Building and operating at scale with feature management
summary
LaunchDarkly explains how feature flags enable live migrations, compliance variants, and architecture shifts with measurable safety: uses ALB, Kinesis, DynamoDB, Elasticache, Elasticsearch and server‑sent events for 24 million concurrent clients and an analytics pipeline doing 16M RPS and 150 TB/day. A MongoDB→CockroachDB migration runs via per-collection flags with four modes (A, A/B, verification sampling, B) across 13 teams; a Graviton rollout saves ~$91k/year and ~20% cost; customers count ~4,000.
AI-written summary · about
chapters (7)
- 0:00Introduction to feature management
- 3:35What is feature management
- 14:55Evolution of LaunchDarkly
- 29:52App modernization with flags
- 39:02Compliance and feature flags
- 45:01Cost and performance optimization
- 47:49Key takeaways and Q&A
related talks
07mScaling with LaunchDarkly: A feature toggle story!
54mFeature management: De-risking migrations and releases for PowerSchool
41mFeature Flags: Separate Deployment from Feature Release
05mFeature Flags
22mBeyond Boolean: Solving Hard Problems in Feature Flag Architecture at Scale
53mTinkertoys, microservices, and feature management: How to build for the future
27mTinkertoys, microservices, and feature management...
30mChoose Your Own Deployment: Interactive Feature Flag Adventure