C-7 Zigbee behavior on 2.3.5.105

Last week I updated my idle C-7 to 2.3.5.105.

No apps, no Z-Wave; started with no devices at all. Joined 3 Zigbee devices, GE dimmer plug, Iris V2 motioin and Iris V2 contact.

Shutdown and powered off C-7 Friday afternoon. Those Zigbee devices joined to it were then reset and rejoined to C-3 .

On C-7 powerup this morning (with no Zigbee devices at all communicating with C-7), child table shows:

Child Data
child:[null, 52DC, type:EMBER_SLEEPY_END_DEVICE]
child:[null, 4161, type:EMBER_SLEEPY_END_DEVICE]
child:[v2 contact zz, 8B6D, type:EMBER_SLEEPY_END_DEVICE]

None of these devices were capable of communicating with C-7, yet its child table remains populated with devices that weren't connected to it since yesterday, and still remain after C-7's having been unplugged from power for twelve hours.

C-7 Zigbee join behavior is now similar to C-8; both C-8 and C-7 don't issue new device ID's to joining devices (unlike other routers in the mesh which follow the stochastic scheme).

Even more unusual on C-7, non-existent devices evidently never get purged from its child table. (on C-3 and C-8 on .105, similar Iris V2 child devices that no longer communicate get aged out of the table within a couple of minutes).

I reset and rejoined the Iris V2 contact sensor that was joined to it a couple of days before (it still appeared in the C-7's device list); I got the same behavior which results when this is done on a C-8 (joined directly via the hub, and not a router as an intermediary handling the join request): its blinking blue light stopped during the 'add device' process, indicating that it was communicating with the C-7. However there was no 'previously joined' notification; the device remained non-functional. Sniffer trace shows the old device ID (issued days ago on its first join to the C-7) is indeed actively communnicating with the hub (8B6D requesting data from C-7, C-7 issuing 'ACK' with frame pending, followed by data) yet C-7 event history shows nothing and device doesn't work.

Same 'remedy' as on C-8 is required to get the device working: rejoin it through a router, in which case 'Add Device' produces the expected 'previously found' and it works normally, or delete it from the database, reset/rejoin, and fix whatever automations used it.

3 Likes

[C-7] now causes some Zigbee devices to change the short Network IDs :

This is an excerpt from a log when re-pairing a previously joined device , paired directly to the C-7 hub

child:[Loratap, 488D, type:EMBER_SLEEPY_END_DEVICE]

...........
dev:38042023-03-11 16:50:38.617debugLoratap received device announcement, Device network ID: 488D
dev:38042023-03-11 16:50:38.614debugLoratap catchall descMap: [raw:catchall: 0000 0013 00 00 0040 00 488D 00 00 0000 00 00 158D484275C6EAFE38C1A480, profileId:0000, clusterId:0013, clusterInt:19, sourceEndpoint:00, destinationEndpoint:00, options:0040, messageType:00, dni:488D, isClusterSpecific:false, isManufacturerSpecific:false, manufacturerId:0000, command:00, direction:00, data:[15, 8D, 48, 42, 75, C6, EA, FE, 38, C1, A4, 80]]

^^^^^15 seconds later ^^^^

dev:38042023-03-11 16:50:23.538debugLoratap received device announcement, Device network ID: 8103
dev:38042023-03-11 16:50:23.534debugLoratap catchall descMap: [raw:catchall: 0000 0013 00 00 0040 00 8103 00 00 0000 00 00 0E03814275C6EAFE38C1A480, profileId:0000, clusterId:0013, clusterInt:19, sourceEndpoint:00, destinationEndpoint:00, options:0040, messageType:00, dni:8103, isClusterSpecific:false, isManufacturerSpecific:false, manufacturerId:0000, command:00, direction:00, data:[0E, 03, 81, 42, 75, C6, EA, FE, 38, C1, A4, 80]]
.........

