Since it hasn’t been mentioned in this forum, let me share a little performance tuning tip.
When you’re running FoundationDB on a VM (e.g. AWS EC2),
use tsc clock source instead of xen clock source whenever possible.
Our in-house microbenchmark showed significant differences, especially for small reads.
GET: 280064 reads/sec => 324192 reads/sec (+15.8%)
GET RANGE (5-key range) 205648 reads/sec => 235312 reads/sec (+14.4%)