A Case Study in Migrating to the Cloud with OBIA
Author: Anu Chaudhary | 5 min read | August 22, 2018
Many people have a misconception that the cloud is an all-or-nothing proposition. Either you keep everything in house – the data, the applications, the database, the hardware, etc – or your turn over your most valuable resource to strangers out there somewhere on the internet.
The truth is that there’s a whole spectrum of possibilities between fully on-premises and fully in the cloud. Here’s the true story of how one enterprise found a perfect middle ground using Oracle Business Intelligence Applications (OBIA).
The Original State
The enterprise was a well-known quick-serve café franchise on the national stage. They had massive amounts of data coming in from franchise stores everywhere and the businesses were rapidly expanding, straining against capacity. They needed to make their IT scalable but contain costs while they were doing it. The cloud was clearly the way to go to answer the scalability problem, but they weren’t entirely sure what that meant or how to get there.
Initially, their system that needed a cloud upgrade consisted of:
- An Oracle e-business suite running on-premises
- Connections to various non-Oracle HR apps that ran in the cloud
- A widely dispersed assortment of data marts
- Many instances of Essbase online analytical processing (OLAP) servers
- Integrations with Oracle cloud applications
- An old Oracle Data Integrator (ODI) running to an antiquated (2013 version) of OBIA.
At the end of it all, complex data analytics were being rendered down into excel for convenience in reporting, even though they had access to more advanced Oracle data visualization tools.
A significant percentage of their existing architecture was already moving to the cloud, but they didn’t know exactly what needed to go into the cloud next, or how expanding capacity in one part of the system would impact bottlenecks elsewhere in the network.
The Solution
Datavail completed an assessment and roadmap to determine which elements should move to the cloud, and which should stay on-premises. The proposal Datavail presented started with an upgrade of OBIA to 11g compute and 12c database, giving them the optimal range of choices for how to proceed next.
From there, our experienced analytics team suggested that development should be kept on-premises for maximum control, but user acceptance testing (UAT) and production environments should go to the cloud where traffic spikes and release schedules wouldn’t impact operations. Both connected to the Oracle Analytics Cloud (OAC), containing all the necessary Essbase components and Oracle data visualization tools.
The only downside to keeping development with on-premises servers is the enterprise’s access to new features. Oracle 12c is excellent for data protection and multitenancy, but as Oracle develops new tools for the OAC, they go to full cloud architectures first before pushing the latest upgrades to on-premises configurations. In this case and for the majority of enterprises in industries like retail or restaurants, speed and security come first.
Final Thoughts
Going fully to the cloud is not always the right solution for every company, though it does work well for some. When considering the options, our team suggests looking at a hybrid solution – especially with regards to BI applications – and finding the mix that works best for the environment at hand.
To learn how Datavail can evaluate your environment and build you a roadmap to cloud analytics, download our white paper, “Packing the Essentials for Your Journey to Cloud Analytics,” or contact us.