2 Likes

Are there other routers in your mesh? I see this happen (short ID change) when a router elsewhere in the mesh accepts a join request; when I've tried this with the C-7 (and no other router) I haven't (yet) see it issue a new one. It's easiest to observe these behaviors with the bare minimum set of devices; any other routers in range can change the results.

I experimented a bit with the C-3 on .105 but was unable to actually get a child device to direct join to the hub (it's not practical for me to disable all the routers since there are dozens and the C-3 is currently running the whole house). The router I did join to it did receive a different short ID on every reset/rejoin, but I can't verify that the C-3 accepted those join requests directly.

1 Like

This is on my dev hub. I haven't tried pairing it away from the hub, so that eventually the device chooses a repeater to join.

But, I just remembered that I am using a newer platform version... so my and your environments are different, as the Hubitat team is working on resolving this issue right now and the changes are probably in this area...

It may be a good idea if you join the Beta program (send a PM with your HUB ID to bobbyD), as the latest changes are implemented in the Beta versions which are not rolled out to te production version yet.

Are you seeing the stagnant child table behavior on C-7? I can't imagine how/why that is happening.

Actually I've tried joining the beta program before, never got accepted :man_shrugging:

1 Like

This is my ChildAndRouteInfo on the main hub (C-7):

Summary

Parent child parameters
EzspGetParentChildParametersResponse [networkId=0, childCount=8, parentEui64=0000000000000000, parentNodeId=FFFF]

Child Data
child:[Sensor KrisiRoom Window Contact, A856, type:EMBER_SLEEPY_END_DEVICE]
child:[null, 5D9D, type:EMBER_SLEEPY_END_DEVICE]
child:[Button Xiaomi Bedside Krisi Room, 2DBC, type:EMBER_SLEEPY_END_DEVICE]
child:[Contact Sensor Bathroom Door Aqara, D2E9, type:EMBER_SLEEPY_END_DEVICE]
child:[Sensor Bedroom Door, 3DE4, type:EMBER_SLEEPY_END_DEVICE]
child:[Sensor Motion Aqara Krisi Room, F18A, type:EMBER_SLEEPY_END_DEVICE]
child:[Button My Room, 8831, type:EMBER_SLEEPY_END_DEVICE]
No information for Child 7

Neighbor Table Entry
[TS011F (Krisi Room), 07CA], LQI:255, age:4, inCost:1, outCost:1
[Power Plug Ikea TRADFRI , 0F69], LQI:250, age:4, inCost:3, outCost:1
[Power Plug OSRAM 01, 0FB8], LQI:254, age:4, inCost:1, outCost:7
[TS011F Lelki (AC LivingRoom), 288D], LQI:253, age:4, inCost:3, outCost:1
[TS011F Wall Socket (LELLKI) My Room, 379E], LQI:255, age:4, inCost:1, outCost:1
[TS0121 MyRoom, 4C5B], LQI:251, age:4, inCost:3, outCost:3
[Radar Tuya _TZE200_ztc6ggyl LivingRoom, 84B7], LQI:254, age:4, inCost:1, outCost:1
[IKEA TRADFRI Signal Repeater, 8DB4], LQI:254, age:4, inCost:1, outCost:1
[Toster, B4B8], LQI:250, age:4, inCost:3, outCost:3
[TS0207 Tuya Zigbee Repeater, C32E], LQI:254, age:4, inCost:1, outCost:1
[Switch Sonoff LivingRoom , F188], LQI:249, age:4, inCost:3, outCost:1
[Bulb Philips Hue LCT007, FF12], LQI:241, age:4, inCost:5, outCost:1

