Skip to content

Not getting all BasicInformation cluster Leave events? #1159

@olavt

Description

@olavt

Test procedure:

  1. Add a Matter over Thread device to Apple Home (using Apple TV as Thread Border Router)
  2. Add the same device to Open Home Foundation Matter Server (using pairing code from Apple Home)
  3. Add the same device to chip-tool (using pairing code from Apple Home)
  4. Open a subscription to Leave events from chip-tool using the command "basicinformation subscribe-event leave 5 10 1 0"
  5. Select "Remove Accessory" from Apple Home and then "Remove from all services"

I then observe the following:

From the chip-tool, I get all the Leave events like this:

[1748617981.204] [21465:21467] [TOO] Endpoint: 0 Cluster: 0x0000_0028 Event 0x0000_0002
[1748617981.204] [21465:21467] [TOO] Event number: 196630
[1748617981.204] [21465:21467] [TOO] Priority: Info
[1748617981.204] [21465:21467] [TOO] Timestamp: 619313
[1748617981.204] [21465:21467] [TOO] Leave: {
[1748617981.204] [21465:21467] [TOO] FabricIndex: 18
[1748617981.204] [21465:21467] [TOO] }
[1748617981.204] [21465:21467] [TOO] Endpoint: 0 Cluster: 0x0000_0028 Event 0x0000_0002
[1748617981.204] [21465:21467] [TOO] Event number: 196631
[1748617981.204] [21465:21467] [TOO] Priority: Info
[1748617981.204] [21465:21467] [TOO] Timestamp: 620040
[1748617981.204] [21465:21467] [TOO] Leave: {
[1748617981.205] [21465:21467] [TOO] FabricIndex: 19
[1748617981.205] [21465:21467] [TOO] }
[1748617981.205] [21465:21467] [TOO] Endpoint: 0 Cluster: 0x0000_0028 Event 0x0000_0002
[1748617981.205] [21465:21467] [TOO] Event number: 196632
[1748617981.205] [21465:21467] [TOO] Priority: Info
[1748617981.205] [21465:21467] [TOO] Timestamp: 620340
[1748617981.205] [21465:21467] [TOO] Leave: {
[1748617981.205] [21465:21467] [TOO] FabricIndex: 20
[1748617981.205] [21465:21467] [TOO] }

In my client connected to "Open Home Foundation Matter Server", I get only this message:

[30.05.2025 15:13:00.056 INF] Received response:
{
"event": "node_event",
"data": {
"node_id": 1,
"endpoint_id": 0,
"cluster_id": 40,
"event_id": 2,
"event_number": 196630,
"priority": 1,
"timestamp": 619313,
"timestamp_type": 0,
"data": {
"fabricIndex": 18
}
}
}

I ignore this message in my application, since the fabric index does not match the one I control (which was fabric index 19 in this test). Fabric index 18 is related to Apple Home and fabric index 20 is the one owned by the chip-tool.

I would have expected to have received the remaining two Leave events, but they are not arriving.

The problem is that I don't get to remove the node from my fabric and end up with a dead node.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions