Skip to content

test(trace): ingest the F461 and MyHomeServer1 captures of 2026-10-25 (actuator status, fan speed and humidity per zone) #815

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
GreenGrassBlueOcean merged 2 commits into from
Oct 25, 2026
Merged

test(trace): ingest the F461 and MyHomeServer1 captures of 2026-10-25 (actuator status, fan speed and humidity per zone) #815

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
b8b72b5
test(trace): ingest the F461 and MyHomeServer1 captures of 2026-10-25…
Oct 25, 2026
50c740e
test(trace): keep the 2026-10-25 replays green across #792; drop the …
Oct 25, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
115 changes: 115 additions & 0 deletions tests/fixtures/traces/f461_actuators_live_2026-10-25/README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,115 @@
# Live capture: BTicino F461 (firmware 2.0.11), 2026-10-25, actuator status at startup and valve switching on zone 2

Every frame the physical F461 gateway put on its event session on 2026-10-25 between 08:31:43 and 08:35:57 UTC
while the actuator-switching capture asked for in [review 5642758778](https://github.com/OpenWebNet-HA/MyHOME/pull/792#pullrequestreview-5642758778)
of MyHOME #792 was run: a local Home Assistant running the integration at the head of #792 (`0e44bcd7`, PyPI OWNd
pin) was restarted with an event-only listener on the bus; its startup sweep of the Living Now KW4691 standalone
thermostat at zone 2 was followed by the all-zones actuator request `*#4*#0*20##` that #792 adds, answered by both
actuators within 0.13 s; then `climate.set_temperature` 26.0 made the heating valve (actuator 1) report itself on
2.2 s after the setpoint echo, and `climate.set_temperature` 23.5 made it report itself off 2.1 s after the echo.
The fan coil (actuator 2) sent nothing but its startup reply. Nothing is synthetic except the gateway's MAC and
LAN address in the `gateway` block (no WHO 13 frame was exchanged); the HMAC handshakes are not included. Recorded
with OWNd master (`OWNEventSession`).

The same plant's earlier captures are `tests/fixtures/traces/f461_live_2026-10-15/`, `f461_live_2026-10-18/`,
`f461_live_2026-10-23/` (run 6 holds the previous `*#4*2#1*20*0##` / `*#4*2#2*20*5##` pair, 2.2 s after a zone
poll, and the plant table in full), `f461_live_2026-10-24/` and `f461_presets_live_2026-10-24/` (both hold a valve
off report 3 - 43 s after a Home Assistant write; none holds the valve switching on).

## Plant

| Item | Value |
|---|---|
| Gateway | F461, firmware 2.0.11 (`*#13**16*2*0*11##`), kernel 5.10.35, distribution 2.26.4, WHO 13 device type `200`, WHO 1013 object model `134.15.5.0`, HMAC SHA-256 on both sessions (identified in `f461_live_2026-10-23`) |
| Climate | Living Now KW4691 standalone thermostat at zone 2 with a BT-F430/2 heating valve (actuator 1) and a BT-F430/4 fan coil (actuator 2); heating season; no central unit; zone 2 is the only WHO 4 device; measured 24.0 throughout |
| Home Assistant | a local instance with the integration at #792's head (`0e44bcd7`) and the PyPI OWNd pin, started 12 s before the first row; `climate.climate_zone_2` = zone 2 (`heat` + `cool`, standalone); every request in the archive is its own, taken from its debug log (not included) |
| Cover | the BT-F411/2 timed shutter `05` answers Home Assistant's post-discovery poll `*#2*05##` (the first row); lights 06 / 07 / 09 are not involved |

## File format: `myhome_trace_F461_actuators_live_2026-10-25.json`

Standard `MyHOME` derived trace archive, in the layout of `f461_presets_live_2026-10-24`:

- `capture` and `gateway`: metadata with the RFC 5737 documentation IP `192.0.2.92` and the synthetic MAC
`00:03:50:00:10:25` (the gateway of `f461_live_2026-10-23` / `-24`, a new synthetic MAC for this capture).
`capture.listener` is the event-only listener (`evt.jsonl`, started 08:31:36.5 UTC for 3600 s) and how Home
Assistant's log was placed on its clock; `capture.runs` lists each run with its UTC window, `t_offset`, purpose
and `home_assistant_requests` (the frames Home Assistant sent in it). There is no probe command session in this
capture, so there are no `probe_segments`. `capture.home_assistant_entity_states` holds the climate entity's
service calls and state changes as recorded from Home Assistant's event bus for the first write (UTC wall clock,
`hvac_action` and the `actuators` attribute of #792); the harness's first restore call (08:33:59 UTC) was dropped
by Home Assistant (the websocket message reused an id, so no service ran and nothing reached the bus) and its
record is listed with that note; the restore that reached the bus went through the REST API at 08:35:55 UTC and
its state change (idle at 08:35:57.464 UTC) was read back, not recorded, so it is not listed.
- `frames`: the 20 event-session `rx` frames (`timestamp`, `direction: "rx"`, `raw`, `run`, `session: "evt"`),
replayed by `test_device_health_corpus.py` and parsed by `test_trace_oracle_validation.py` (tier 4).
- `rows`: the 51 session-log entries in capture order:

