What is a Hyper-Converged Infrastructure... and Why Use It?

By Scott Drucker ·

Back in the day, cloud was the big buzzword. Are you in the cloud? Have your started to use the cloud? Every company, big or small, was offering some sort of cloud solution. Today, cloud is mainstream.

The new buzzword in my opinion is hyper-converged. Sounds like a mouthful. Say it with me slowly, h y p e r - c o n v e r g e d.  By defintion hyper-convergence or hyper-converged is a type of infrastructure system with a software-centric architecture that tightly integrates compute, storage, networking and virtualization resources and other technologies from scratch in a commodity hardware box supported by a single vendor.

Let me repeat that, "tightly integrates compute, storage, networking and virtualization resources and other technologies from scratch in a commodity hardware box supported by a single vendor".

Companies like Nutanix, SimpliVity, Dell, VMware, and Atlantis Computing offer such a solution.

But why would you want to use a solution like this versus the traditional separation of hardware vendors? What is the selling point, the upside, the benefit of changing everything you are used to doing today that works?  In its easiest form, simplicity.

Having a single piece of hardware that integrates your storage, compute, networking, resources, and virtualization gives you a single pane of glass into your infrastructure. If storage is compromised, you can see that from a single console. If your virtual hosts are over taxed, you can see that from a single console.  Need to deploy 100 VM's in a matter of minutes, you can do that from a single console. Need to expand your environment, you can quickly see what hosts are available and what resources you have available.  The administration of hardware and software has never been easier.

I've listed some of the benefits below:

  • Software Based
    • The software-based approach of hyper-convergence provides the flexibility required to meet current and future business needs without having to purchase new hardware components every year; and because its software based, new features in updated software releases allow customers access to those features and performance enhancements via a few clicks of the mouse in the management console.
  • Low Cost
    • Because all of the resources live in a single appliance, you save on the cooling and power consumption costs of having to maintain multiple physical servers.  Case in point, if you had to maintain four 2U Servers and an extra 2U Storage Server, the rack space cooling costs is 5x to a single 2U Server.
  • Commodity x86 Hardware
    • Using commodity hardware significantly lowers cost of ownership thus reducing the costs of acquiring the hardware in the first place.
  • Centralized Management
    • Because all components — compute, storage, network, etc. are combined in a single shared resource pool, it allows the IT Management team to more effectively manage the entire ecosystem. All data and metrics on everything running in the platform is a available with a simple mouse click.
  • Agility Streamlined
    • The ability to have all resources in all physical data center reside under a single administrative hyper-converged roof allows for easy and tactical workload migrations in the event of an environment failure.
  • Scalability and Efficiency
    • The ability to scale out quickly and cost effectively is always a concern for any business.  By implementing a hyper-converged infrastructure, you can simply bolt on additional appliances as you expand.  One click capability allows you to add drives and storage in an instant; without having to contact a storage vendor to come in and do this for you.  The control is back in the hands of your IT team.

Other benefits include easy automation, support of major hypervisors including VMware, Citrix, Hyper-V, and support for Virtual Desktop projects. You have redundant data protection, High Availability, Node Failure, and IOPS that can equate to 100,000.  100,000 IOPS was unheard of years ago. Today that's common place with hyper-converged; and those numbers will just get bigger.  Want to know the best part of all of this? It's software, and with a software release you can see IOPS continue to increase x amount of times over.