Beginner: We have the problem that our sensors are often inactive in blocks

Hello, I am a beginner and want to ask you for help with a problem. I am a member of a group doing a scientific project in Berlin with CO2-sensors. We have about 75 sensors in The Tings Network and we also installed (and will install) new gateways. But now we face a problem, since a lot of our sensors have blocks (of days or even weeks) of inactivity, i.e. the sensors do not transfer any data anymore. However after this block of inactivity the sensors start working fine again. What could be the reason for this? Does this sound like a problem of the sensor or could this be a problem of the network? (The problem also exists for sensors close to our installed gateways) Thanks a lot in advance!

I have not seen any Network issues or repotted issues on the network, so I will say that is ruled out.

The question is where are the sensors in relation to the gateway?

  1. Distance between gateway and sensors?
  2. What buildings, trees or other obstructions are between the sensors and gateway?
  3. Were the gateways actually online? (Are the gateway yours or community gateways?)

Where is the data missing? In the gateway console, in the application/end device console, in the integration/back end database. As noted no major reported issues wrt network in general and if they ‘restart’ sending data after a period of loss unlikely to be device firmware/NS software issue so I would investigate back-haul and integration issues - look concurrently and see where data goes astray…

Has the uplink counter reset or has it increased during the time no data was reported?

1 Like

Check the live traffic feed from your gateways (or even better use whatever remote administration solution you hopefully deployed these with to connect in to your gateways and check their logs locally) to make sure that they’re seeing raw packets now and then. You can probably even unreliably correlate these to the device addresses of your nodes. Your gateways losing their Internet backhaul connection to TTN could be another cause of an area-specific outage.

Thanks a lot for your ideas. Since we have 75 sensors, it’s different for every sensor. But I figured out that we have 4 Sensors in the same building, close to one of our gateways and all of the sensors stopped at some point with transferring data, but not all of the 4 at the same time.

Thanks for your ideas. I will check this.

Thanks for your response. Actuall I don’t know, so I will check this.

Thanks for sharing this idea, I will check also this.