Data Volume vs. Data Operations

Connectivity Cellular
1

In May. I am seeing very high cellular data usage (90.8 Mb) as compared to data operations (4,464). This occurred while operating on the Free Plan in my Sandbox. This activity was driven by 14-15 B-series units using LTE in a small city (plenty of cell coverage).

My question is: how can I drill down to see where all of this cellular usage is coming from?

I know that cellular usage is driven by more than Data Operations (mine are generally 175 bytes each). I have read that it also includes OTA code flash, overhead, and 3rd-party services. I have not done a huge number of OTA updates in May (guessing 20-30 total, at about 8-10 sec each - 71Kb RAM is typical). My .bin files run about 48Kb. Best I can tell, that leaves overhead, as I don't engage 3rd party data services. I have a few units which seem to be disconnecting and reconnecting a lot (5-10 times/day, despite cell signal strength > 75%) - if there's overhead associated with that, I suppose it could drive the usage up.

Is there a utility which will show me the source of cellular data consumption?

DataUsageMay2023
DataUsageMay20231350×641 105 KB

2

Hi there!

Try this tool: Cellular data usage | Troubleshooting | Particle

Will likely help quite a bit!

3

Excellent tool - as you predicted, it helped me a lot. I have identified three of my units with unusually high usage (as compared to the other 12). They are running the same .bin as the rest, though (it's a product). Is there a utility I can use to break down the source of the usage (separating Data Operations from OTA , "overhead" and 3rd party - which I think should be 0)?

I have a new .bin I plan to drop after this weekend (this weekend is a critical one for my devices, and I'm reluctant to drop new software prior to Tuesday). It may help, but I'd also worry about the "Huh, it got better effect," because it can just as easily get worse again!

Thanks for pointing me to that utility. I plan to keep using it to monitor fleet health and identify outliers.

Bill

4

Hi Bill, if you DM me the DeviceIDs of a happy unit and an unhappy unit, I can take a look next week.

1 Like