Start Small, Scale Big: Building and Scaling Platforms and SRE Culture...
summary
Starts small with three microservices and five engineers, then platformizes infrastructure as reusable Terraform modules, moves from Elastic Beanstalk then ECS-style container deployments, and builds a productized internal platform that cuts release time from 30+ minutes to ~10 minutes. At scale it supports eight products and ~25M API requests/month with two platform engineers, ELK logging, Packer AMIs, PagerDuty on-call, and developer-owned CI/CD. A second case keeps a monolith, uses GitHub Actions and migrates from Datadog to self-hosted Grafana/Loki/Tempo to save costs while preserving continuous delivery.
AI-written summary · about
chapters (13)
- 0:00Start
- 1:47Platforms Engineering
- 4:56Reliability & Platform Engineering
- 5:46Alignment of platforms with the needs of the business
- 6:54Staying adaptable while scaling
- 10:05Setting up SRE culture from the first Engineer
- 11:57Adapting SRE philosophy based on constraints
- 13:46Decisions regarding suitable technologies and services
- 16:25Synergy between Platform Engineering, DevOps, and SRE Principles in Practice
- 29:08How architectural decisions flow into the Platform?
- 29:47Impact
- 34:19State of Platforms in June 2023
- 38:29Framework for setting up a Culture of Site Reliability Engineering
related talks
45mFrom 5 to 100 in Product Development team in just 5 years
56mSmarter, Not Harder: Scaling without burning out in an always more landscape
31mBuilding an Enduring Culture of Engineering Excellence
33mMy Journey to SRE
43mThe Scale-up, Autonomy and the Nuclear Submarine
12mKeynote: What It's Like to Bring DevOps to a Software Factory in Central America
30mThe hidden art of scaling yourself and your app through 15million users
32mScaling Up for Scalable Apps: The CIF-N26 Experience