November 23, 2015
The brain of a business is its data center. Everything, from customer accounts to financial records, is housed within the company’s data center. It is so important, in fact, FEMA has calculated that 40% of companies who suffer a disaster, lose their data, and have no way to recover it will close their doors forever. However, data centers are also incredibly expensive to own, operate, and maintain.
Whether on-premise (on-prem) or off, a company needs to house their data somewhere. So, great efforts have been made across the industry to make these centers as efficient as possible. This is an ongoing refinement of resources that will likely never end, but even today current designs have made great leaps over the data centers of the past. The current trend is a distinct shift
towards hyperconvergence.
What is HyperConvergence?
Hyperconverged infrastructure seeks to alleviate the headaches and inefficiencies of the modern data center by combining (converging) each component of the data center rack into a single device. Compute, storage, networking, virtualization, and data protection resources are integrated into a single software-centric architecture.
The best way to understand the challenges hyperconvergence aims to solve is by looking at traditional data center architecture and the steps that have been taken to reach where we are today.
Traditional Data Center Architecture
In a traditional architecture, individual components of the stack are purchased piecemeal from a number of manufacturers. The components are wired together manually within a chassis to create a stack. Then, every stack in the center is wired together to create a network.
There is no virtualization involved in this kind of data center, so every application the company needs to function requires its own server to operate. This approach leads to a few pain points.
- Complex Design Process: The sheer number of cords alone make designing a traditional data center extremely complex. Factor in the installation and configuration of each piece of hardware, and a data center deployment could quickly become a quagmire.
- Different Vendors: Each component of the stack is typically made by a different manufacturer, so shipping, building, configuring, and testing the necessary pieces can take an enormous amount of time.
- Prone to Errors: With so many parts needing connection, there is ample opportunity for error.
- Operational Maintenance: This architecture requires significant maintenance in order to keep functioning normally.
Reference-Based Architecture
Reference -based architecture attempts to ease some of these pain points by providing a data center template. IT directors would simply download the template or whitepaper from a major manufacturer and follow its instructions to build their data center. This removes some of the design responsibility from the IT director and reduces the opportunity for error. The in-house IT team now has a guide with which to build their racks. While this is a step in the right direction, it still has its fair share of deficiencies.
- Outdated Designs: The design templates are not always regularly updated. It is possible for an IT director to purchase and construct an entire data center full of components that are End of Sale.
- Time Consuming: The team still needs to order, build, and test the data center by hand.
- IT Departments Operate in Silos: A referenced architecture does nothing to address the problem of siloed experts in the IT department. Certified specialists are still required for each piece of technology in order to deploy correctly and maintain proper function.
Converged Infrastructure
A subset of reference architecture is converged infrastructure, the goal of which is to simplify the data center construction process as much as possible. Instead of several vendors offering the various components of the rack, multiple companies have combined together to offer a “data center in a box” with everything a company needs to house their essential business applications and data.
This approach has several key benefits:
- Single SKU: An IT director only needs to make a single purchase and wait for the components to arrive at the site.
- Ready-to-Deploy: Converged infrastructure such as the FlexPod, SmartStack, and VBlock, arrive at the company site ready to deploy. Because all the hardware has been designed for compatibility, installation and configuration time is drastically reduced.
- Best-of-Breed Solutions: Major manufacturers have partnered to develop these converged infrastructures, and they have each included their finest equipment.
HyperConverged Infrastructure
Finally, we reach what looks to be the next step in data center evolution, hyperconverged infrastructure. Championed by companies such as Nutanix and VMware, hyperconverged infrastructure is based upon Webscale IT and represents a pivot in how we conceptualize data center design. Abandon the notion of piecemeal hardware and converged bundles like the FlexPod, SmartStack, or VBlock. Now, the data center is 100% software defined and comprised of equal nodes.
A software defined dispersed node system has a few notable advantages for today’s data centers.
- Nodes: Each node includes compute, virtualization, network management, data protection, and storage in a single device.
- Resilient: A hyperconverged infrastructure exhibits excellent resiliency. Every node in the system is exactly the same, and responsibilities between nodes are shared equally. So, should a single node break down, the system can immediately redistribute workloads to different nodes and maintain service.
- Scalable: The environment can grow without limits. All scaling requires is the addition of more nodes.
- Ease of Management: Through the Nutanix and EVO:Rail software, the entire data center can be managed from a single screen.
- Simplified Support: In traditional architectures when problems arose in the data center, IT directors would need to contact multiple manufacturers to find a solution. Should trouble arise with an EVO:Rail or Nutanix solution, there is only a single number to call for immediate assistance.
- Efficiency: Hyperconverged infrastructures require far less power to run and occupy less space than most traditional data centers.
- No Specialized Skills Needed: IT departments do not require certified experts to deploy or maintain a hyperconverged solution.
Like what you read?
About Mindsight
Mindsight, a Chicago IT consultancy and services provider, is an extension of your team. Our culture is built on transparency and trust, and our team is made up of extraordinary people – the kinds of people you would hire. We’ve always prided ourselves in delivering the full spectrum of IT services and solutions, from design and implementation to support and management. Our highly-certified engineers and process-oriented excellence have certainly been key to our success. But what really sets us apart is our straightforward and honest approach to every conversation, whether it is for a local business or global enterprise. Our customers rely on our thought leadership, responsiveness, and dedication to solving their toughest technology challenges.
For Further Reading:
5 Benefits of Hyperconvergence (HCI): An Infrastructure Report