Weather Driver

[Vit]

Installed 3.3.7.0 and poured some water

Of the newly added child devices, only one got updated - RainToday (Rain Daily).

In the HS log, I am getting this error:

Sep-12 4:14:03 PM Error Posting back to plugin web page HubitatConfig: Object reference not set to an instance of an object.-> at System.Runtime.Remoting.Messaging.LogicalCallContext.Propaga teIncomingHeadersToCallContext(IMessage msg) at System.Runtime.Remoting.Proxies.RealProxy.PrivateInvoke(Mess ageData& msgData, Int32 type) at HomeSeerAPI.IPlugInAPI.PostBackProc(String page, String data, String user, Int32 userRights) at Scheduler.proxyPage.postBackProc(StateObject& state, String Data)

The error seems to be specific to the 3.3.7.0 version. I tried reverting to the 3.3.6.0 and the error disappears.

Please see the attached debug file.
hubitat_debug.zip

Thanks!

Forgot to mention, the rain features appeared in all sensors, not only the Rain Gauge. Is it intentional?

[Michael McSharry]

The error is odd as I can play back your debug file and the following is what I get for HS view. The debug does show that the new rain devices were not recognized on your system. This is likely due to the error that was raised. I will need to try to understand where the error you are seeing is being raised.

[Vit]

Let me know if there is anything I can do on my side.

[Michael McSharry]

I see three Ecowitt RF Sensor device being reported by Hubitat. All three report as having a WaterSensor capability and attributes of the various flavors of rain. The updates from Hubitat are based upon the one that I showed in my screenshot. The other two did not have any updates for rain.

Is your question based upon these three sensors all having rain devices created or are there other devices besides Ecowitt that had the rain devices?

[Michael McSharry]

The HS Log message "Error Posting back to plugin web page HubitatConfig" is something related to the Config page. When I bring up the Config page and make changes in the Debug checkbox or Connection Timeout I do not get any messages from HS. What changes were you making on the Config page when the error is raised?

I also noticed several restart of the plugin in your debug where the most recent around 4:15 PM. The last report of rain sensor update in the debug was at 2:57. Could it be in your case that you did not wait long enough after the rain devices were created to get an update from Hubitat?

[Vit]

All three report as having a WaterSensor capability and attributes of the various flavors of rain. The updates from Hubitat are based upon the one that I showed in my screenshot. The other two did not have any updates for rain.

Is your question based upon these three sensors all having rain devices created or are there other devices besides Ecowitt that had the rain devices?

Yes, I was asking about the three additional sensors that I have. It is all gray to me; and I don't know how it all gets sorted out on the Hubitat side but for these three sensors , the only item that is visible and (may be) related to Water(Sensor) is humidity. For instance, Soil Moisture sensor displays these attributes:

• battery : 57
• batteryIcon : 60
• batteryOrg : 1.5
• humidity : 15
• status : OK

Is it possible that the driver suppresses some of the available attributes based on the sensor model number?

The last report from the Rain Sensor at 2:57 is correct. That's when I climbed the roof.

~ 4PM events are also understandable. I noticed the error in the HS log, pressed 'Connect/Re-Scan Devices' button several times and tried going back to the released version.

I went back now to do some "controllable" button-pressing. And I believe I see two scenarios that result in the error.

Say, I press 'Connect/Re-Scan Devices' button on the plugin Config page. For the first 30 (or more) seconds, there is no response on the screen to acknowledge the button press. (the red 'Connecting to Hubitat' message appears after the said period [I am using Chrome]).
If (by association with the HS Z-Wave plugin) I get scared and restart HS, the natural but wrong inclination is to reuse already-opened plugin Config Page. The error is recorded every time I 'rescan devices'.

In the second scenario:

When the plugin (re)started, I see ~40 seconds between 'Finished initializing..' and 'Initializing is done..'. If I don't wait long enough and open the plugin Config page too soon, attempts to rescan devices generate this error.

So, as far as the error is concern, I think it is fair to attribute it to the user stupidity.

However, having rescanned the devices and letting everything settle, I took another walk to the roof and unfortunately, I still see only the RainToday field updated.

[Michael McSharry]

Your prior debug contained several messages about a device not in the device map. Do these still exist in the current debug? Look only at the top as the bottom will be older data. The times are in the file. The lines containing DEVICES are the events being reported by Hubitat/MakerAPI.

The other piece of information of interest is \config\hubitat-devicemap.json. It can be posted or sent to mcsSolutions at CenturyTel dot net.

I doubt if MakerAPI has insight into the available sensors connected to the Ecowitt. The plugin creates devices based upon the capabilities reported by MakerAPI for a Hubitat device. The same capabilities are reported for all three of your sensors. In the past HS has the ability to hide devices so they do not clutter the display.

[Vit]

A clean slate:
Returned VM to the pre-plugin state, installed the plugin (released version). Immediately replaced the executable with v. 3.3.7.0.
Enabled the plugin and scanned devices.

Original 'Hubitat Device not in Device Map" were mostly related to my 'MQTT bridge' device and 'Lifeline' virtual switch. Both are un-checked in the MakerAPI selection box. However, the box is titled "Allow Endpoint to Control These Devices"; and so, I guess it may be that the messages from the devices are being sent regardless of the selections.
The only other Not-In-Map device absolutePrs. This may be related to the fact that MakerAPI exposes Ecowitt WiFi Gateway in addition to the actual sensors and I have all of them selected.

At 2:07:38 PM, it 'rained' again.
The only message that went through was 'rainDaily'; all others were 'not-in-map'.

How she decides that

9/13/2020 2:07:38 PM 40855519 | Message received: {"source":"DEVICE","name":"rainDaily","displayName":"Rain Gauge Sensor","value":"0.51","unit":"mm","deviceId":226,"hubId":nu ll,"locationId":null,"installedAppId":null,"descriptionText " :null}
is a 'good' message and

9/13/2020 2:07:38 PM 40855486 | Message received: {"source":"DEVICE","name":"rainHourly","displayName":"Rai n Gauge Sensor","value":"0.51","unit":"mm","deviceId":226,"hubId":nu ll,"locationId":null,"installedAppId":null,"descriptionText " :null}
9/13/2020 2:07:38 PM 40855487 | Hubitat Device not in Device Map: rainHourly
is not so good, is certainly beyond me, especially, given the fact that she's apparently smart enough to associate "rainDaily" with "RainToday".

[Michael McSharry]

The device map looks good to me so I do not understand why the various rain devices are not in the map. What I did was add more info around these messages. It is in the attached. Also is the simulate file that was obtained from your initial debug. You can use this file to produce the event messages rather than dealing with the real sensor. Put hubitat_simulate.json in \Data\hubitat and then click on the button on the Config Settings page to exercise the simulated data. The new debug file is what is of interest The missing device map items will look like below.

9/13/2020 2:49:44 PM 18373 | Hubitat Device not in Device Map of 148 keys for key: 4_message with name message

[Vit]

Got them.
I also see 'Hubitat Device not in Device Map of 150 keys for key: 3_Switch with name switch' - that's probably about my virtual switch.

[Michael McSharry]

It looks to me as if the rain events are no longer producing the Device Map issue. I would expect your HS Device to have been updated. Is that the case or do we have another thread to investigate?

[Vit]

Everything seems to be working now. Thank you!!