DevOps at Ordnance Survey: Turning the Dinosaur Into a Unicorn
summary
Ordnance Survey converts bulky legacy geospatial systems into an automated pipeline using infrastructure as code, Eclipse, Puppet, Git and open‑source tooling to provision VMs and run vendor install scripts. Embedding infrastructure engineers in projects and moving to a single common backlog shifts provisioning left: a simple VM drops from two days to 30 minutes, environment builds fall from four weeks to half a day, backlog lag drops from two weeks to half a day, and the first 10 projects saved about £1m in 12 months. Continuous delivery is next.
AI-written summary · about
chapters (20)
- 0:00Intro
- 1:03ABOUT ME
- 1:51OS BUSINESS CONTEXT
- 4:05OS INFORMATION SYSTEMS STRUCTURE
- 6:12THE BUSINESS DRIVERS
- 6:49THE CHALLENGE PARTI - TECHNOLOGY
- 8:31THE CHALLENGE PART 2 - ORGANISATION & METHODOLOGY
- 10:46VALUE STREAM ANALYSIS OF FEATURE DELIVERY
- 12:40INFRASTRUCTURE PROVISIONING METRICS
- 14:53STEP 2 - AUTOMATION
- 17:31INFRASTRUCTURE DELIVERY PIPELINE
- 19:55WORK IS SHIFTED LEFT
- 20:34MEASURING PROGRESS
- 21:19SINGLE COMMON BACKLOG Backlog
- 22:50BRING OPERATIONS CLOSER
- 25:13SUCCESS ENABLES FURTHER ROLLOUT
- 26:10MEASURING SUCCESS - PROVISIONING STATS
- 27:09MEASURING SUCCESS - THE KNOCK-ON EFFECT
- 27:48INTANGIBLE SUCCESS
- 29:04LESSONS LEARNED
related talks
36mDevOpsSup - Extended DevOps approach
45mDevOps Transformation — Breaking Down Walls
17mBeyond DevOps – Holistic transformations in uncertain times
27mCultivate DevOps
55mHow to make an ops team get along with N development teams
24mHow DevOps changed the way we operate software
35mDevOps. Evolution of the Species
29mDevOps Transformation at Scale