Every now and then (every 20min?), motes print error messages indicating that their time correction is large. A couple of observations:
sometimes this happens in bursts, for both motes
the number of ticks can be always the same (e.g. always either 14 or 25 ticks)
sometimes this results in the mote(s) desynchronizing.
In the design of the IEEE802.15.4e layer all radio/timing activity happens in an ISR, so there should be very little jitter on the timing.
This is totally strange and we should get 100% to the bottom of this before dealing with (potentially more complex) timing-related tasks such as FW-10.
1 DAGroot, 2 motes. Forcing topology to a linear network.