Seeing odd polling traffic on my z-wave network


I’ve started noticing some performance issues with my z-wave network these last few days. In particular, one GE 14291 switch stopped sending instant updates to HA when turned on/off (though it would respond to commands from HA). So I removed the switch and reset it - that was a process - and it seems to be working fine again.

However, I’ve now noticed some big lags in my system which indicates heavy traffic. I checked the OZW logs and I noticed a GE 14294 dimmer being polled every few minutes. Here is a log excerpt from 4:30 am:

2018-03-30 04:35:51.096 Detail, Node012, Polling: COMMAND_CLASS_SWITCH_MULTILEVEL index = 0 instance = 1 (poll queue has 0 messages)
2018-03-30 04:35:51.096 Detail, Node012, Queuing (Poll) SwitchMultilevelCmd_Get (Node=12): 0x01, 0x09, 0x00, 0x13, 0x0c, 0x02, 0x26, 0x02, 0x25, 0x84, 0x6e
2018-03-30 04:35:51.096 Detail,
2018-03-30 04:35:51.096 Info, Node012, Sending (Poll) message (Callback ID=0x84, Expected Reply=0x04) - SwitchMultilevelCmd_Get (Node=12): 0x01, 0x09, 0x00, 0x13, 0x0c, 0x02, 0x26, 0x02, 0x25, 0x84, 0x6e
2018-03-30 04:35:51.443 Detail, Node012,   Received: 0x01, 0x04, 0x01, 0x13, 0x01, 0xe8
2018-03-30 04:35:51.443 Detail, Node012,   ZW_SEND_DATA delivered to Z-Wave stack
2018-03-30 04:35:51.655 Detail, Node012,   Received: 0x01, 0x07, 0x00, 0x13, 0x84, 0x00, 0x00, 0x16, 0x79
2018-03-30 04:35:51.655 Detail, Node012,   ZW_SEND_DATA Request with callback ID 0x84 received (expected 0x84)
2018-03-30 04:35:51.655 Info, Node012, Request RTT 558 Average Request RTT 522
2018-03-30 04:35:51.655 Detail,   Expected callbackId was received
2018-03-30 04:35:51.794 Detail, Node012,   Received: 0x01, 0x09, 0x00, 0x04, 0x00, 0x0c, 0x03, 0x26, 0x03, 0x00, 0xd8
2018-03-30 04:35:51.794 Detail,
2018-03-30 04:35:51.794 Info, Node012, Response RTT 697 Average Response RTT 660
2018-03-30 04:35:51.794 Info, Node012, Received SwitchMultiLevel report: level=0
2018-03-30 04:35:51.794 Detail, Node012, Refreshed Value: old value=0, new value=0, type=byte
2018-03-30 04:35:51.794 Detail, Node012, Changes to this value are not verified
2018-03-30 04:35:51.794 Detail, Node012,   Expected reply and command class was received
2018-03-30 04:35:51.794 Detail, Node012,   Message transaction complete
2018-03-30 04:35:51.794 Detail,
2018-03-30 04:35:51.794 Detail, Node012, Removing current message
2018-03-30 04:35:51.794 Detail, Node012, Notification: ValueChanged

I only have polling setup for one of my switches on 40+ device network and it is an EATON dimmer that I had to use and the polling_intensity is set to 15.

Does anyone have any suggestions? it seems odd that this device is suddenly starting to flood the network and I vaguely recall having the same issue with another dimmer a few months ago - I think a reset and re-add fixed it.

Would adding a custom entry in the zwave_device_config.yaml with polling_intensity: 0 help resolve this issue? I’m just trying to figure out first what causes this random issue and second, the quickest fix.

Thanks for the help,

I’ve also just started seeing dead nodes pop-up, some near the receiver, some far away. I am now working on the notion that this is related to HA updates.

I restored HA to a known working version, 0.63.3 (pretty old) and I don’t have any dead nodes appearing and I don’t see any unusual traffic. It appears post 0.63.3 the developers fixed the issue of the GE switches toggling back to the on state when they were turned of, I didn’t see this at all with 0.65.6 (I haven’t updated yet to 0.66.0). I am now wondering if this fix caused OZW to be more sensitive to failed communications with nodes and caused my dead node issue.

Does anyone know of any z-wave related updates or changes post 0.63.3?