Hyperconverged Computing: What is it, and why should you care?

Written by Steve Garnjobst, Senior Solutions Architect, Datatrend Technologies, Inc.

Hyperconvergence is an emerging IT market trend that has been gaining momentum in the IT infrastructure marketplace of late. Hyperconverged vendors promise lower costs and increased speed of application deployment. Hyperconverged solutions are a relatively recent phenomenon, with industry pioneer Nutanix shipping their first products in 2011. The ensuing four years have seen hyperconvergence evolve from a niche market to a significant emerging trend with dozens of alternatives and virtually all the major server manufacturers offering some sort of hyperconverged solution. So, what is hyperconverged computing, and why should you care?

The emergence of hyperconverged solutions is largely a result of innovations in two related fields: web-scale/scale-out computing, and software defined storage (SDS). Web-scale refers to the type of server infrastructure initially developed for data centers at companies like Amazon, Google, and Yahoo – who needed a cost-effective, massively scalable architecture to support rapidly expanding data centers. The resulting solutions are typically built using commodity rack-mount servers with onboard compute, storage, and networking resources. These standardized servers are combined to form highly scalable pools of resources for applications. While early web-scale deployments proved highly effective for specialized web companies with a relatively narrow range of highly customized applications, further development was required to support the wider range of applications found in most enterprises. That’s where developments in SDS come into play.

At a basic level, SDS involves de-coupling control of storage resources from the physical hardware. While hypervisors such as industry-leading VMware support virtualization of compute resources, storage resources are still largely limited by physical hardware and vendor platform limitations. SDS solutions seek to remove this limitation by aggregating physical storage hardware into pools of storage resources. This virtualization of storage hardware, when combined with the rack-mount building blocks from web-scale computing, provides the foundation for hyperconverged solutions.

The hyperconverged marketplace is generally separated into two tiers: vendors offering complete hardware/software solutions as an alternative to traditional enterprise storage, and vendors offering SDS applications that enhance the functionality of customer-provided infrastructure. Figure 1 highlights key hyperconverged and storage vendors:


Hyperconverged solution vendors typically claim a wide range of benefits, including:

  • Fast Deployment
  • Management Simplicity
  • Operational Simplicity
  • Simple Elasticity/Scalability
  • Cost Efficiency

While hyperconverged vendors are generally claiming the same benefits, the solutions are not “one size fits all”, as significant differences exist between the various options. In particular, many of the solutions only support VMware-based environments, or only support specific hardware platforms/configurations. Such narrow requirements can radically speed deployment and simplify management, but don’t support all use cases or workloads. Scalability also varies widely between the options, with many vendors setting relatively modest limits on their solutions, due to concerns with complementary infrastructure components or simply due to the reality that most hyperconverged solutions are at a relatively early point in the development cycle.

For smaller companies, a hyperconverged solution may be capable of replacing your entire existing compute/storage infrastructure. In fact, SimpliVity claims that 65% of their customers run 100% of their applications on a SimpliVity hyperconverged solution. However, hyperconverged deployments for larger enterprises are typically focused on specific workloads – with Microsoft SQL, Exchange, and VDI being common use cases. Specialized applications such as CAD/CAM and industry-vertical software have also proven a good fit for hyperconvergence. Finally, hyperconverged deployments for enterprise applications such as Oracle and SAP are becoming more common, as hyperconverged solutions become more established and stable.

The Datatrend team has a strong focus and extensive experience with a diverse range of hyperconverged options – partnering with industry-leading vendors such as Atlantis, Cisco, Dell, HP, IBM, Lenovo, Redhat, SimpliVity and VMware to deliver turnkey hyperconverged solutions. Please contact us to discuss the applicability of hyperconverged computing to your specific environment and use cases.

Dunk, C. (2015). Cisco Hyperconverged Computing Strategy (BRKCOM-1125). Presented at CiscoLive! 2015, San Diego, CA.

TechValidate (2014). Survey of 55 users of SimpliVity OmniCube. TechValidate content identifier: TVID 0EE-75E-9F4.

Morgan, T.P. (2014). Hyperconverged Systems go mainstream. EnterpriseTech, October 15, 2014. Retrieved from: http://enterprisetech.com/2014/10/15/hyperconverged-systems-go-mainstream/

Steve Garnjobst is a Senior Solutions Architect with over 25 years of experience designing, and deploying complex IT infrastructure solutions. Steve may be reached at steve.garnjobst@datatrend.com.