Building and Deploying Netflix
summary
Netflix moves core delivery tooling from Groovy to Kotlin: Nebula (43 Gradle plugins) and Spinnaker’s orchestration engine Orca are gradually rewritten, with Orca’s new Kotlin queue processing about 7 million messages a day across four instances, backed by Redis and achieving near‑once delivery semantics. About 86 Kotlin projects exist at Netflix and new services (declarative CD) are being written 100% in Kotlin. Testing remains tricky—Spock and JUnit hybrid approaches continue.
AI-written summary · about
chapters (19)
- 0:00Introduction
- 0:38Programming languages at Netflix
- 1:04Netflix is culture
- 2:46Nebula
- 5:51Astrid and Niagara
- 9:34Spinnaker
- 10:06Cloud Control Plane
- 12:49microservices
- 15:20why Kotlin
- 20:54Java
- 22:36Scala
- 25:53Kotlin
- 27:46Kotlin Approach to Null
- 30:31Orca Queueing
- 32:27Orca Spring Boot
- 34:23Kotlin Queueing System
- 36:24Testing
- 39:29Future of Kotlin
- 41:39Outro
related talks
45mNetflix’s Secret to Safe Automation at Scale
37mNetflix + Gradle A Journey in Developer Productivity
29mHow Netflix Thinks About DevOps
29mFast, Safe, and Continuous -- Picking all 3 at Netflix
58mBuilding a Culture of Continuous Delivery
27mDeploying Docker Containers with Netflix Spinnaker on Kubernetes and Titus
31mHow Netflix Thinks About DevOps
21mKeynote: Cloud Native CD: Spinnaker and the Culture Behind the Tech