#I have a ton of these messages in my
1 messages · Page 1 of 1 (latest)
and then these messages are the failures I guess: matter-server | 2024-01-05 03:42:55 sdk4 chip.DIS[1] ERROR Timeout waiting for mDNS resolution. matter-server | 2024-01-05 03:43:09 sdk4 chip.DIS[1] ERROR OperationalSessionSetup[1:000000000000000C]: operational discovery failed: src/lib/address_resolve/AddressResolve_DefaultImpl.cpp:119: CHIP Error 0x00000032: Timeout
If you upgraded to the latest matter server, a simple restart of the server should clean up that orphaned node.
pretty sure i’ve done that, but i can try again. if im right about what this node is, its been around for a while and then somehow disappeared from home assistant
Now that you are working on the webinterface/websockets yourself, simple solution could be to just issue a node remove command for that node id to clean up this orphaned node.
Is this node not visible in HA with unavailable entities ?
I see it in HA with unavailable entities - HA says matter is not reporting the node - and a get_nodes command also does not report it
When you get into the beta stage, I can test this for you. I have such an orphaned node in my logs, that is not visible in HA.
matter-server | 2024-01-05 13:58:45 sdk4 asyncio[1] ERROR Task exception was never retrieved
matter-server | future: <Task finished name='Task-13059' coro=<MatterDeviceController._check_interview_and_subscription() done, defined at /usr/local/lib/python3.11/site-packages/matter_server/server/device_controller.py:870> exception=NodeNotExists('Node 12 does not exist.')>
matter-server | Traceback (most recent call last):
matter-server | File "/usr/local/lib/python3.11/site-packages/matter_server/server/device_controller.py", line 876, in _check_interview_and_subscription
matter-server | raise NodeNotExists(f"Node {node_id} does not exist.")
matter-server | matter_server.common.errors.NodeNotExists: Node 12 does not exist.
matter-server | 2024-01-05 13:58:59 sdk4 matter_server.server.client_handler[1] ERROR [547252040912] Error handling message: CommandMessage(message_id='d57e5635-df0c-4d34-83ad-50299df4d3aa', command='remove_node', args={'node_id': '12'})
matter-server | Traceback (most recent call last):
matter-server | File "/usr/local/lib/python3.11/site-packages/matter_server/server/client_handler.py", line 188, in _run_handler
matter-server | result = await result
matter-server | ^^^^^^^^^^^^
matter-server | File "/usr/local/lib/python3.11/site-packages/matter_server/server/device_controller.py", line 515, in remove_node
matter-server | raise NodeNotExists(
matter-server | matter_server.common.errors.NodeNotExists: Node 12 does not exist or has not been interviewed.```
so it's attempting to interview it repeatedly, but when i try to remove it, it's not found
what server version is this ?
because in the most recent version I added some logic to just forget such nodes at startup
this is all old stuff btw, you cnat get into this state anymore with more recent versions but with early versions a partial commission wasnt cleaned up
it's 5.1.4 - latest code
i don't know if partial commission is what happened here - it was working for quite a while
i updated the server from an older version to 5.1.3 this week
Ah you mean this node dropped off
then yesterday updated to 5.1.4 + some of my patches (which i think are separate from this)
yeah - it was in HA for months
it is, but it has this banner: This entity is no longer being provided by the matter integration. If the entity is no longer in use, delete it in settings.
Yes, and now go to settings and hit delete
so is matter server getting this node id from HA?
but matter server doesn't have it anymore?
No, it is not reachable anymore
yes it does have it because its trying to reach it over and over
The nodes listing filters out the unavailable nodes now
the one you called on the websocket
I should probably alter that
but then remove_node should have found it?
But its a bit chicken and the egg, we have no idea of the state of the node because we couldt reahc it from the start
yes
maybe the type was wrong or something - string vs int
it should auto coerce it to int
hmmm maybe it isnt coercing to int then
ah I spot one tiny issue here
so the node would be deleted but you get an error
its not what your are seeing but there is a small bug
in your case its most porbably int vs string then
ok - i fixed that so it's sending int now, but don't have another node i want to remove
adding it back now i get a new issue:
matter-server | 2024-01-05 14:34:01 sdk4 chip.CTL[1] ERROR Failed to perform commissioning step 18
matter-server | 2024-01-05 14:34:01 sdk4 chip.EM[1] ERROR Failed to send Solitary ack for MessageCounter:28937974 on exchange 19662i:src/messaging/ExchangeContext.cpp:103: CHIP Error 0x00000002: Connection aborted
matter-server | 2024-01-05 14:34:01 sdk4 matter_server.server.client_handler[1] ERROR [547681815760] Error handling message: CommandMessage(message_id='ac406574a6ac4eb89c34fe15ddad04b9', command='commission_with_code', args={'code': '<redacted>', 'network_only': True})
matter-server | Traceback (most recent call last):
matter-server | File "/usr/local/lib/python3.11/site-packages/matter_server/server/client_handler.py", line 188, in _run_handler
matter-server | result = await result
matter-server | ^^^^^^^^^^^^
matter-server | File "/usr/local/lib/python3.11/site-packages/matter_server/server/device_controller.py", line 197, in commission_with_code
matter-server | raise NodeCommissionFailed(
matter-server | matter_server.common.errors.NodeCommissionFailed: Commission with code failed for node 21```
Was that using Android perhaps ?
ios
Did you factory reset it ?
it's an eve motion
you need to start commissioning mode from apple home and then use that code in the wizard from HA (or use the websocket directly now you are there)
yeah - that's what i did. i can try the websocket later i guess
If i go to the "Matter accessories" page in apple settings, it shows it as connected to my HA matter server
shouldnt matter - the app just calls the commission_with_code under the hood in this case
yeah, its borked
maybe remove it from there first
Your log above shows ocnnection issues btw
i may have to get out wireshark or something