For the billing month 26th March to 25th April, we have experienced significant increases to the Cellular Data (MBs), yet we have not implemented any changes for hardware, firmware, or added devices etc.
We are now exceeding the limit for the account which will be suspended on the 18th of April. As this has never been an issue in the past 6 months with the same devices, we need support to identify the root cause of the abnormal usage.
The devices in question are deployed at one site where we have a total of 7 devices. All devices use the Boron BRN310. Five devices use the embedded SIM and two use local SIMs (Airtel Rwanda). The devices have the same hardware and firmware.
Below is our analysis based on data from the console for the period 26th of March to 16th April 2024:
Three devices on embedded SIM use 41.39 + 26.72 + 38.67 MBs, an average of 35.59 MBs. These devices are on OS 4.2.0.
Two devices on embedded SIM use 5.2 + 5.17 MBs, an average of 5.09 MBs. These devices are on OS 4.1.0.
Two devices on local SIM and do not consume MBs. These devices are on OS 4.1.0.
The below is a screenshot of the data. We can also provide the Gsheet or Excel file with the full data.
Looking at the dates for the increase for the devices on OS 4.2.0, a significant reduction in data operations seems to align with the increase in MBs. The exact detail is difficult to gauge due to the lack of specific data on MBs per day, since that data is not included in reports generated on console.
We are also unable to determine why the increase occured or started at different times for the three devices? This is peculiar as the devices are identical, in the same location (within only a few meters of each other), on the same firmware and OS.
It's hard to say for sure, but there are numerous tips in the cellular data usage guide for possible reasons for data usage increasing but data operations not increasing or falling. The most common reason is poor network connectivity, but there additional reasons in the guide.
As Rick mentions, data usage can vary due to various reasons.
A key factor is the frequency of connections - each connection attempt will consume data, even if not successful. Devices that are programmed to aggressively connect will eat up MB quickly.
One possibility for your location is a local carrier re-aligning antennas/taking a tower offline. That can cause devices that previously worked to have trouble. New plant growth/structures/electrical equipment can also influence signal and cause excessive connections.
Nothing on Particle's end, or our cellular partners, have changed that would cause such a difference in usage.
In this instance, the devices are very close to one another. We are literally talking centimeters apart. Shouldnt the logic then be that any external influence should result in the same changes to MBs across all devices, in particular when the devices are identical when it comes to hardware and firmware. The only identified difference is the OS which we shall now try to change/update to see if that makes a difference.
In the meantime, the most pressing challenge for us is that the account is not suspended. Obviously, we are very much hoping that the support team can assist in identifying the root cause.
The issue has been marked as resolved. Small regional outages do not typically have post-mortems as the local carriers do not easily share information on why outages occur.
In this instance however we believe it was related to the multiple undersea cable cuts around that time - most of the continent had unstable connectivity for a couple of days.
RF is sometimes strange in the way it behaves - devices cannot be directly compared as two devices in the same room can be connected to different towers/carriers, or pick up on interference in unexpected ways.
Individual device antenna orientation and proximity to another RX/TX device will also affect the performance. This is not directly related, but it shows how multiple transmitters work together to shape the signal propagation pattern: https://youtu.be/49bwwQxUgcs?si=_nx35LvEzUGDfvkZ.
Making sure DeviceOS is identical will help to ensure the devices behave in a similar manner.
Regarding your usage - your account has already been unpaused until the 26th when the new billing period starts. We are not able to indefinitely suspend limits on a free Sandbox account due to usage that is beyond Particle's control. If your product is in an unfortunate location/situation where it exceeds the free 100MB limit, you will need to consider a paid Growth plan.
Data usage is driven by a device, it's behaviour, and local RF factors.
Changing the device behaviour (driven by firmware) will assist in driving down data usage.
