Blue Series 2-1 Firmware Changelog | VZM31-SN

@EricM_Inovelli I have an issue on 2.18 and I am not sure how to best debug it so its kind of circumstantial at the moment. Every time I put my coordinator in pairing mode nearly all of my blue’s start the interview and initialization process. Near 30 of them, I am also seeing a lot of zigbee broadcast traffic and I know you helped remedy an issue early on where Zigbee Broadcast and messaging was inefficient. Is there some way to control or further dig into this? I’ve spent a lot of time debugging my network lately and am starting to suspect its 2.18 as that is the largest change I’ve made in the past year.

Following up, it appears many of the switches keep rebooting which causes my lights to turn on at all hours of the day. I think the reinstalling and configuring occurs after each reboot. There has to be some correlation with heavy traffic and reboots to the devices that is causing this.

For the broadcast traffic, can you post any logs? Do you see a lot of these:

Or these:

I have a very large Z2M network and over 100 of the 2-1 switches. I removed all the reporting for the 2-1 switches as that caused a lot of un-needed traffic for me and I have no issues with broadcasts now. I also use a high quality coordinator from TubesZB that has kept my network performing well.

What is your coordinator? What version of firmware? What Zigbee platform are you running?

@cscott ZHA, I’ve used both the ZBdongle-e and zbdongle-p. I’m on all firmwares to get it working. How did you turn off all the reporting?

@EricM_Inovelli is there a good way to pull that? I just look in the HA UI and I always see every inovelli broadcasting.

@cscott I was able to finally set all reporting on the blues to 0 for parameters 18,19,20. It took several passes due to instability. I think the first pass I was able to do 15, then 5, etc up until I was able to read/set the attribute on all 30 switches. Network seems a little bit better but am going to let it settle. Which coordinator from Tubes did you end up getting? I was looking for a PoE coordinator anyway so now is a good time.

I have the MGM24…newer chip with more performance and memory. Supports my 340 device Z2M network. I have done a lot of tweaking etc…to get to that device limit, but anything under 250 should be amazing on that chip from my experience.

@EricM_Inovelli Tons of these:

0x77f7 and 0x9d11 are Inovelli 2-1 Blue switchs

What device is at this address?
image

That is my coordinator, a ZBDongle-E currently with a modified firmware from puddly to increase broadcast group sizes for better queueing.

Since I can’t post a link here is a screenshot:

I would run the latest from here:

That was one of the problems. Running that stock, zha could never initialize with constant network busy errors. It’s why I went to the modified firmware.

That being said I was using Zigbee groups not Light Groups for commands with Adaptive lighting and didn’t have Inovelli reporting off so I’m sure it was getting spammed more. I’ll give it a try again but something was or is spamming the network.

Alright, debugged this some more. That e8:e0 address doesn’t exist. I think I need to repair the Innovelli’s to the address of my coordinator. That e8:e0 was showing up as a coordinator but wasn’t actually it. Like a phantom coordinator.

1 Like

@johnlento maybe you are seeing something like this? Integrating Zigbee Canless Lights For Entire Home - Juno Connect - Switches / General Discussion - Inovelli Community

Anyone that uses Zigbee2MQTT I have made a significant change (as requested by the repo owner) to the converter and it is currently in the dev version. If anyone is running dev or would like to try it please let me know if you notice anything strange.

Hopefully if anything is broken it can be caught before it makes it to production next month.

Inovelli - Adding custom device specific cluster to converter by InovelliUSA · Pull Request #7822 · Koenkk/zigbee-herdsman-converters (github.com)

To go along with the above post, the specific cluster has been removed from the zigbee-herdman configuration so hopefully everything is still working. My tests say yes, but if anyone notices anything let me know.

Inovelli - Removing manuSpecific Inovelli because it was moved to the… by InovelliUSA · Pull Request #1161 · Koenkk/zigbee-herdsman (github.com)

1 Like

@EricM_Inovelli

Receiving a lot of these errors now:

z2m: Publish 'set' 'brightness' to 'study_vanity_lights' failed: 'Error: Cluster with name 'manuSpecificInovelli' does not exist'

It’s always set brightness but is on most groups that are bound with the VZM31-SN

Are you running dev or production? Also, you don’t have any custom converters do you?

I just noticed the same errors as @bgreet in my logs. In my case, this is a Z2M group with Hue bulbs and a VZM31-SN. Full disclosure, this happened while I was pushing a firmware update to one of the group’s bulbs, not sure if that’s related. I’m on Production, Z2M v1.40.0, no custom converters. Intending to update to 1.40.1-1 in the next day or two.

z2m: Publish 'set' 'brightness' to 'Upstairs Bath lights- Sconces' failed: 'Error: Cluster with name 'manuSpecificInovelli' does not exist'

Edit:
Just noticed the same error on another group that does not have a bulb actively getting updated. Similar group though, VZM31-SN and Hue bulbs.

z2m: Publish 'set' 'brightness' to 'Kitchen lights- Ceiling' failed: 'Error: Cluster with name 'manuSpecificInovelli' does not exist'

Here is my information. Seems like all groups that have an inovelli switch on them are affected. I’ve included a sample log