This is the user on VM forums, a graph showing when they were having the problem - ignore the large red chunk.
And this is now after changing to DNS Forwarding mode.
That's me. The first graph above shows the SH3 going offline briefly around 3:30 am, which is when it did the firmware upgrade (I checked the Hub logs the next day). The three weeks prior to that red line (which was when I had the SuperHub replaced) are like the graph to the left of around 3:30am. Everything after that is like the graph to the right.
The first thing I did was turn off the 'add DHCP leases to DNS' entry in the DNS resolver in pfsense. That improved the situtation to this:
That was a dramatic improvement, but then a few days ago I ticked the box on the DNS resolver in pfsense to make it operate in 'forwarding' mode. I'm now getting graphs more like this:
Still room for improvement, but again, much better.
I should add that the same version of pfsense was running perfectly for around 6 months before the firmware upgrades to the Super Hub 3 started rolling out in around January of this year.
To conclude, there seems to be some sort of interaction between the recent firmware update and the 'resolver' mode of the DNS resolver in pfsense. I'd love to get to the bottom of it if possible.
Andy