Anybody else having issues with this? My environtment canada sensors have been down for 20hrs now.
Logs are showing
Logger: homeassistant.components.environment_canada
Source: components/environment_canada/__init__.py:66
and
Logger: homeassistant.components.environment_canada
Source: components/environment_canada/__init__.py:46
Things seems to be running fine on my side. What version are you using, and have you tried restarting?
May have just been an outage since it’s back up and running for me today.
@Bondo2 There’s a frost advisory here tonight so I was able build this out. Let me know if it works for you, just double-check the entity name.
# Frost Advisory
template:
- binary_sensor:
- name: "Frost Advisory"
device_class: cold
icon: mdi:snowflake-alert
state: >
{% if states('sensor.ottawa_kanata_orleans_advisory')| int > 0 %}
{{ 'Frost' in state_attr('sensor.ottawa_kanata_orleans_advisory', 'alert_1') }}
{% else %}
{{ false }}
{% endif %}
Works for me!
I’d like to know who to call to turn it off!
I rebuilt my HA and my forecast TTS script broke. Turns out that
sensor.regina_forecast
changed to
sensor.regina_summary
In case anyone else is wondering why their forecast isnt coming down anymore.
That’s odd, because mine is still sensor.vancouver_forecast and works just fine.
I wonder why there is a difference?
I checked my old installation and it still uses …forecast and it works. My new installation wasnt giving me any forecast and the sensor which contains the forecast info is called …summary
I doubled checked and it is true the new installation has no sensors containing the word, “forecast”
Interesting.
I’m not sure why my implementation is different, but as long as it works, I guess we are both good.
If it breaks one day, I will know where to look - thanks!
How is sensor.ottawa_kanata_orleans_advisory configured. I have sensor.advisories which is always 0 (sometimes unavailable) and doesn’t have a attribute called alert_1
The state of each alert sensor is the count of current alerts of that type, and the titles of the current alerts are in numbered attributes. So, when there’s a frost advisory the state will be 1 and you will see an attribute called alert_1 with a value of Frost Advisory.
Hi,
I have many sensors that are always or almost always unavailable. I was wondering if there was a reason for this?
I saw in the doc that the humidex and the wind chill are on only on certain temperature, but the others?
Here is the list of the sensors:
chance_of_precipitation
current_condition
humidex
icon_code
visibility
wind_chill
wind_gust
Thank you
Also how come that in the daily forecast for today, the precipitation_probability = 0,
in the hourly forecast for today, each hour has a precipitation_probability = 90,
and the sensor chance_of_precipitation is unknown?
That’s certainly odd. If you can tell me what station you’re using I’ll test it out on my side.
Sure it’s this one:
s0000361,Deux-Montagnes,QC,45.53N,73.88W
Awesome, thanks.
Environment Canada does not report all current conditions for that station, which explains why many of the sensors are unavailable: Deux-Montagnes, QC - 7 Day Forecast - Environment Canada
The forecast precipitation probabilities are also consistent with their data, which looks to me like an error on their side: https://dd.weather.gc.ca/citypage_weather/xml/QC/s0000361_e.xml
Ok. Thanks a lot for taking your time.
Don’t suppose you would mind throwing an example of your script up and how you trigger it, etc. I have used this integration for years and its been a dream of mine to have it announce weather watches, advisory’s and warnings during waking hours.
I was trying to automate off a state change but every time the sensor would update it looked like a state change so google would announce, but there would be nothing. This got old with the wife pretty quick. It appears the text for the warnings would be an attribute too but now I see nothing when I check sensor.mycity_warnings ?
Was this ever resolved? I would like to be able to use the radar images again as it was an integral part of our climate dashboard.
Yes. It was fixed by adding caching to the library, reducing queries by 95%.
