Smoothing the continuous delivery path- a tale of two architectures
summary
Compares two delivery journeys: a greenfield stack of Scala, mongodb, Docker and a constrained .Net/Windows monolith, with one platform running 300+ microservices and just ~20 webops people. Shows concrete practices that reduce pain: think about production first in standups, make testing an activity, use release-notes automation that pulls commits and story states, ship dashboards (Kibana/Grafana), runbooks and per-team ownership. Reports fewer incidents on the microservice side, weekly releases on the monolith become manageable, and the team avoids internal API versioning by rolling forward with notice.
AI-written summary · about
chapters (19)
- 0:00Introduction
- 1:26Benefits of continuous delivery
- 2:30Etsy and Netflix
- 3:15The two very different teams
- 3:56No single silver bullet
- 4:54Learning French
- 6:09Production all the time
- 7:06Production in your standup
- 8:29Testing activity
- 10:43Astra Hawk story
- 12:28Root cause analysis
- 17:48Release notes
- 20:20HMRC Digital best practices
- 23:53Breaking changes
- 25:16Questions
- 30:20Hiring the best people
- 31:21Monitoring tools
- 32:22Automated dashboards
- 33:14Runbooks
related talks
45mUsing modern .NET technologies to decompose a monolith into micro-services
49mCase study of a good enough architecture
23mOn the road from Monolith to Micro
53mThe Road to Continuous Deployment: a Case Study
01h 02mTools for simplified large scale systems
01h 00mBeyond the innovator’s dilemma
27mBuilding a Better Software Supply Chain
36mIt's Simple...