CMX interactions with a WLL station

[billy]

Trying to get my head around how CMX interacts with WLL and have some questions (I've read some, but not all of the forum posts that might be relevant).

My setup is a WLL station that gets its data from a VP2+. (And I have an airlink as well - back to that at the end of this post).

So far, I understand the WLL uploads the VP2+ data to WL (I have a Pro subscription). It also has storage "on board" so that in the event of a power failure or loss of internet connectivity it can continue to collect the VP2+ data and on their resumption upload the data collected in the meantime. The WLL also "broadcasts" the data locally.

Mark's announcement for b3221 made me think a bit more about how CMX interacts with the WLL and WL. In the "changed" section of that post (can't work out how to quote the relevant bit!) I infer that
1. CMX, as a matter of routine, gets its data from the "broadcast"?
2. If that fails it gets it from the "local API"?

I would have guessed these were the same thing , but clearly that's not right! Could someone enlighten me on what these terms actually mean please? (It would be nice if there was a "Ways in which cmx gets data from a wll station" for dummies. I'm very happy to be the test dummy )

Also, I ended up rebooting my rpi after installing b3221 and I noticed that after CMX resumed it missed 10-15 minutes of airlink data. WL didn't miss any of it ... because the airlink data goes direct to WL. So it appears to me that cmx didn't recover the data from WL? Yet it recovered the VP2+ data. CMX runs as a service and I don't have a delay for CMX to start after a rpi reboot. Maybe I need that?
Edit: Just realize this is probably because the VP2+ data comes from the WLL whereas the airlink data has to come from WL

[mcrossley]

Normally CMX both listens for the WLL broadcasts (which only contain wind and rain data), *and* polls the WLL via the API (for temperature, solar, UV, extra sensor data).

The WLL sends the wind/rain broadcasts every 2.5 seconds.

For missed data, both the WLL and AirLink should be downloaded from wl.com (mine both catch-up). If you are missing the Airlink data during catch-up then you probably have something awry in the config.

The AirLink is a bit odd in that it is actually a standalone station type as far as Davis are concerned. However, you can "associate" it with an existing station when you install it by adding it as Node. Depending on whether it is fully standalone or added as a node you must configure CMX slightly differently. I'd recommend you add it as a node, that is the simplest option.

[billy]

Thanks Mark for your very helpful comments.

I thought I had set up the AirLink as a node but at the time I did this (late last year) it wasn't clear to me how I would know if I succeeded ... and my usual bumbling process to get it going almost certainly didn't help What is/are the tell tale sign(s) that show it is a node of the VP2 vs stand-alone? The attached screen shot shows how it appears in WL - is this indicating I didn't succeed in making it a node? ("Transmitter 2" is today's development of placing the VP2+ solar sensors away from the ISS)

Screenshot 2023-01-12 160316.jpg

BTW, I have had some concerns about my weather vane and in checking the CMX data using http://10.1.1.6:8998/api/tags/process.j ... x&wdirdata at about 10 minute intervals, I see the index is incremented at exactly 2 second intervals, so I guess there are some duplicate values in the data - ie about 20%?
Also, #nextwindindex appears to be the latest value, not the one about to be replaced as suggested in the Wiki.
When it gets a little cooler later today ... and I won't feel like a cat on a hot tin roof ... the VP2+ wind sensor will relocate to the new transmitter.

[mcrossley]

That looks like it has been added as a node.

The wdirdata is updated every time a wind packet is received, so every 2.5 seconds.
I'd have to check the code for the nextwindindex question, but will be a zero relative index.

[billy]

Thanks - I was pretty confident I had set it as a node because forum posts (your advice I think) suggested that was the sensible way to go. Not sure now why it missed some AirLink readings during catchup when the rest of the data was OK.

Should the AirLink logs be available to edit in the admin interface under Data logs/Extra data logs? Mine aren't.

[mcrossley]

Do you have the MXdiags file from when the catch-up did not happen?

I checked nextwindindex and it points to the next location to be used, there is a +1 on it after it used to write the last value.

[billy]

I've got the trifecta.
First, AirLink catchup did work ... I was looking in the wrong file
Second, nextwindindex is correct ... I had thought about the positions starting at zero but when I check for that I counted incorrectly
Third, I was right to have concerns about my weather vane measures ... it is getting "stuck", intermittently, around SW . SW is our usual afternoon direction ... when I often check it ... but the night time/morning direction is typically E. The vane itself points in the correct direction, but it's the value transmitted that is often incorrect - in blocks of time that are getting longer and longer and have become the majority. I'm guessing this means replacing the vane and anemometer.