Mercurial (acquired by IMS Health) is a Business Intelligence company that operates a cloud based service to deliver live decision support capability to its clients across mobile and browser based platforms. Mercurial’s customers include many leading pharmaceutical industry companies who require analysis in sales and marketing effectiveness, compliance and other business optimisation areas.
Mercurial needed to process large data sets on demand to meet customer requirements for up to date reporting. They needed a scalable solution because of the large size of their data sets and static sets. Their pre-existing technology methodology led to an explosion of on premise physical servers and then to large farms of static virtual machines that were unused for long periods of time. Significant amounts of time were spent coordinating and executing data processing, which consumed valuable staff and introduced occasional human error into a time critical environment.
The size of their data set was growing so fast that it was becoming challenging for them to provision additional storage capacity, servers, and application servers to their development and operations teams. They were also running out of storage space on their archival systems and were unable to store their large data sets efficiently.
They needed to build a robust and scalable private cloud infrastructure that would grow in a managed fashion and respond to the very dynamic business environment that they work in. The private cloud solution also needed to be able to integrate with their publicly hosted web front ends and web applications servers. They also wanted to be able to burst into public cloud capacity in times of heavy load.
The Aptira Solution
Aptira analysed Mercurial’s requirements and proposed an OpenStack private cloud solution that would seamlessly burst into Aptira’s or any other OpenStack based cloud. Aptira recommended Mercurial move towards an internal IaaS model. This gave their development and operations team the ability to scale and grow according to demand and brought a reduction in the overhead of managing physical servers. This model allowed the teams at Mercurial to acquire resources on demand with self-service and automated provisioning built in to reduce overheads and time to deploy. Aptira’s ADAPT (Aptira Database Analysis Processing Tool) coordinates the execution of jobs across the virtualised infrastructure.
The solution allows Mercurial to scale to greater than petabyte object storage for archival of processed data using cheap commodity hardware. Easy to use browser-based configuration management and control applications allow Mercurial staff to monitor, control and modify the execution of the services while Aptira’s ADAPT ensures that adequate resources are automatically provisioned to meet the workload.
New data is now brought into the reporting solution automatically. As reporting workloads increase, new server instances are automatically provisioned, data loaded and analysis automatically executed. Once the work is complete, the server instances are torn down and the infrastructure can be repurposed to other uses.
What once required a full person-day of effort each day is now fully automated, saving Mercurial over $100,000 p/a and improving the timeliness of the service to Mercurial’s clients. Furthermore, as Mercurial’s business grows, they have a flexible infrastructure model that can rapidly scale to accommodate new workloads both within their internal IaaS environment and into those external providers as required.