Recent Zwave problems - significant delay/lag

Hardware
, ,
1

So a few weeks ago I started getting a significant delay when executing a script that turns off 4 Z-Wave switches at the same time. I’ve not noticed more than 1-2 seconds in the past (system has been up for ~9 months), but this is taking 4-8 seconds (possibly 16 according to a log entry I noticed).

I tried adding 0.1 second delays to the command, however it doesn’t seem to have made a difference. I still see the following error log entries that stand out to me:

Start of command:

2025-02-11T21:20:22.639Z DRIVER » [Node 006] [REQ] [SendDataBridge]

First error @ almost 3 seconds - these or similar repeat for 2 seconds

2025-02-11T21:20:25.473Z DRIVER   Dropping message with invalid payload
2025-02-11T21:20:25.473Z DRIVER « [Node 010] [REQ] [BridgeApplicationCommand]
                                  │ RSSI: -76 dBm
                                  └─[Security2CCMessageEncapsulation] [INVALID]
                                      error: Duplicate command (sequence number 255)

A different error at almost 5 seconds

  • 2025-02-11T21:20:28.161Z CNTRLR [Node 003] failed to decode the message, retrying with SPAN extension…

There are a bunch more of the invalid payload errors, and then @ 19 seconds we get:

2025-02-11 13:20:41.345 ERROR Z-WAVE-SERVER: The node failed to decode the message. (ZW1405)
ZWaveError: The node failed to decode the message. (ZW1405)
    at Driver.sendMessage (/snapshot/build/index.js:275348:25)
    at Driver.sendCommandInternal (/snapshot/build/index.js:275472:31)
    at Driver.sendSupervisedCommand (/snapshot/build/index.js:275497:29)
    at Driver.sendCommand (/snapshot/build/index.js:275525:34)
    at Proxy.set (/snapshot/build/index.js:225089:24)
    at Proxy.set (/snapshot/build/index.js:224869:34)
    at Proxy.<anonymous> (/snapshot/build/index.js:225128:23)
    at ZWaveNode3.setValue (/snapshot/build/index.js:278424:44)
    at NodeMessageHandler.handle (/snapshot/build/index.js:291556:37)
    at Client.receiveMessage (/snapshot/build/index.js:292920:92)

What I’ve tried so far:

  • updated Z-Wave JS UI to current (zwave-js-ui: 9.30.1, zwave-js: 14.3.8)
  • updated HA Core to 2025.2.2
  • individually Rebuilt Routes and for the whole network, multiple times.
  • tried assigning priority routes to bot a few, and most hardwired nodes in an order that seemed to make sense to me (all cleared now (except 2 as it didn’t seem to help)

image
image754×862 44 KB

How should I troubleshoot this further?

Thanks in advance!

Here’s the script I ran, nothing special

Script 
sequence:
  - action: light.turn_off
    metadata: {}
    data: {}
    target:
      entity_id:
        - light.living_room_hall_cans
  - delay:
      hours: 0
      minutes: 0
      seconds: 0
      milliseconds: 100
  - action: light.turn_off
    metadata: {}
    data: {}
    target:
      entity_id:
        - light.living_room_cans
  - delay:
      hours: 0
      minutes: 0
      seconds: 0
      milliseconds: 100
  - action: light.turn_off
    metadata: {}
    data: {}
    target:
      entity_id: light.dining_room_chandelier_group
  - delay:
      hours: 0
      minutes: 0
      seconds: 0
      milliseconds: 100
  - action: light.turn_off
    metadata: {}
    data: {}
    target:
      entity_id:
        - light.bar_pendants
  - delay:
      hours: 0
      minutes: 0
      seconds: 0
      milliseconds: 100
  - action: light.turn_off
    metadata: {}
    data: {}
    target:
      entity_id:
        - light.kitchen_cans
alias: Turn off Main Floor Lights w/Delays
description: ""

And finally, the raw log for the command from Z-Wave JS UI Debug:

2

So, the only thing I can think of that may have changed is the update to Core 2025.2.2 I restored a backup that brought me back to

  • Core2025.1.0
  • Supervisor2025.02.0
  • Operating System14.1

I think the delay of the lights is back down to a second or two. This time the invalid payload only shows up once, and the log for the same script is 50% as long:

3

Hi all, just bumping this up with some new info.

I downgraded to 2025.1.1 for a couple weeks and it all seemed to work normally. All 5 lights turn off within a second or two of each other. As of a couple days ago I updated to 2025.2.5 and I’m back to having the same issue again.

Does anyone else have a setup with 5 zwave switches in a group on all current software that they can test with?

4

This seems related.