The right solution depends on the task at hand

Back in the day, as a Delivery Manager, I had the opportunity to lead a project focused on migrating and upgrading an outdated Oracle Exadata platform for a client in the financial sector. The primary objective was to restore the platform to a supported state, minimize business risks and costs, and ensure the stable operation of mission-critical systems.

Key Aspects of the Approach

1. Parallel Infrastructure: We gradually rolled out the new on-premise infrastructure alongside the existing system. This approach helped us avoid downtime and mitigate potential risks.
2. Phased Transition: The migration and upgrade were carried out in stages, starting with lower-tier environments (Dev, QA) and concluding with production. This phased approach ensured safety at every step and allowed for early detection and resolution of issues before they impacted production.
3. License Optimization: A phased migration strategy enabled the client to avoid additional Oracle licensing costs, ensuring a more efficient use of their budget.
4. Go-Live Strategy: Due to limited go-live windows and other constraints, including licensing considerations, we opted to roll out the new DR environment on Oracle 12c first. After performing a switchover to declare it as the primary environment, we proceeded with the upgrade to Oracle 19c. This two-weekend go-live plan ensured a seamless transition without disrupting business operations.

Why Oracle 19c and On-Premise?

• Oracle 19c vs. 21c: The decision to choose Oracle 19c was driven by the client’s specific needs. Compared to the then-new Oracle 21c, version 19c was more stable, reliable, and offered long-term support, making it a safer choice.
• On-Premise vs. Cloud: The project’s primary objective was to perform a migration and upgrade "as-is" within tight deadlines, without altering or optimizing the business logic. Migrating to the cloud would have required significant redesigns, introducing unpredictable costs and delays that didn’t align with the client’s priorities.

Outcomes

• The migration and upgrade of the platform were successfully delivered by a relatively small team in just over 10 months, from kick-off to full production transition.
• The optimized architecture significantly improved data warehouse performance, fully meeting the client’s requirements.

Lessons Learned

This project highlighted that, even in the era of widespread cloud adoption, the most suitable tool depends on the specific task at hand. In this case, a non-cutting-edge solution proved to be more appropriate and effective.

What are some examples of "fit-for-purpose" solutions you’ve encountered?