| `dir` | Meaning | Fields |
|---|---|---|
| `tx` | frame Home Assistant wrote on its command session | `frame`, `is_status_request` (how OWNd classifies it), `source: "home_assistant"`, `t_estimated: true` |
| `rx` | frame read; `session: "cmd"` for the replies and the closing `*#*1##` (from Home Assistant's log, `source` / `t_estimated` as above), `session: "evt"` for the listener | `frame`; event rows also carry `ownd_class` and `ownd_log`, what OWNd master made of the frame at capture time. Unlike the probe captures, the command-session replies sit at the time the log collected each one (the F461 delivers them one by one), between the request and its ACK |
| `result` | what the command session returned for the `tx` | `ack` (always true here), `reply_time_s`, `replies`, `events_after` (only the event-session frames that arrived while the request was pending, `tx.t <= t <= result.t`), `utc`, `note` on every request (what it established) |

Sessions and timing: the listener's `t` is the archive's axis, shifted so that `t = 0` is the first row (Home
Assistant's `*#2*05##`, whose reply is the listener's first frame). Home Assistant's debug log (local clock, UTC +
2 h, millisecond resolution) was placed on that axis through the 14 frames both logs hold (log seconds - listener
`t` = 36.544 - 36.547 s), so every `tx` / command-session `rx` / `result` row carries `t_estimated: true`: `tx.t` is
when the worker handed the frame to the session ("unqueued"), `result.t` the ACK ("successfully sent"), each
reply the moment the log collected it. The two setpoint writes opened a new command session first (the previous
one had been closed idle 15 s after the startup); their `reply_time_s` includes that HMAC handshake (24 - 27 ms).
`capture.runs[].t_offset` gives each run's first row. `step` is the request in flight.

## The runs

| `run` | UTC | What it was |
|---|---|---|
| `run1_startup_sweep_and_actuator_request` | 08:31:43-08:31:48 | Home Assistant's startup, 12 s after the climate entity was added: `*#2*05##` (cover 05 stopped), `*#4*2##` (`0`, `12` = 23.5, `*4*1*2##`, `14` = 23.5, `13` in 1.31 s), `*#4*2*60##` (34 %, 124 ms), `*#4*2*19##` (`*#4*2*19*5*0##`, 103 ms), then `*#4*#0*20##` 0.75 s later: `*#4*2#2*20*5##` at +88 ms, `*#4*2#1*20*0##` at +128 ms, ACK at +2.13 s; the command session was closed idle 15 s later |
| `run2_setpoint_26_0_valve_on` | 08:32:59-08:33:02 | `climate.set_temperature` 26.0 from 23.5 (measured 24.0): `*#4*2*#14*0260*3##` ACKed in 90 ms with no reply, echoed 26 ms after the ACK as `*#4*2*12*0260*3##`, `*4*1*2##`, `*#4*2*0*0240##`; 2.2 s after the echo `*#4*2#1*#20*1##`, then `*#4*2#1*20*1##` 37 ms later |
| `run3_restore_23_5_valve_off` | 08:35:55-08:35:57 | `climate.set_temperature` 23.5: `*#4*2*#14*0235*3##` ACKed in 104 ms with no reply, echoed 50 ms after the ACK as `*#4*2*12*0235*3##`, `*4*1*2##`, `*#4*2*0*0240##`; 2.1 s after the echo `*#4*2#1*#20*0##`, then `*#4*2#1*20*0##` 3 ms later; the listener saw nothing else until it was stopped at 08:37 UTC |

Home Assistant's requests before the listener started (`*#1*0##`, `*#2*0##`, the `*#4*0##` and `*#16*0*5##`
NACKs, the WHO 18 meter reads) are not in the archive.

## What the plant established (each row has its frames in the archive)

* **The all-zones actuator request is answered by both actuators within 0.2 s.** `*#4*#0*20##`, sent 0.75 s
after the startup sweep's last ACK (2.4 s after `*#4*2##` went out), was answered on both sessions by
`*#4*2#2*20*5##` (the fan coil, off) at +88 ms and `*#4*2#1*20*0##` (the valve, off) at +128 ms; the ACK came
2.13 s after the request, 2.0 s after the last reply (the gateway closes the all-zones batch on a timeout, as the
MyHomeServer1's 43-actuator reply of `myhomeserver1_live_2026-10-24` did not). With #792 the entity showed
`hvac_action: idle` and `actuators: {"1": "off", "2": "off"}` 12.1 s after it was added; on the previous build
the first dimension 20 frame after a start was the F461's periodic broadcast, 56 min later (not in the archive).
* **A setpoint write above the measured temperature switches the valve on 2.2 s after the echo.**
`*#4*2*#14*0260*3##` (ACK 90 ms, no reply) was echoed as `*#4*2*12*0260*3##`, `*4*1*2##`, `*#4*2*0*0240##` (no
dimension 14 frame, as every dimension 14 write of `f461_live_2026-10-24`); 2.203 s after the echo the valve
reported itself on, first in the command form `*#4*2#1*#20*1##` and 37 ms later in the status form
`*#4*2#1*20*1##`. The same 2.2 s separated the setpoint echo and the valve report on the old build the evening
before, so the delay is the thermostat's, not the integration's.
* **The restore switches it off 2.1 s after the echo.** `*#4*2*#14*0235*3##` (ACK 104 ms) was echoed the same
way with 23.5; 2.126 s later `*#4*2#1*#20*0##`, then `*#4*2#1*20*0##` 3 ms later.
* **The fan coil stays 5 / off.** Actuator 2's only frame is the startup reply `*#4*2#2*20*5##`; `*#4*2*19*5*0##`
(cooling fan coil off, heating valve off) says the fan coil is the cooling side. The valve's two switches are the
only dimension 20 frames outside the startup reply.
* **The `#20` command form.** Each spontaneous switch arrives twice: `*#4*2#1*#20*V##` (the dimension written as
`#20`, the form a write would use) 3 - 37 ms before `*#4*2#1*20*V##`; the startup reply has no command form.
The pinned OWNd parses the command form as `OWNHeatingCommand` with no zone, so the integration logs it as
unsupported and the entity acts on the status form only (no OWNd change decodes the `#20` write-form
echo yet; the tests accept an event decoding as well).
* **Entity replay (base branch, pinned OWNd).** Fed the event frames, the climate entity shows `hvac_action`
`None` until `*#4*2*19*5*0##`, then `idle` through both startup actuator replies, `heating` on `*#4*2#1*20*1##`
and `idle` on `*#4*2#1*20*0##`; the setpoint follows the dimension 12 echoes 23.5 -> 26.0 -> 23.5 in `heat`;
`fan_coil_cooling: true`, `fan_mode: auto`, `running_fan_speed: off`, humidity 34 %. Pinned in the tests; the
`actuators` attribute is asserted only when the entity exposes it (#792).
* **Timings.** `*#4*2##` 1.314 s, `*#4*2*60##` 124 ms, `*#4*2*19##` 103 ms, the two writes 90 / 104 ms including
the session handshake; every reply reached the listener 1 - 3 ms before Home Assistant's log collected it.

## What this capture backs

| Pull request / issue | Frames in this archive |
|---|---|
| MyHOME #792 (actuator status request after the startup sweep, `hvac_action` and `actuators` from dimension 20; review 5642758778 asked for this capture) | run 1: `*#4*#0*20##` -> `*#4*2#2*20*5##`, `*#4*2#1*20*0##` within 0.13 s; runs 2 / 3: `*#4*2#1*20*1##` 2.2 s after `*#4*2*12*0260*3##`, `*#4*2#1*20*0##` 2.1 s after `*#4*2*12*0235*3##`; `capture.home_assistant_entity_states` |
| MyHOME #791 (the setpoint in force is dimension 12; merged) | the two write echoes carry dimension 12 and the season frame, no dimension 14 frame |
| OpenWebNet-HA/OWNd#140 (`actuators_status` builder); the `#20` write-form echoes have no OWNd decoder yet | `*#4*#0*20##` sent by the integration (`is_status_request: false`); `*#4*2#1*#20*1##` / `*#4*2#1*#20*0##` parsed as commands by the pinned OWNd |
| own-firmware-oracle#60 (`*#4*#0*20##` forwarding matrix) | the physical F461's reply order (fan coil first) and the 2.13 s ACK |

Tests: `tests/test_f461_actuators_live_2026_10_25_trace.py`.
Loading
Loading