How to get the CPU and memory details of Kong and Cassandra nodes for monitoring purpose?
I am using an application called Netdata to get nginx server metrics exposed visually real time + CPU/System metrics + can also use it for Kong StatsD api metrics - https://github.com/firehol/netdata , you can also have Prometheus scrape Netdata if you want your metrics in storage for a longer amount of time.
Cassandra I will probably end up going with a JMX Exporter but I have not solved that piece yet. Looking forward to seeing if Kong devs have any secrets to spill the beans on things they have seen for best practice 
-Jeremy
For monitoring infrastructure details like CPU and memory usage, you’ve got lots of options - for example, Datadog and New Relic both offer monitoring at the OS and Nginx-level.
Kong Vitals is a new feature in Kong Enterprise Edition that monitors and reports 1-second resolution metrics like Kong latency, usage of Kong’s internal cache, API requests per consumer, and more. Reporting is available via both Kong’s Admin API and the Kong Admin GUI.