#Hi Everyone

1 messages · Page 1 of 1 (latest)

outer dome
#

I managed to capture the moment a light turned on in the ZHA debug logs which can be seen here. The light in question has Nwk: 0xec80. Unfortunately I dont understand the log output well, could anyone point out the reason why this specific light turned on?

#

The light in HA says it turned on at exaxtly 11:10:38

sterile topaz
#

I think you'll have to wait for the bigger brains to see this, I have no idea sorry

outer dome
#

@sterile topaz lets continue here.
Yes I have now unplugged the coordinator from power to see if the lights turn on even without it, and in 2 hours nothing happened, will wait a few more since the behaviour is random

sterile topaz
#

so it's receiving commands

#

as in, it seems to be coming from the coordinator

outer dome
#

well I cant confirm that yet.
thank you. Its veeery odd though and makes me so angry. I dont understand exactly how within a Zigbee network you can send a command to a light to turn on, like could there bre a malfunctioning device that is sending random commands? I'm so confused

sterile topaz
#

seems very strange indeed

#

however unlikely that it is a power issue, what happens when the bulb is off at 80% brightness and the power cut is very brief from off to on

#

actually, nevermind that

outer dome
#

I would take this into consideration, were this not to be so random that currently it affects various lights (GU10 220v powered lights, light strips powered via a separate trafo to 12v and also other lights that have a trafo to 24v. It would make so little sense that its a power issue because they are all on different circuits, though its possible, its really highly unlikely )

#

(Also I have once circuit where 4 of the same lights are connected together. Only one of the 4 would turn on at a random time)

glossy laurel
toxic epoch
#

I’ll take a look at the log when I get back to computer I’m on mobile most of this week while away.

outer dome
#

I think I might have solved the issue unintentionally. I power cycled all the lights I have and since then, none have turned on for the past 12 hours. What I struggle to understand, however, is how is this possible? Like I'm fairly certain that it cannot be that all the lights started turning on at random due to a fault in each light, instead it was probably one thing that was signalling them to turn on, no?

Can a Zibgee network get one bad actor that can mess the whole thing up? Does anyone know?

toxic epoch
#

Check the on power startup state (if the bulbs are zigbee 3.0) . It’s possible the bulb “rebooted” and came on from that I suppose. Nwk conflicts will cause a device to rejoin silently (in most cases, sengleds eventually fail to do that). I don’t think the z-stack fw reports conflicts though and just deals with them internally. You se them with ezsp coordinators. I have a group of 2 bulbs come on frequently on its own but I attribute that to flakiness from a non-neutral inovelli blue which controls them and the power to them 🤷🏼‍♂️

glossy laurel
#

my network is still "healing" itself i believe too

outer dome
#

I tested the lights on reboot and they are configured to power on remember last state and I tried the power cycle and it worked fine, so I'm excluding a power issue here. I do think that something was messing up the network though. So Now I know next time this happens I will just power cycle all the lights in the house

glossy laurel
#

@toxic epoch atDuck: OK, so in my efforts to figure out my zigbee network, I removed some devices to re-add them back. However ZHA isn't even finding them to be resdiscovered.
[10:28 AM]CatDuck: I wish I could just tell my network to start from scratch, but now I have a few devices that won't re-interview. (it is not a reception issue, new devices paired perfectly)

#

currently looking how to clear any memory from the controller, debating just deleting the entire thing, lol

#

if i need to wipe it, to be ablet o get it ready for matter/thread it would be a good time, or debating buying another, one for z2m and keep this one for matter/thread in future

toxic epoch
#

you can use the migrate flow to set up a new network and reset the coordinator.

glossy laurel
#

sorry, probably beyond your scope

toxic epoch
#

I’d look at the ezsp counters

#

There is one that will tell you about interference

glossy laurel
#

so its not interference,

#

new devices, "perfection"

#

its like something remains of devices removed

toxic epoch
#

It’s the phy_cc_fail_count

#

I’m not really sure. If you turned off source routing it should clear out the stored source routes in the coordinator.

glossy laurel
#

Yea, I started with 2 devices, now I'm at 50 and growing,