Hyperconverged infrastructure (HCI) is a combination of servers and storage into a distributed infrastructure platform with intelligent software to create flexible building blocks that replace legacy infrastructure consisting of separate servers, storage networks, and storage arrays. More specifically, it combines commodity datacenter server hardware with locally attached storage devices (spinning disk or flash) and is powered by a distributed software layer to eliminate common pain points associated with legacy infrastructure.
Hyperconvered Infrastructure Basics:
How it works?
HCI converges the entire datacenter stack, including compute, storage, storage networking, and virtualization. More specifically, it combines commodity datacenter server hardware with locally attached storage devices (spinning disk or flash) and is powered by a distributed software layer to eliminate common pain points associated with legacy infrastructure. Complex and expensive legacy infrastructure is replaced by a distributed platform running on industry-standard commodity servers that enables enterprises to size their workloads precisely and to scale flexibly as needed. Each server, also known as a node, includes x86 processors with SSDs and HDDs. Software running on each node distributes all operating functions across the cluster for superior performance and resilience.
Hardware platform configurations are available to fit any workload by independently scaling the various resources (CPU, RAM, and storage) and can be provisioned with or without GPU for graphics acceleration. All nodes include flash to optimize storage performance, and all-flash nodes are available to deliver maximum I/O throughput with minimum latency for all enterprise applications.
In addition to the distributed storage and compute platform, HCI solutions also include a management pane to enable you to easily administer HCI resources from a single interface. This eliminates the need for separate management solutions for servers, storage, storage networks, and virtualization.
The birth of hyperconvergence technology
Datacenter infrastructure has been designed around SAN Storage since the 90’s to protect data and to power critical databases, and became pervasive with the explosion of virtualization in the early 00’s.
But as organizations have grown increasingly dependent on technology, traditional SAN-based infrastructure can no longer keep up with IT needs. It’s complex, unwieldy, and can’t scale as flexibly or efficiently as IT teams need to keep up with changing business priorities.
The world’s largest web companies faced the realities of traditional infrastructure’s limitations long before the broader market, and developed distributed systems technologies to meet their scalability, reliability, and operational efficiency challenges.
In 2009, engineers from several of these web scale companies realized that the technologies they had developed to solve their own operational challenges were applicable to the market at large. The realities of bringing these technologies to enterprise computing required a new approach, and the concept of HCI was born.
Today, HCI is the infrastructure of choice for companies that want to stay competitive and evolve with the changing realities of the technology landscape. While the actual date and person who first coined the term hyperconvergence can be up for debate, Nutanix was the first technology company to bring to market an HCI-specific product in 2011 called Complete Cluster.
The tie between hyperconvergence and cloud
Organizations increasingly utilize public cloud services such as Amazon Web Services (AWS), Microsoft Azure, and Google Cloud for deploying IT applications to run their business. Public cloud services are flexible and dynamic, and enable organizations to dynamically adapt to changing business needs.
But despite the increase in flexibility, cloud computing has its own challenges. Building and deploying applications in public clouds requires specialized skill sets that diverge from traditional IT teams, increasing the specialization in already highly siloed organizations. In addition, utilizing public cloud resources is more expensive than on-premises infrastructure and creates control and security challenges.
Hyperconverged infrastructure is underpinned by many of the same distributed systems technologies as public clouds, enabling IT organizations to build private clouds that bring benefits of cloud computing into organizations’ datacenters. Hyperconverged infrastructure services can also be extended into public clouds for true hybrid cloud infrastructure that enables applications to be deployed and managed with the same tools and procedures while making it easy to migrate data and services across clouds.
