Skip to content

Simon Griffiths

Thoughts about Technology and Architecture

Guide

Shaping the project

Waves, decomposition, scope reduction, cost and ownership.

The technical work on these migrations is largely well understood. The scoping, ownership and decision discipline around it is what varies.

Waves or big bang The most consequential project-shape decision, usually made by default rather than deliberately. Making an architected platform decomposable Dual-run the foundation; migrate the top layer by domain. Slice horizontally, not vertically. Buying credibility: the early-value shortcut Deliberately breaking the architecture to fund the programme — and how to make sure it gets thrown away. Finding out what nobody uses Invert the default: migrate what the evidence shows is used, not what nobody will admit is dead. The BI, ETL and consumption tail Owned by everyone and nobody. Where migrations actually overrun. Security, sovereignty and the path the data takes Security requirements constrain the method, not just the implementation. Measuring the whole data path The slowest stage sets the pace, and it is rarely the one that gets optimised. What it costs Dual running and non-production are the two costs that consistently surprise people. Operating ownership A managed service reduces operational work. It does not transfer accountability.

Reviewed 25 August 2026

These are my own views, written from years of building and moving large Oracle data platforms. They are not Oracle’s official guidance and this is not Oracle documentation. Volatile product claims are quarantined in Product facts that will date.

Pages

Start Here
About
Guides
All Posts

Categories

AI Architecture
AI-Assisted Software Engineering
Data Architecture
Oracle & Database Engineering
Microservices & APIs
Digital Safety
Friday Coffee

Series

Digital Safety Series
Craft & Code
Friday Coffee
Designing the Stack
APIs in the Age of Agents
Database in the Age of AI

RSS

Feed