English | EN HOW HYPERCONVERGED INFRASTRUCTURE CONVINCED ME?
August 10, 2015
Following my colleague Philippe André’s blog post about Hyperconvergence, I’d like to add my point of view regarding this topic.
A few years ago, the data-center infrastructure was a client-server generation. Then arrived Virtualization — with different tiers for compute (standard servers) and storage (SAN), connected with the network. Today, we are in a Cloud generation, where hyperconvergence is redefining the way we work. Hyperconvergence integrates compute and storage in the same box.
Hyperconvergence offers multiple benefits. I’ve had the opportunity to work with the Nutanix system. And I am putting forth my point of view, based on my experiences on this system.more–>
Cost-effective. First of all, the cost. With a traditional infrastructure, you have to invest in the storage technology, meaning the first investment is very high. With hyperconvergence, the concept of “pay as you grow” is making its mark on business models. Indeed, you buy only what you need. If you need more compute and storage, you just have to buy one or more nodes, depending on your needs.
Simple to manage. With most storage technologies, you need some specific skills. With hyperconvergence, you don’t. It’s so easy to install, deploy and manage. The best word to describe this benefit is “simplicity”. I’ve never been an expert in storage; but now, I can easily manage the storage of my clusters in the data-center.
Easy to scale up. With hyperconvergence, you can very easily scale up your infrastructure. If you need to add some nodes to your cluster, you can do so with just three clicks of the mouse. And with Nutanix, there is no limit to the number of nodes you can add.
High performance. Finally, you are assured of high performance. On a Nutanix node, you can have a part of the storage on the SSD disk, and the other part on a standard disk. When a VM is accessing a data pretty often (hot data), this data is stored on the SSD disk and assures you of high performance. When it is less often accessed (cold data), this data is stored on the standard disk. The storage is a “software defined” model, meaning that it is managed by a specific software hosted on the node itself.
Of course, you have the possibility to implement some scenarios for high availability and disaster recovery. So, for me, hyperconvergence is less expensive, assures high performance and availability, is easy to scale up and manage. Hyperconverged infrastructure is now here and is the new way of designing infrastructure. If you need more information regarding hyper-convergence or Nutanix, do not hesitate to contact me.
About the author