdowns at seemingly random times. These are not reporting slowdowns - I've trimmed all functions except motion way back so as not to flood the network further. What I'm seeing now thanks to setting up ZWave sniffing (thanks community for help with that) is what appear to be singlecast 'storms' of the NOP Power command.
I wouldn't normally be super concerned, except these correlate in the logs with ZWave network busy messages and visible slowdowns I see when ZWave commands either take > 20 second to arrive at devices (mainly Aeotec Nano Dimmers) or get lost completely.
All but 8 of these MS6's are on mains-power, running firmware version 1.13. So, my question to all of you is: Am I basically hosed until the new MS7 comes out and hopefully addresses poor MS6 repeating and apparently this issue, or is there something I can do to alleviate this?
Before anyone suggests another multisensor, I have to say I'm in a bit of a bind, as the WAF of the physical dimensions of the sensors, plus the installation of all of the mains-powered ones in our walls with the Aeotec recessor mounts means that convincing her to accept another sensor, plus patching holes in the walls, etc. isn't going to be pretty.
Any thoughts, suggestions, etc. gratefully accepted. Thanks everyone!
I haven't done a ZWave full repair in months (on the advice of @bobbyD). There have been one or two single node repairs, never on these MS6's. As @njanda states, 8 of the 24 are battery powered, but what is strange is that these NOP Power storms are happening on some of those nodes as well, and in fact, seem to be contributing (not surprisingly) to increased power drain - those battery MS6's experiencing this phenomenon require more frequent battery changes.
I have a ton of ZWave packet logs like the one I posted, and in the one I posted, it's happening at 1 pm in the afternoon. I have tried to determine a pattern of time, and haven't been able to - it can be fine for several hours with no slowdowns, and then be bad for 15 minutes to several hours.
Do you have a link to a specific thread about this? I have been interested to get some data like this but did not want to waste time and money going down the wrong path.
Same - thanks @Brandon for pulling the link in - was going to do it, then got sidetracked trying to see what other logs/etc. I could pull on my issue - @bcopeland, any ideas?
Yes.. I have experienced this as well.. Right after I started zniffering my network for the first time, I decided to trash all of these from my network..
These being mains powered especially.. When mains powered they also act as repeaters and they make terrible repeaters… You can have them mains powered and not be repeaters, you have to include them on battery and then switch to mains power..
The latest I have seen from a thread with @csteele is that the latest firmware no longer lets you include them on battery then switch to mains - they apparently act as repeaters based on the voltage detected now.
Also, if I'm seeing NOP storms on my battery powered MS6s (with associated power drain), I'm not sure just trying to fix the mains-powered ones would work?
Ok, I can definitely try that - in order to not lose these in my automations, would it work to disconnect them from mains, refresh and then use the Z-Wave replace? Or do I need to remove them from automations, hard remove them and then switch?
I started with a remove of the single mains-powered MS6 that isn't in any automations (so far). Removed it cleanly, full hardware reset on it, then re-included it on battery. Using @csteele's driver, it shows as powerSource: battery. Once I took the batteries out and powered it via USB, his driver reports it as powerSource: dc.
It does seem that there are some confusing schools of thought as to whether trying the remove/reinclude as battery but power it via DC actually turns off the repeating function. I guess I'll need to pick one that is repeating according to my Zniffer logs and try that. Unless anyone else has conclusive evidence that trying this trick to get the MS6 to stop being a repeater works... BTW - these are all on firmware v1.14.
Not confusing at all.. It’s clearly laid out in the standard.. If it is a repeater it’s part of the NIF and that can’t be changed without excluding and re-including.
