Breaking a Monolithic Devops
summary
Describes a two-year migration from a monolithic EDW to integrated data marts that yields concrete results: deployment lead time drops 30 days to 2 hours, monthly deployments rise from ~30 to ~250, and an operations headcount falls 40 to 4 while shifting work to development. Implements templated CI/CD pipelines (Git + Jenkins), containerization and Docker, config tools like Ansible/Chef, Datastage/Hadoop for ETL, automated code review and monitoring, plus a semi‑local big‑data dev environment; testing and test‑data management remain the biggest unresolved risk.
AI-written summary · about
chapters (7)
- 0:00Breaking a Monolith
- 5:27The journey
- 14:41High-level CICD Infrastructure - Overview
- 24:41Bi Integration and Deployment Pipeline Legacy and new
- 28:36CI-CD pipeline - demo
- 32:03EDABI Mirror - Demo
- 34:03Our journey - what we achieved
related talks
33mUnleashing the Gorillacorn - DevOps in Very Large Enterprises
17mBeyond DevOps – Holistic transformations in uncertain times
28mDevOps in a Data Warehouse: Inside Out
43mAbout to Start Your Nth DevOps Transformation?
26mDevOps and Lean in Legacy Environments
30mWhat's It Like When DevOps Is Just Another Day at the Office?
26mDevOps in a Data Warehouse: Inside Out
23mDevOps and Legacy Code