TechPrufer
Get started
Stripe
Stripe logoStripe · Feature trail

Database Fleet Auto-Remediation

2 posts · Jun 2024 – Jul 2026

What this trail covers

This post continues the evolution of Stripe's advancements in automating the management and remediation of its global database fleet. It details how Stripe's document databases, specifically their implementation of the Data Movement Platform, supported 99.999% uptime with zero-downtime data migrations. This builds upon previous work in using graph search and state machines to model MongoDB infrastructure, enabling dynamic computation and execution of recovery plans.

All Stripe trails
  1. How Stripe uses graph search and state machines to auto-remediate a global database fleetThis post details how Stripe's automated database fleet remediation system, which models MongoDB infrastructure as a traversable graph and uses pathfinding algorithms for recovery, reduced pager volume by 30% (~200 pages/year) and eliminated 12 days of unhealthy shard states annually. It also highlights the system's ability to support new shard layouts with zero manual effort.
  2. How Stripe’s document databases supported 99.999% uptime with zero-downtime data migrationsThis post details how Stripe's Data Movement Platform, built on top of their document database infrastructure, enabled 99.999% uptime during zero-downtime data migrations. It highlights the technical strategies and architectural considerations that allowed for seamless data movement without service interruption, contributing to the overall reliability and manageability of Stripe's database fleet.