Hello, I'm in trouble with some devices trying to rejoin faster than they should, with some reception issue causing the Dev Nonce to be moslty used is a short period of time. When things are restored, device still have reconnection problem due to the devNonce saturation (these device does not implement the devNonce increment and use random values...)
I'd like to be able to reset the DevNonce history for a device or an Application to avoid this.
- Is that already possible (api, command line, setup to change the cache duration, restart...) ?
- Is there a way to make it done ?
- Where this cache is stored ? (db, redis ?) can we manually clear it ?
Background - unfortunately device developper are mostly not understanding how LoRaWan works and regulation and many devices try to rejoin en every 2s to 15 second basis. Where the device got noisy environment for a couple of day, or when a device is getting low battery and recycling on join request, it can easily reached 65K join requests killing the whole devNonce space and prohibiting device to rejoin once the radio/batteries conditions are restored. (For those who read, the rejoin faire duty cycle (about 0.01%) is to be applied according to the LoRaWan specification and a 1min wait on restart before any communication is highly recommanded)
Paul
