API Changes in Rheem Econet Vacation Mode

[arcivanov]

Describe the bug

Error setting up entry <redacted>@gmail.com for econet

Traceback (most recent call last):
  File "/usr/src/homeassistant/homeassistant/config_entries.py", line 762, in __async_setup_with_context
    result = await component.async_setup_entry(hass, self)
             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/src/homeassistant/homeassistant/components/econet/__init__.py", line 60, in async_setup_entry
    equipment = await api.get_equipment_by_type(
                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
        [EquipmentType.WATER_HEATER, EquipmentType.THERMOSTAT]
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    )
    ^
  File "/usr/local/lib/python3.13/site-packages/pyeconet/api.py", line 207, in get_equipment_by_type
    await self._get_equipment()
  File "/usr/local/lib/python3.13/site-packages/pyeconet/api.py", line 174, in _get_equipment
    _equip, __ = self.check_mode_enum(_equip)
                 ~~~~~~~~~~~~~~~~~~~~^^^^^^^^
  File "/usr/local/lib/python3.13/site-packages/pyeconet/api.py", line 98, in check_mode_enum
    if (value != enumtext.index(status)):
                 ~~~~~~~~~~~~~~^^^^^^^^
ValueError: 'Vacation' is not in list

Error doing job: Task exception was never retrieved (task: None)

Traceback (most recent call last):
  File "/usr/src/homeassistant/homeassistant/components/econet/__init__.py", line 89, in resubscribe
    await api.refresh_equipment()
  File "/usr/local/lib/python3.13/site-packages/pyeconet/api.py", line 201, in refresh_equipment
    _equip, __ = self.check_mode_enum(_equip)
                 ~~~~~~~~~~~~~~~~~~~~^^^^^^^^
  File "/usr/local/lib/python3.13/site-packages/pyeconet/api.py", line 98, in check_mode_enum
    if (value != enumtext.index(status)):
                 ~~~~~~~~~~~~~~^^^^^^^^
ValueError: 'Vacation' is not in list

Source: /usr/local/lib/python3.13/site-packages/pyeconet/api.py:324
First occurred: 12:11:30 PM (9 occurrences)
Last logged: 6:35:08 PM

    Failed to parse the following MQTT message: b'{"@ALERTCOUNT":0,"@ACTIVE":true,"@STATUS":"","@MODE":{"value":5,"status":"Vacation"},"device_name":"<redacted>","serial_number":"re-da-ct-ed","transactionId":"AUTOPOLL_2026-01-26T20:48:09.846Z"}'
    Failed to parse the following MQTT message: b'{"@ALERTCOUNT":0,"@ACTIVE":true,"@STATUS":"","@MODE":{"value":5,"status":"Vacation"},"device_name":"<redacted>","serial_number":"re-da-ct-ed","transactionId":"AUTOPOLL_2026-01-26T22:53:50.567Z"}'
    Failed to parse the following MQTT message: b'{"@ALERTCOUNT":0,"@ACTIVE":true,"@STATUS":"","@MODE":{"value":5,"status":"Vacation"},"device_name":"<redacted>","serial_number":"re-da-ct-ed","transactionId":"AUTOPOLL_2026-01-26T23:25:25.425Z"}'
    Failed to parse the following MQTT message: b'{"@ALERTCOUNT":0,"@ACTIVE":true,"@STATUS":"","@MODE":{"value":5,"status":"Vacation"},"device_name":"<redacted>","serial_number":"re-da-ct-ed","transactionId":"AUTOPOLL_2026-01-26T23:28:23.999Z"}'
    Failed to parse the following MQTT message: b'{"@ALERTCOUNT":0,"@ACTIVE":true,"@STATUS":"","@MODE":{"value":5,"status":"Vacation"},"device_name":"<redacted>","serial_number":"re-da-ct-ed","transactionId":"AUTOPOLL_2026-01-26T23:35:05.175Z"}'

To Reproduce
Steps to reproduce the behavior:

1. This occurs in the current latest HASS.
2. Set Rheem Hybrid Tank Heater to Away using Rheem app.
3. Disable and then enable Rheem integration.
4. Observe the exceptions.

Expected behavior
Parsing is correct

Screenshots
N/A

Additional context
This is sudden as of 2026-01-26

[arcivanov]

@w1ll1am23 So this stacktrace (api.py:98) has been overwritten in 3605d9b 5 months ago. HASS carries an old version of the lib that may not be affected by this problem, hence my request for home-assistant/core#161647 to bump the pyeconet dependency to latest stable

[w1ll1am23]

This was merged two weeks ago so I guess it should be out next month.

home-assistant/core#164859

[arcivanov]

If there are further problems I'll still be happy to maintain the pyeconet insofar I actually have hardware to test it (unless I get API dumps from people with different hardware).

fixed in home-assistant/core#164859

[arcivanov]

The most recent version of HASS with latest pyeconet isn't functioning due to changes in protocol related to Vacation/Away mode. While the latest pyeconet version prevents a crash changing from and two away mode has no actual effect on the state reflected in the Rheem app.

[mbedworth]

Adding a data point that may help narrow this down. Running HA 2026.7.2 (pyeconet 0.2.2, the version currently pinned in the econet manifest) against a Rheem heat pump water heater. Tonight, the very first time the device's own app was used to enter Vacation Mode, Equipment.update_equipment_info() threw and logged:

ERROR (paho-mqtt-client-...) [pyeconet.equipment] Failed to update with message: {'@AWAY_MSG': {'value': 1, 'constraints': {...}, 'title': 'Unit is in Away Mode', 'status': ''}, '@AWAY': True, '@ACTIVE': True, '@STATUS': '', '@MODE': {'value': 5, 'status': 'Vacation'}, 'device_name': '...', 'serial_number': '...', 'transactionId': '...'}

Looking at update_equipment_info (src/pyeconet/equipment/init.py, current main branch same shape as 0.2.2), the dict-value branch does:

if isinstance(value, Dict):
    for _key, _value in value.items():
        self._equipment_info[key][_key] = _value

This assumes self._equipment_info[key] already exists as a dict. @AWAY_MSG doesn't appear to be part of this water heater's normal (non-away) equipment snapshot, so the first time a dict-valued @AWAY_MSG update arrives, self._equipment_info.get('@AWAY_MSG') is None and the subscript assignment raises, which the broad except Exception swallows and logs as the "Failed to update" line above — silently dropping that field's update.

Separately (may or may not be related): @AWAYCONFIG — the field Equipment.supports_away gates on — never once appeared in any MQTT payload from this device all evening, successful or failed, even while genuine Vacation Mode was actively toggled on and off from the Rheem app multiple times and correctly reflected in @AWAY/@MODE. So supports_away stays False and HA's AWAY_MODE feature/turn_away_mode_on/off are unreachable for this unit regardless of the crash above. Not sure if this model/tier of water heater just doesn't send @AWAYCONFIG at all, or if it's a separate gap. Happy to grab more debug logs or test a patch against this hardware if useful — the unit is my daily driver so I can validate on real traffic.