A history of GOV.UK's infrastructure: the long voyage to Kubernetes
summary
A decade of platform moves from VMware vCloud through multiple datacenter migrations to AWS ends with a replatform to Kubernetes live since March. Infrastructure uses vcloud-tools, Puppet, Jenkins/Capistrano, many databases (Postgres, Mongo, Elasticsearch) and a bespoke monitoring stack; a postcode-driven traffic spike causes mass 500s and 30 minute VM scaling is the catalyst for change. The current stack runs containers with Helm, Prometheus and Argo CD, autoscaling cuts deploy/scale time to minutes.
AI-written summary · about
chapters (8)
- 0:00Introduction and speaker background
- 1:28Early history of GOV.UK product
- 4:57Codebase decisions and architecture
- 9:27Early infrastructure choices
- 15:28Challenges with existing platform
- 18:21Exploring alternatives and Kubernetes
- 24:17Migration to Kubernetes
- 26:24Lessons learned and Q&A
related talks
39mUsing Kubernetes to Change Legacy Systems and Processes in the Public Sector
17mOur Journey to Kubernetes at Scale
34mKubernetes from Dev to Prod at GoEuro
35mKubernetes journey
35mPerforming Infrastructure Migrations at Scale
31mMigrating Legacy Infrastructure to Kubernetes: Olark’s Lessons From the Trenches
05mBare Metal to Containers to K8s: How We Got Here
31m5 Years with K8s: Container Deployment at Pôle Emploi