With over 14,000 active employees and nearly 15,000 user accounts at any given time, Identity and Access Management was a major challenge. Add in the fact that over 200 applications are used on a daily basis, from point of sales, to financials and human resource information systems, you’ve got yourself a seemingly impossible task to manage identities securely and efficiently.
Arisant is a specialized Oracle Platinum partner that can provide you both expert IT services as well as the expertise to manage your IT budget and craft a vision for the future. Contact us to find out how our team can help your organization through its IT journey.
The client in this case study is a global enterprise which specializes in gaming machines and casino management systems. Their gaming analytics system, which allows casinos to capture and analyze user data in real time, as well as their casino gaming machines are used in many well-known casinos in Las Vegas and around the world.
As the enterprise expanded and deployed their gaming analytics system at multiple customer locations, they started experiencing several challenges. Complex manual installations, updates and configurations at each customer site became time consuming and error prone tasks. Furthermore, maintaining different versions of the system at different client sites, as well as customer created content, made configuration management challenging. Finally, the solution was based on outdated reporting software that could no longer meet the modern analytics needs of users and data scientists.
As the company continued to grow and support more and more sites, it became clear that the current approach was not viable. A new, modern and scalable solution would need to be designed to allow for fully automated deployments, system operations and change management across all customer sites. At the same time, when delivering updates to the various customer sites, the update mechanism should be able to preserve content created by end users at each site without manual intervention. A further constraint was that the solution should only be hosted at end customer premises – no public cloud or any type of third party hosted solutions were allowed.
Arisant looked at different options and determined that the best plan of action would be to design a solution based on Oracle Analytics Server (OAS) – an on-premises self-service visualization and augmented analytics platform. OAS would allow the client to bundle and ship their gaming analytics system as a ready to use physical (bare-metal) server appliance. The packaged solution would also allow end customers to extend it with their own content. The challenges of provisioning, deploying, and maintaining this type of turn-key solution at scale were solved with:
The result was a set of standardized, predictable, and repeatable processes for provisioning and maintaining the appliances.
Containerization
Everything is containerized so as to provide standardized environments – the OAS software components and their dependencies, the maintenance workflows as well as the change management workflows. The container images and build pipelines provided by Arisant ensure that the required versions and patches for all underlying software are shipped with new appliances and can be reliably updated on existing deployments. Additionally, all necessary testing and quality assurance is performed against the same containers before they are made available to end customers.
Maintenance and Change Management
Updates to the gaming analytics system are quite frequent, so Arisant developed workflows which allow the customer to push updates to deployment sites when they become available and automatically apply them on each on-premises appliance. Since each site can also create their own content on top of the prebuilt solution, the update mechanism ensures that this content is preserved and not overwritten by updates. With the old solution, custom content required to be manually saved and reloaded after each update.
Similarly, when the OAS software stack needs to be updated, new versions of the container images with the updated software are pushed to on-premises appliances. Automated workflows can then re-create the affected containers with their new image versions. The same method is also used for updates to the workflows themselves.
Automated provisioning of container-based appliances on bare-metal servers
The automated provisioning for the physical appliances is a simple three step process:
At the end of this workflow, an appliance is fully loaded with the operating system, required software, user accounts, containers, the gaming analytics system, and per-site configurations. It can then be shipped to customer sites ready to use.
In the past, deployments, updates and operations were performed manually at each site. With the new solution a new appliance, running OAS with the gaming analytics system can be produced in about 30 minutes. Updates can be applied to any customer site with the push of a button in a matter of minutes. Since updating the OAS software is only a matter of replacing the OAS container with a new one, the application downtime is reduced to seconds. Since everything is standardized and automated, human errors and inconsistencies between deployments are eliminated.
The solution requires almost no human interaction or expertise to deploy and operate. Complex manual installations and operations are now a thing of the past; using this approach of leveraging containers to build appliances the client can scale to infinity.