My device seems to go offline every few weeks, always when I'm travelling. I'm setup to where my Hubitat can cycle my internet if it goes out but I can't seem to find a way to get a value on my Tempest on if it's connected or not. Anything anyone can think of?
To be clear my device stays connected to my network but it loses it's connection to the internet somehow. A quick power cycle resolves it.
@user2927 I can't think of anything unfortunately. I would reach out to Weatherflow customer support. I did see a couple instances on their forum of people having wifi issues. It's possible you have a defective hub.
There is a ping app (Web Pinger) that you could use that periodically can be set to send a ping to an IP address. After configurable number of failures it can be used to trigger an action. You would need to set up your weatherflow hub to a static IP address. Then periodically ping that with the app. If it fails you could then power cycle it.
Thanks for the driver! I have a voice announce every hour that includes "chance of rain" if >20%. However, seems the only attribute I know to use is for the daily forecast. Any way to get the "precip_probability" for the hour? I guess that would be something like "Current Hour Chance Precipitation".
From my following this driver and the related REST one, my understanding is that this information you desire is available from the API. But it dumps a ton of info and makes parsing heavy. So especially for this driver which is designed to be lite, it’s been deemed to not fit well with it. If anyone has interest in developing a “heavy” version of the driver, you could get it! I may have an interest in it and develop it come spring, so that I can proactively park my robot mower if rain chances creep above 80-90%.
EDIT: Well the hourly forecast might not actually be available via websocket, so definitely wouldn’t come to this driver if that’s the case since it used the websocket API. I know it’s available from the REST API though.
Apparently I'm slow....how do I set this up? Checked for basic newb documentation and didn't find any. I installed via HPM but don't see an app or a device. If I go to 'uninstall' in HPM it's there. Tried to add device and search for 'WeatherFlow' and 'Tempest', no results. What am I missing?
Thank you for all your work putting this together!
Go to Developer Tools section Drivers Code and ensure you have the WeatherFlow Lite driver in that list. If yes, go to Devices and click the Add Device button and choose Virtual. Give it a name and set type to WeatherFlow Lite. You can quickly search for it by typing in “weather”
@augoisms This was brought up elsewhere (not sure if for WF) but it happens here. For the strkedetected attribute is it showing maybe an epoch time? Anyway to have it show x-mins or x-hours ago instead?
@rlithgow1 Unfortunately showing time since would not be practical because you would have to constantly update the value every min/hour. That’s something better left to custom javascript in a dashboard.
I realize that the timestamp is not user friendly, however the formatted date time is available in the event description,
What about a preference to allow users to format that date plus the state variable dates to the hub local toon format? Default leave as is but making true would format those dates before event is fired.
@ritchierich Yes, that could be an acceptable solution.
There were previous issues with too many strike events coming through (during a storm?), so I probably need to add some rate limiting in there as well. No guarantees on timing, but I will try and work on both in the near future.
@rlithgow1 It appears that you're not getting any observations through the websocket connection. There is a health check that is run every 5 minutes and if there hasn't been an observation within 3 minutes it resets the connection. For me, the observations come in about once per minute.
If you enable debug logs and PM them to me, I can take look to see if anything stands out.
@rlithgow1 The logs appears to show that things are working... I would need logs from around when the not healthy warnings occur. Send me a private message with the logs if the warnings pop up again.