Hewlett-Packard Enterprise’s announcement in January that it would acquire hyperconverged infrastructure specialist SimpliVity for $650 million ensured that the trend toward hyperconvergence would stay in the spotlight.
Hyperconverged platforms combine computing, storage, networking and virtualization capabilities, all preintegrated and controlled by a single management layer. A key benefit is that, because those core IT components are so tightly integrated, data center administrators spend less time launching new resources and more time applying them to business needs.
The use of hyperconverged infrastructure (HCI) is growing rapidly, and, according to Gartner, deployments are expected to jump by 79 percent this year, based on nearly $2 billion in sales.
There are many competing hyperconverged solutions for small and medium-sized businesses to choose from, including from HPE, Cisco Systems, Dell EMC, Nutanix and VMware. What should SMBs be considering as they decide to shift to a hyperconverged setup?
HCI enables cloud-like operations inside data centers, and it lets businesses save money by slimming down their investments on physical server infrastructure. HCI can make data center operations more efficient and faster, and allow administrators to more easily add virtual machine and capacity.
Stu Miniman, an analyst at research firm Wikibon, told TechRepublic that “HCI falls into the larger trend of transitioning infrastructure to the equivalent of” Infrastructure as a Service for on-premises environments.
“Wikibon defines a True Private Cloud as a layer that has simplified operations and greater levels of automation and orchestration compared to traditional infrastructure,” Miniman said. “The goal of leading infrastructure solutions over the last decade has been to become invisible, allowing IT more agility and time to meet ever-changing business requirements.”
What are the main considerations SMBs need to sort through as they choose an HCI provider? Liviu Arsene, a senior e-threat analyst for cybersecurity and anti-virus firm Bitdefender, writing on the website Dark Reading, notes that HCI can “reduce the complexity of managing multiple hardware and software platforms, and lower operational costs by using hardware as a single pool of resources shared across multiple applications.”
IT leaders at SMBs can take those benefits to their company’s management as they promote HCI adoption. However, they also need to consider several factors that will help them decide what kind of HCI to embrace.
George Teixeira, CEO of DataCore Software, which provides storage virtualization, server array and resource management software, notes in Forbes that HCI lets companies “buy an efficient ‘rack and stack’ server that consolidates compute and storage resources to simplify and reduce networking complexity. This simplicity makes it easy to manage and grow your system.”
However, there are ways in which such systems can grow unnecessarily complex, he says. “While efficiently designed systems can achieve high-availability data access and protection against failures (by having data redundancy across two server nodes), most use three or four nodes to achieve acceptable data protection,” he says. “Adding these extra nodes adds complexity and cost — and the system is no longer simple. This is a key trade-off to understand when evaluating hyper-converged systems for your particular setup.”
Using more than one node allows HCI systems to sustain failure to one node while still protecting a company’s data, according to Teixeira, since it lets applications recover or “fail-over” to another node.
“This is useful in remote office environments, where the simplicity of using only two nodes is an advantage over the added costs of using three or four servers,” he says. “But while a hyperconverged system works for simple use cases such as a remote office with virtual desktop infrastructures — where only a few nodes are needed — it’s not always suitable for more challenging cases that require ‘real-time’ response or depend on input-output (I/O)-intensive workloads like databases.”
In situations where high performance and fast response times are critical (for instance, how quickly a user can query a database and get a response), Teixeira says that “many more nodes are required,” which can add cost and complexity.
As a result, Teixeira says, when businesses are evaluating HCI solutions, IT leaders “should ask for performance benchmarks that allow the comparison of performance — especially response times — across different kinds of hyperconverged storage to determine which is best for specific application(s) or use cases.”
An optimal tool for measuring database and business-critical workloads for response times and I/O performance is the Storage Performance Council’s (SPC-1) benchmark, he says.