Orchestration and IO performance optimization using storage analytics

The Cloud, mobile technology, IOT, Social Media has completely changed the storage needs as these applications are generating massive amount of data.
At the same time the IT organization is moving from siloed infrastructure to Software defined storage, Hybrid Storage. It’s a challenge for Data center administrator to understand the storage IO access patterns and tune the storage system to optimize for performance keeping the cost optimum.
The goal of this paper to define a solution on how to understand the IO pattern so that the storage system can be optimized using the deep learning and analytics. The deep learning analysis can be used as feedback into the Multi-Tire Hybrid storage system to have an adaptive caching system to optimize the read IO performance. Another objective for this solution to provide facility to test any new storage system with predefined load and compare the performance with the existing storage system.