At the University of North Texas (UNT), Information Technology Shared Services creates and maintains the system-wide infrastructure that provides services to students, faculty and administration – from email to parking passes to provisioning. Our four separate campuses each require storage and compute resources.
Before we started our move into the cloud, each campus had its own storage and compute resources, using a traditional storage infrastructure of building block drive systems. They didn’t scale. They multiplied. And the configurations for each campus expanded in different directions, at different times with different systems.
The burden on my operations team was significant. Routinely, one campus would max out their storage while another had excess capacity. There was no way to take advantage of that unused capacity across the siloed campuses. Data management processes – such as tiering and others had to be done manually: application-by-application, sometimes workload-by-workload.
We began researching hybrid cloud solutions, including one from EMC built on the VMAX3. We wanted to transition existing infrastructure– storage, networking, and telephony– to a next-generation, centralized and consolidated hybrid cloud infrastructure.
VMAX3 Takes Us to the Cloud
We asked a lot, far more than most enterprise storage vendors could deliver. Of the companies we vetted, only EMC – with its Federation Hybrid Cloud – had the broad product portfolio, with VMAX3 at the core, and the vision that would get us across the finish line.
Much of my team’s time was devoted to ironing out performance issues around tiering across our four campuses. VMAX3 was the only solution that provided fully automatic tiering – allocating storage tiers at any time based on an application’s specific workload. With VMAX3, tiering became something we no longer had to think about.
VMAX3 solid-state drive support improves application performance and significantly lowers latency. Here too, auto tiering comes into play, as VMAX3 identifies which data from which applications demand high speed caching, and automatically (and dynamically) directs that application data to the SSD.
All of this resulted in a continuously available storage array with an 80 percent performance improvement for core database I/O. A lot of that resulted from EMC’s Dynamic Virtual Memory, which allocates cores and cache dynamically to meet the performance demands of our Oracle, Blackboard, Peoplesoft and other critical applications.
The broader EMC product portfolio delivers improvements beyond “just” I/O performance. Our primary data center uses VMAX; our secondary center uses VNX. EMC VPLEX brings them together into a continuously available infrastructure. Storage is spread across both data centers and presented seamlessly to our applications. EMC RecoverPoint and Data Domain Boost improve the performance, efficiency and protection level of our backup and DR architecture. With them we’re able to replicate efficiently, dedupe quickly, and overall establish a highly productive recovery workflow.
With EMC, we’ve dramatically improved performance and eliminated massive amounts of IT administration. We now have an expanding catalog of 60 services that our customers request via our self-service portal. And EMC’s technology serves as the foundation for our strategic move into the Federation Enterprise Hybrid Cloud.