Running FoundationDB at Scale: Wavefront's Experience - Management and Architecture

Explore how Wavefront manages FoundationDB at scale in this 26-minute conference talk by Matthew Zeier and Mike McMahon from VMware. Gain insights into the management of 500 FoundationDB instances across 35 Wavefront clusters, with 400 running as both memory and SSD systems. Learn about automated cluster provisioning, expansion, and contraction techniques, as well as strategies for replacing entire FoundationDB fleets. Watch a live demonstration of using Terraform for quick capacity addition and Ansible for managing exclusions. Discover the architecture design of a single FDB instance, including disk layouts and instance store caching that enable high write rates of 800KHz for memory and 70KHz for SSD. Delve into topics such as Linux tuning, read caching, instance lifecycle, database configuration, fleet replacements, trace logs, event tracing, latency, observability, and metrics to enhance your understanding of running FoundationDB at scale.

Introduction
Tuning Linux
Read caching
Instance lifecycle
Database configuration
Landingparty
Fleet replacements
Trace logs
Event tracing
Latency
Observability
Metrics