You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In smartfin-fw, there was a limitation that we could not connect to the Particle network more than once a second. There was some logic to rate limit the number of connections made.
Investigate whether such a limitation exists in Tracker SOM/Gen3 Particle devices.
Ensure that the logic of DataUpload and sf::cloud consolidates network connections to minimize churn on network state.
The text was updated successfully, but these errors were encountered:
So this corresponds to 33 requests per second across the entire fleet. We can probably incur many connect/disconnect events in rapid succession per device, but we want to avoid this anyhow to conserve energy. So we want to probably implement a per device limit. I think it will probably suffice to do a limitation of one publish per second, and an exponential back off for connect from once per second to once every 6 hours on connection failure. Probably want to do a connection timeout of 1 second or so, that way the system isn't sitting there churning, maybe an exponential gain of 2 or 3?
smartfin-fw3/src/cellular/dataUpload.hpp
Line 48 in 91b5bb9
In smartfin-fw, there was a limitation that we could not connect to the Particle network more than once a second. There was some logic to rate limit the number of connections made.
DataUpload
andsf::cloud
consolidates network connections to minimize churn on network state.The text was updated successfully, but these errors were encountered: