[2.2.4.158 C7] Z-Wave locked up

[rob9]

@bcopeland Not sure what happened, but this morning when I woke up and did my normal "disarm", it triggers the "Set At Home when Many Triggers" app. That turns on the right lights, unlocks the door into the garage, etc.

While it appeared to turn on my Hue strips (thru Hue, not zigbee) and unlock the garage door (Z-wave), only the "Entryway Chandelier" and "Family Room Ceiling Light" actually were turned on. So, it appears the Z-Wave "locked up" in the middle of all this and quit issuing commands.

Of note, the Entryway Chandelier and Family Room Ceiling lights were the first 2 things to be turned on, followed by the LED Status lights, then the HUE lightstrips.

In looking at the logs, I'm not even seeing entries for most of those devices (e.g., "Master Bedroom Ceiling Light"). So, despite the app supposedly issuing that command, nothing really got sent.

No event in that device's event log:

But, no other substantial errors or warnings appear in the logs.

Also, the routine that sets the LED Strips on my Inovelli seemed to run--and that did issue device commands (z-wave). However, none of them actually received/acted on those commands (they appeared to be lost).

Another app ran to set a lightstrip and two z-wave devices. The lightstrip came on (HUE integration), but the Z-Wave lights did not. This ran a few seconds after the "Set At Home..." app, when the Z-Wave things seemed locked up.

Now, after waiting a while, I am able to issue commands to the Z-Wave lights that do work.

In fact, retriggering the "Set at home" actions did seem to make everything turn on properly this time. It's just clear that everything was really hosed up for a bit.

[kahn-hubitat]

You are doing a lot of crap at once... I have seen similiar if one of the devices is failing or at the time was not responding and starts trying to find a new route the mesh sometimes can get locked up till it is resolved.. No real way to see that in the logs, unless maybe you had zwave logs running.

It is especially likely if the node having issues is a repeater or switch that a bunch of your other devices you are controlling is routing through.

Would be nice if there was a "show past logs" for the zwave logs.

[kahn-hubitat]

it also could be the issue here ... doesn't have to be ecolink as i understand could be any battery device

[2.2.4] Automations using Ecolink sensors delayed or unresponsive - Support / Feedback - Hubitat

[rob9]

@kahn-hubitat

it also could be the issue here

It seems unlikely. The rules fired and got past any contact sensors, etc.

It just stopped doing things suddenly.

[kahn-hubitat]

in those cases the contacts sensors fire and the rules also fire, but then the mesh locks up and responses are delayed or never happen.. so similiar. But most likely all devices would have failed or be delayed not just some.

[lewis.heidrick]

You could easily test it by changing all the zwave contact sensors to use the Ring Alarm Conact Sensor driver until @bcopeland puts in the fix in 2.2.5. It's a very plausible cause in this case given all the crap you're triggering and how the contact sensor spams the mesh trying to update the battery status.

[2.2.4] Automations using Ecolink (and other battery powered) sensors delayed or unresponsive

There was a change in some of the generic drivers that was intended to enforce battery updates for some stubborn devices that refuse to update battery % in a decent time frame.. The consequence being that it was sent after the device was already asleep causing a 15+ second flood of activity trying to get the message to the device through every conceivable route and the through explorer frames. This was sent if the device sent a sensor update and there was no battery update in 12 hours.

[rob9]

@lewis.heidrick

Ring Alarm Conact Sensor

All my contact sensors are Ring G2 sensors.

[lewis.heidrick]

I don't think the fix is applied in the G2 version but @bcopeland could confirm.

[rob9]

Although, I am setting my Ring G2 Extender sending a ton of battery status messages in a row, for some reason.

[eric.dalquist]

I'm seeing issues on 2.2.4.158/C7 as well. I see a lot of hub log activity, for devices I don't think arer actually that active, very little zwave log activity, and most of my devices are not responding to ZWave commands.

[lewis.heidrick]

Check settings, zwave details and make sure there aren't any devices without anything in the clusters column (ghost, failed pair). Safe shutdown and pull power for a minute then power it back up and see if things are back to normal. Change any battery contact sensors to the Ring Alarm Contact Sensor driver and see if things clear up.

[rob9]

Everything was working fine 15 min or so later

[lewis.heidrick]

Yes, thats how that bug behaves. It floods the mesh then is good for 12 to 18 hours or whatever till a contact sensor is triggered again that hasn't reported lately. Or just wait for it to get fixed in 2.2.5.

[rob9]

It rarely happens, so I'd be hard pressed to know if it's "cleared up" or not.

I do have 2-4 devices (5-15%) that either don't respond or doing update their state during group on/off events (with optimization off). So, there are still some issues, but it's getting better.

I noticed as zooz multi-relay with no cluster entries, but the children are triggering events.

[eric.dalquist]

I had two orphaned nodes from bad joins that I hadn't got around to removing. I did that via a ZWave stick and the PC Controller software and things seem to be behaving better.

[rob9]

Interesting--seems to have done the same thing this morning (just a few lights came on).