Blog: How to determine Oracle migration ROI for complex systems

A global pharmaceutical company has been supporting hundreds of applications on the Oracle platform. Their complex system became outdated over the years with 90% of end of life or extended support Oracle versions with very expensive contracts. They needed a cost effective solution, so they looked for a way to move their workloads to a higher value SQL Server solution. See how we helped our customer determine Oracle migration ROI (Return on Investment) for their highly complex environment.

Let’s see how we approached the analysis to demonstrate that our customer could achieve an Oracle migration ROI within 1-year by moving to SQL Server.

Customer’s problem

Our customer system was constantly growing in size over the years and eventually become highly outdated. This leads to poor maintainability and increasing operational expenses. Moreover, the product owner faces financial and reputational risks caused by data losses, frequent downtimes, and security issues.

So our customer needed detailed analytics across their system to find out all modernization opportunities and benefits along with upgrade costs.

We based our approach on analyzing the existing environment to differentiate the applications according to their architecture, functionality, migration complexity and possible database licensing savings in case of migration to SQL Server. At first, we collected information on Oracle versions across their databases. We find out that 90% of Oracle version (mostly 11g Release 2 or older) in their system are currently unsupported or in an extended support phase. So, instead of buying new licenses for all outdated Oracle databases, we offered migration to the latest version of SQL Server as the most effective solution.

Understanding the complexity and relationships between databases and applications

Let’s take a look how we assessed the effort and cost required to perform the migration. Knowing application architecture is very important for the correct assessment of migration complexity. In the existing customer’s system, more than 50% of applications are packaged. When talking about migration complexity, in case of custom application we have to convert all components that depend on the database connection. On the contrary, packaged applications already have versions that were specially designed to work on SQL Server. So, the effort needed to reconnect these applications to the new database will be minimal.

One of the essential steps during the database migration is the conversion of SQL code. For precise assessment of this migration step, we used SQL Server Migration Assistant (SSMA) for Oracle. This tool not only automates database migration processes but also shows the manual conversion effort. After running SSMA tool against source Oracle database, we reviewed automatically generated man-hour estimate for manual conversion. We manually reviewed all conversion errors reported by SSMA tool and defined automatic estimate more precisely basing on previous experience. As a result, we delivered detailed analytics on database migration effort.

Knowing the exact number of database objects is another key point in the determination of migration complexity. We developed a script to automate counting of objects across all databases. Usually, having a small number of database objects means low migration complexity. So we outlined that our customer should start the migration with these databases.

Building a decision tree to determine Oracle migration ROI for 1-year

The customer’s system included several hundreds of applications connected with around 800 databases. Simultaneous migration of that amount of applications is a time and cost-consuming task with profit shifted to distant future. So, we decided to group application into complexity categories and then focus on the set of applications with maximum migration profit. In order to visualize this approach we built a decision tree, where we categorized the applications to determine Oracle migration ROI based the following factors:

  • the number of licensing cores;
  • migration complexity;
  • possible savings in case of migration to SQL Server.

How to use a decision tree to determine Oracle migration ROI

The process used to determine Oracle migration ROI allowed our customer to choose the most profitable application sets as top-priority migration candidates. As a result, that would lead to short-run profit objectives and provide long term benefits with SQL Server innovations.

Deep dive assessment

Having complete migration picture, we could focus on high-priority migration candidates. We have chosen several most promising applications in terms of possible profit and performed deep dive assessment for them. For these applications, we assessed all components affected by migration. They included conversion of embedded SQL statements, ETL workflow, and reports. Using that information, we validated that migration could be performed with minimal costs and risks and in the accelerated timeframe.

Benefits

By using system analytics to determine Oracle migration ROI, our customer looks forward to achieving the following benefits:

  • deep visibility into customer’s system that allows making informed decisions in future;
  • effective migration plan with minimized risks and maximized profits;
  • impressive 1-year ROI by migration of a limited number of high-priority applications;
  • upgraded systems using the cost effecient and secure SQL Server platform.