Quick update: I've been watching this over the past few days and the CPU and memory usage seems to have returned to normal. So I'm guessing those spikes are related to the ping issue we saw at around the same time which brought the server down. I will post updated graphs once the server has been running for a week.
As promised here is the CPU usage for the first week:
and here is the memory usage for the first week:
After the spikes at the start and the subsequent restart on the Wednesday, both CPU and memory usage appear to have stabilised and are even at slightly lower levels to ActiveMQ when I look back at the graphs for the past month. Good news!