Hi, I'm experiencing an infuriating problem starting a few days ago, in which the Z-Wave radio stops sending commands to devices after it has been running for about a day. Statuses update in hubitat if physically toggle a switch, etc. Only way to get Z-Wave working again is to shut down, unplug for a while, then reconnect to power. I have seen threads by other users explaining the same symptoms that were found caused by ghost nodes or spammy devices -- however I see no ghost nodes and have not added any devices in weeks. Z-Wave logs are empty. Is there not any way on this hub to see exactly WHY commands are not being sent? What further can be done to fix?
Can you associate anything you might have done to the installation around the time this started happening? New device, etc?
I notice you have a number of devices that haven't reported in for a while, ie, the last report date is blank. I'd see if you can get that squared away.
Maybe by trying to get them to report, you'll come up with a bad actor. It only takes one.
It takes a while after startup for some, but they've now all populated. The only change I can think of is that I broke my phone three days ago so it's no longer running the hubitat app... but obviously that is not a Z-Wave device.
I fired up z-wave logs for the first time in a long while. Quite different than it used to be. Quite lively. But, I'm on a c8pro with zwjs, latest update. Sample:
Thanks for confirming. And Z-Wave logs stay empty while you turn lights on/off at the switch, or try to command them from the Device pages?
You could install the Hub Info Driver from Hubitat Package Manager (assuming you have HPM installed, or install Hub Info Driver manually) and use the Hub Info Driver to monitor hub memory... your hub may be running down memory really quickly for some reason.
When you get in the bad state, what do you see when you look at Logs (Live & Past), and App Stats and Device Stats? Go to those pages and refresh them to update the info and see if your hub Device or App load is unusually high.
Let me say I'm no expert, for sure.
I suppose it would appear, but would that take down a mesh network? I don't know. Not necessarily.
Can't be a bad sign they're all reporting.
Have you also tried a reboot w/advanced option of rebuilding database? Can't hurt.
I enabled description and debug logging on some suspect devices and found an HSM200 that was sending motion updates approx every second. Memory and CPU were fine. I changed drivers back and forth between “EzMultiPli new” and the built-in driver, and eventually it settled down (no restart of hub or device). May be resolved but I will continue to monitor and post back if I have more info to share. Thanks for your help!
