Many enterprises running Oracle as their database of record are looking to migrate their implementations to the public cloud, to escape from the age-old problems with in-house system deployment: large capital outlays, time consuming infrastructure deployment and management, and lack of elasticity and adaptability.
The benefits of migrating to a cloud environment are compelling, and the message is getting through to enterprises: According to a recent cloud usage survey conducted by data virtualization company Denodo, 36% of organizations are currently in the process of migrating their data infrastructure to the cloud, while nearly 20% are in advanced stages of implementation.
However, it’s one thing to want to move to the cloud; it’s another thing to actually do it. Migrating the most important workloads in the enterprise to the cloud is a high stakes game—so it needs to be done right, with minimal interruptions in application performance and availability. This leads to the first two questions of any migration: “How much will it cost?” and “Do we do it ourselves, or enlist the help of a partner?”
D-I-Y or S-O-L?
Similar to home projects, do-it-yourself (DIY) can be an attractive proposition due to the perceived cost savings and the skills that one can develop along the way.
On the other hand, just like when home DIY projects go awry, hiring outside experts to fix what you’ve done wrong and complete the original project is always more expensive than if you’d just hired them to begin with, potentially leaving you sadly-out-of-luck (SOL).
Ultimately, whether or not an IT organization should conduct DIY Oracle migration comes down to the answers to three basic questions:
- Do we have the skills in-house to successfully execute an Oracle migration? These would not only include Oracle skills, but also the requisite cloud platform skills. These can be difficult to find, and it takes time to develop them internally, which is why it is relatively rare for organizations to have these skillsets readily available.
- Will my staff have the proper focus for an Oracle migration? People tasked with doing a DIY migration usually have other responsibilities, which can lead to their existing work suffering, or them not being able to focus enough time and effort on the migration. These are hidden costs that need to be considered.
- Do we have a really good understanding of our environment, including how our applications are used, the number of users, and the metrics required to fully understand migration paths? This can be a challenge because many Oracle systems have been in place for many years, and the people who built them have often left the business, which can burden migrations with an enormous amount of “technical debt.”
To accurately answer these questions, one must first fully understand the extent of cloud and Oracle skills that are required to conduct a migration. To do this, it can be helpful to examine the critical phases of an Oracle migration to the cloud, and determine if resident skillsets are sufficient to execute the project. These phases include:
- This is how IT organizations can answer the second question above. A thorough assessment will detail the current Oracle operating environment, both from a technology and end-user perspective. The assessment will also identify any open security issues with the environment and define the steps required to ensure data integrity during migration.
- Proof of Concept. This phase requires a thorough understanding of the target cloud environment, and how the different features and capabilities can be used to support the Oracle implementation, so the database and workloads perform up to required standards.
- Test, test and test again. Cloud infrastructure and every component subject to the migration must be tested to ensure a successful outcome.
- Cutover plan. A comprehensive and detailed plan for cutting over to the new cloud implementation is required to avoid business disruption.
The prospect of migrating Oracle to the cloud can be highly attractive to IT personnel, because it is an excellent resumé builder to be able to take credit for a successful migration. But, resumé building is hardly an adequate reason for undergoing a DIY migration—so the personnel involved must be highly skilled and demonstrably able to attain the skills required for a successful migration, or else seeking the help of a third-party consultant is likely a better decision.
Tools for the Job
Cloud providers offer tools and services to assist with migrations, including AWS Migration Services, and Azure Migrate. However, these services are often not comprehensive enough for Oracle implementations, particularly if multi-cloud strategies are utilized. For these more demanding migrations, there are also Oracle migration templates and images from private companies.
Finally, should an enterprise wish to actually migrate off Oracle and onto Amazon Aurora or another flavor of MySQL or PostgreSQL, they will need to consider the skill and work required for database refactoring. There are free schema conversion tools available from cloud providers, but they will only get enterprises about halfway to where they need to be for a data migration. To complete the migration and ensure that all applications are running properly, the rest of the data refactoring will need to be performed by IT staff with appropriate “tribal knowledge.” Refactoring can be very time consuming, so organizations may want to enlist the assistance of a partner with the skills to advance the migration—however, even under the best of circumstances, there will still be the need for IT staff to help with organization-specific coding, although it should only be for about 20% of the project, not 50%.
Which brings us back to the original question—when is it viable to do a DIY Oracle migration? If an IT organization has the appropriate skills and capacity to perform this ambitious project, then yes, it is possible. However, if there is any doubt about staff skillsets or capacity, then acquiring the assistance of a partner experienced in Oracle migrations is a better route. The important thing is to conduct a dispassionate analysis so you can make the right choice—otherwise, a venture into DIY could leave you SOL.