"Target having two JVM instances, (three if you need more scale or redundancy), grow those instances up to the NUMA-node size but preferentially try to stay below either 4GB or 32GB max heaps; above NUMA-node RAM you need to test whether a bigger heap single JVM or per-NUMA-node memory restricted JVMs are more efficient"
"The probability that a customer will see a worse than a nth centile request is 1-(n)^requests, eg the average page makes 42 requests to display the page, then 1-(0.95)^24 = 88% of pages see the 95th centile or worse in latency!"
