IoT service outage fails to recover

Development Tools Cloud IoT
1

I have a similar problem to a post earlier today. Since start of Cloud problems around midnight my sensor variables have yet to all recover data flow. Most did but one is still locked at the value it took around midnight, ~19 hrs ago. Rebooting ESP8266 and sensors or re-logging the Cloud does not improve problem. The sensor in question functions fine when read locally on an UNO.
How do I re-establish Cloud data service on my nonfunctional variable?

2

Hello @dbfenner – are you still experiencing the issue?

3

Yes. No change in the issue. That one variable is still frozen while the others function fine.

4

Hello @dbfenner can you please tell us which is the name of the variable that is not working properly? Can you please paste here any relevant part of the sketch that edits the value of the variable and the output of the serial monitor?

5

It displays as "Pressure" and enters in the code as float pressure. There is limited math/editing done on the sensor reading (SparkFun BMP581). Here is that code section with comments:

    pressPa = data.pressure; // Read BMP581 sensor
    
    // Serial.print(pressPa); // Send to Serial Monitor, if desired
    // Serial.print("\t");
    
    if ( abs (pressPa/100 - pressure) <= 2) {  //Tests new data to see if it jumped unphysically
          pressure = pressPa / 100; // Calc pressure in millibar and load into Cloud variable
    }  // If new data was high or low then do not update pressure variable
6

@dbfenner I see data from the pressure variable correctly coming into our backend. Are you still experiencing the issue?

7

Yes, that pressure variable is reporting to me correctly again. Started up about 2 AM local time. Thanks for your help.
Is there something I can do from my end to repair this if I see it happen again?
David

8

Unfortunately, I can't say anything specific, since when I was checking your situation everything was working as expected

9

Okay. It is suspicious that my code to reject single-point glitches from the pressure sensor (the short segment I sent you) and not send that that point to the cloud, was somehow preventing recovery. None of the other sensors' data have needed such a glitch rejection filter and they had no trouble recovering from your data outage.
But, again, I appreciate your help.
David

10

This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.