Route Table Entry
status:Unused
status:Unused
status:Active, age:64, routeRecordState:0, concentratorType:None, [Sensor Bedroom Right Window Contact , 05E2] via [TS0121 MyRoom, 4C5B]
status:Active, age:64, routeRecordState:0, concentratorType:None, [Sensor LivingRoom Left Window Contact , 81F3] via [Radar Tuya _TZE200_ztc6ggyl LivingRoom, 84B7]
status:Unused
status:Active, age:64, routeRecordState:0, concentratorType:None, [Sensor Motion My Room (lumi), 4572] via [TS011F Wall Socket (LELLKI) My Room, 379E]
status:Active, age:64, routeRecordState:0, concentratorType:None, [Switch Sonoff LivingRoom , F188] via [Radar Tuya _TZE200_ztc6ggyl LivingRoom, 84B7]
status:Active, age:64, routeRecordState:0, concentratorType:None, [Sensor Vibration Aqara 1, 3091] via [TS0207 Tuya Zigbee Repeater, C32E]
status:Active, age:64, routeRecordState:0, concentratorType:None, [TS0207 Tuya Zigbee Repeater, C32E] via [Radar Tuya _TZE200_ztc6ggyl LivingRoom, 84B7]
status:Active, age:64, routeRecordState:0, concentratorType:None, [Button Philips Hue Dimmer 2 Livingroom, D800] via [TS0207 Tuya Zigbee Repeater, C32E]
status:Active, age:64, routeRecordState:0, concentratorType:None, [Toster, B4B8] via [TS011F (Krisi Room), 07CA]
status:Active, age:64, routeRecordState:0, concentratorType:None, [Button TS004F Bedside Living Room, EDE2] via [TS0207 Tuya Zigbee Repeater, C32E]
status:Active, age:64, routeRecordState:0, concentratorType:None, [Button TS0044, 17D9] via [TS0121 MyRoom, 4C5B]
status:Active, age:64, routeRecordState:0, concentratorType:None, [Button Philips Hue Dimmer Livingroom, 8397] via [Radar Tuya _TZE200_ztc6ggyl LivingRoom, 84B7]
status:Active, age:64, routeRecordState:0, concentratorType:None, [Button Sonoff SNZB-01, 8CA3] via [TS0207 Tuya Zigbee Repeater, C32E]
status:Active, age:64, routeRecordState:0, concentratorType:None, [Power Plug OSRAM 01, 0FB8] via [TS011F Wall Socket (LELLKI) My Room, 379E]

Please be sure to submit your request by joining the Beta group. It doesn't look like you have requested access or if you did, your hub id may have been missing.

https://community.hubitat.com/g/Beta

3 Likes

Ok, you'd be able to easily tell if one of your devices exited the child table (or should have, and didn't). The seemingly non-volatile behavior of the child table (I only see this on the C-7 now) would be obvious after a shutdown/reboot. I'd guess that this is one of the things that has been changed in the firmware you're using.

Ok; I had done this a couple of times before (once when I just owned a C-3, and again when I got a C-7 years ago). Will give it another go.

3 Likes

Please do, Tony, we need your sharp eyes on stuff that is happening in the beta releases!! :slight_smile:

5 Likes

Just did it for the C-7.... wow that process has been simplified. Before I recall having to fill out a questionnaire. If I had known it was that easy now I would have been more persistent.

3 Likes

Great to have you in the beta, thanks for diving into this stuff.

FWIW, you can also get your C-8 added and decide when you want to put beta releases on it as well. Bobby's happy to add multiple hubs from us into the beta program. [As he shamelessly nudges Tony towards the "send" button... :wink: )

1 Like

Thanks! I tried the same link, entering the UID of my C-8; it says 'already requested membership'. Must be another link separate for C-8 beta?

1 Like

I think there's only one link. @bobbyD can sort it out...maybe just PM him your C-8 UID.

1 Like

No, is not, I saw it coming through. Please give me 15 min, will get to it shortly.

4 Likes

No problem; also just submitted C-3.

2 Likes

Superstah!

2 Likes

Did you mean C-7? I enrolled your C-8, but don't see a C-3, just a C-7 beside your C-8. Send me a private message with the C-3's id.

1 Like

Yes, C-3; just did that.

I freaking knew it !!! Thank you discovering this. I had the same issue when I did a dual channel swap between C4<->C7 and for months after I had major issues pairing devices without shutting down the "other" hub. Below from March 2022

3 Likes