How Kubernetes Was the Secret Sauce in Our Globally Distributed Team’s Transition to Microservices
summary
A remote engineering team of ~25 engineers (company size 80) replaces a monolithic app with Node and PHP microservices on Kubernetes, cutting deploys from ~3–4 minutes to 45–90s deploys and moving 54% traffic on k8s so far. The team builds a Slack deploy bot that runs kubectl rollouts and rollbacks, standardizes logging with Fluentd -> Kinesis -> Elasticsearch (DaemonSet pattern), adopts Datadog for metrics, and uses an exemplar service to validate autoscaling, canaries and rollback workflows.
AI-written summary · about
chapters (22)
- 0:00Intro
- 0:35Outline
- 1:06About Me
- 2:16Our Infrastructure
- 3:17Our Remote Team
- 5:47Our Challenges
- 8:11Roadhouse
- 11:07Patrick Swayze
- 16:20Deploy Bots
- 18:15Deploy Rollback
- 19:24Logging
- 21:40Besidecar
- 23:37FluentD
- 27:40Standardizing
- 30:15Building an Exemplar
- 30:58Goals
- 33:37Progress
- 36:17Whats Next
- 38:53Wrap Up
- 39:31Thank You
- 39:54Automation
- 44:32Not yet
related talks
20mKeynote: The Challenges of Migrating 150+ Microservices to Kubernetes
44mMigrating a Monolith to Kubernetes
36mUsing K8s to Power Microservices at Scale: The GitHub Way
35mTo Infinite Scale and Beyond: Operating Kubernetes Past the Steady State
46mModern DevOps with Kubernetes
46mAround & After Kubernetes: The Principles and Ideas that Guide Us
46mKeep Building Fresh: Shopify's Journey to Kubernetes
35mKubernetes journey