The following use case is a real-life scenario submitted by a CoreSite IT employee.
The compelling event for change:
Our IT Operations team implemented an Information Security program at the end of the year that centered on following SANS best practices and general industry standards, as well as receiving external consultation to ensure the most comprehensive business continuity plan with regards to infrastructure resiliency. During this process, we identified the need to set up a hot site, or a site with hardware, software, and network installed that was compatible with the production site. This would provide us with the ability to sustain operations in the event of a severe disruption of service with minimal or no downtime during failover.
As it stood, our workloads were split evenly across two production sites, each running approximately 150 mixed-use VMs, including file and web servers, SQL clusters, load balancers and a host of other critical systems to support the enterprise. Our production clusters were built on commodity hardware designed by our systems engineers and were mirrored in terms of network, storage and compute capabilities. Because of this, building a hot site would simply require a call to a hardware reseller to quote similar components, followed by travel arrangements to rack and stack the gear.
Reevaluating legacy IT architecture:
At the start of the year it was time to gather quotes from vendors, purchase equipment and put our plan into action. However, in a preplanning meeting, it was brought to our attention that our existing stacks were three years into their lifecycles and would need to be replaced the following year. Several of our engineers made a passionate case to move to a hyper-converged infrastructure.
Hyper-converged infrastructure would integrate our hardware components (network, storage and compute) into a single chassis, built using commodity hardware. The solution would not only significantly reduce vendor complexity through converging three separate solutions into one, it would also improve performance through collapsing compute and storage into a single box and eliminating the need for a complex storage network connection.
The hot site would not run production systems for some time, as we wanted to let the new equipment bake-in and see how it integrated into our existing infrastructure, which presented a rare opportunity to run a prolonged proof-of-concept in a preproduction environment. Although it was a bit late in the game to pivot, we agreed this was the right opportunity to evaluate new hardware and determine if it was a viable option for future buildouts.
Exploring a cloud solution:
After weeks of research, analyst calls, sales meetings and sizing discussions, we chose a vendor and requested pricing. Unfortunately, that pricing came in much higher than we anticipated due to support and hardware costs. We now had a choice: do we implement a hyper-converged solution with performance benefits and a streamlined support model, or do we deal with the complexity of building out our existing commodity architecture at a lower cost?
While the hot site was an essential component of our overall business continuity plan, we already had the ability to failover business-critical applications between our existing production sites. This relieved the pressure to make an immediate decision, and allowed us time to reevaluate how we would deploy capital towards this investment. This is when we looked inward and realized that we had secure, dedicated access to alternative cloud solutions within our own data center ecosystem.
The migration to hybrid cloud:
This revelation was the catalyst for our hybrid cloud evolution. We have now started the process of offloading targeted workloads to cloud providers within our facilities. This solution offers the streamlined performance and reduced vendor complexity we were aiming for in hyper-converged solution, while also addressing issues of cost and capital expenses.
Unlike the dilemma experienced when selecting a hardware vendor, the biggest challenge with implementing a hybrid IT solution was the lack of subject matter expertise involving cloud migration services. This was the first time that we implemented this type of environment, and it was difficult to know where to begin. Once again, we looked within our rich data center ecosystem and were able to bridge the knowledge gap with the help of managed service providers during the planning, deployment and post-deployment phases.
Reaping the benefits of hybrid cloud
While our cloud story is still being written, the benefits of implementing a multi-vendor, hybrid-cloud environment are apparent. Not only has our IT department lowered costs and increased performance, but we are able to enhance feature functionality for our employee end users.
By embracing our chosen cloud platforms, we are now able to offer more options to our employees due to the improved licensing structure and feature-rich offerings. For instance, users were previously licensed to use the application suite of services on one device, however after restructuring our agreement, they can now run applications on up to five devices while preserving the performance of on-prem hardware. This is important in a predominately mobile-era where users employ a wide range of devices from laptops, phones, and tablets. Additionally, we adopted a new enterprise storage platform that will allow the entire organization to collaborate on documents and is also accessible from mobile devices.
While business continuity was the original driver of change, we were able to vastly improve our IT infrastructures through reevaluating a hybrid cloud strategy. This ties back to the Business Continuity Plan (BCP) story as our users can still access critical systems such as email and file storage during outage events associated with our on-prem gear. Not only have we successfully implemented a business continuity plan to ensure the availability of critical systems, we have enhanced performance and created a streamlined solution for our organization.
Our hybrid cloud implementation has been a success due to the insights, capabilities and performance gains realized from leveraging the providers within our own data center ecosystem.