This post is a follow-up to the one about picking the right storage solution (DAS, NAS, SAN: Which is the right one for you?), so if you haven’t read that yet, make sure to do so before continuing, so you’ll have a good understanding of how both decisions impact the other.

Table of Contents

  1. Traditional infrastructure
  2. Converged infrastructure
  3. Hyper-converged infrastructure
  4. Conclusion

In this post, I’ll cover the types of infrastructures commonly used by organizations and the primary differences between them.

Protect Your Data with BDRSuite

Cost-Effective Backup Solution for VMs, Servers, Endpoints, Cloud VMs & SaaS applications. Supports On-Premise, Remote, Hybrid and Cloud Backup, including Disaster Recovery, Ransomware Defense & more!

The types of infrastructure I’ll be covering are:

  • Traditional infrastructure
  • Converged infrastructure (CI)
  • Hyper-converged infrastructure (commonly referred to as HCI)

Traditional infrastructure

With traditional infrastructure, you purchase components such as storage arrays, servers, and network switches and put them in a rack. The benefit this brings is that you can handpick the right components for you, for example, in terms of performance or from a cost perspective. This also means you can select everything from the same vendor or combine different ones to your liking.
Another advantage is that you can also upgrade individual components. Need more compute power? Just replace the servers with newer and faster models.

Download Banner

Traditional vs Converged vs Hyper-converged infrastructure

This benefit also has a downside, as you need to know what you are doing. Suppose you select a component that isn’t supported by the other. You might not have a functional infrastructure or one that frequently crashes.

Another downside is that all these components must be set up and configured individually, as there is no central management plane to manage them.

And remember that first downside? Now combine that with the second when you want to update the platform. You’ll have to check all the interoperability once more for each and every following update.

Converged infrastructure

With converged infrastructure, things are getting more interesting. For example, some of those separate components now come in a single box, referred to as an appliance. This makes designing and deploying your infrastructure easier, as you have fewer moving parts to worry about and have to be physically put in a rack.

Traditional vs Converged vs Hyper-converged infrastructure

The good thing is that you no longer have to worry about compatibility. As the appliance you purchased is typically validated upfront. And the same goes for managing the lifecycle. The vendor takes care of checking interoperability before releasing a new version.

Now, this might sound like an ideal world, but there are certain things to consider. First, these individual components still exist, so they’ll still have to be managed individually, albeit from a single management platform. Another thing that sometimes worries organizations is something called vendor lock-in. By using equipment from a single vendor, your organization might be at risk when the vendor decides to no longer support your product or is acquired by a different company.

And if you need to scale out? You increase the number of appliances. Something seen in the early days is that when you needed more storage, you were “forced” to increase the compute resources as both came in the same appliance. In the present day, it’s pretty common to scale with compute or storage only nodes, which, as the name implies, only contribute compute or storage resources to their brothers and sisters.

Though converged infrastructure isn’t as popular as HCI, it’s still commonly available in solutions such as FlexPod (which is a joint effort from NetApp and Cisco), Dell EMC VxBLock, and HPE ConvergedSystem.

Hyper-converged infrastructure

Now, if you liked what you’ve read so far, this might be the best section yet. Hyper-converged infrastructure, or HCI as we’ll call it, takes the appliance from the converged infrastructure but embeds everything in software, or software-defined as we call it.

In a software-defined infrastructure, everything is controlled through software. This also opens the door for new automation capabilities (for managing the platform and its workloads)

and simplified (and centralized) management capabilities. The same goes for scaling the infrastructure, as the software layer takes care of it for you.

Traditional vs Converged vs Hyper-converged infrastructure

HCI is often considered a step toward becoming your own cloud provider, as consuming resources becomes a cloud-like experience.

The combination of these factors allows IT departments to focus on supporting the business and the workloads required instead of maintaining the underlying infrastructure.

Does this mean there aren’t any downsides to HCI? Well, that depends on your use case. HCI appliances are typically purposely built, so you might have limited options if you require GPUs for VDI workloads. In addition, though scaling an (H)CI platform is easy, it is also limited to a certain degree by the same software. This is done to guarantee stability and performance, but something typically seen at all vendors is that these maximums are raised with every new software release.

With a traditional infrastructure, you could offer storage to a particular physical host, for example, for archiving purposes. In the early days, storage from an HCI platform was only limited to that platform. Now, multiple HCI vendors allow storage to be shared outside of the HCI platform.

As I mentioned in the converged section, HCI is booming. This also means that there are many different options to choose from. The most well-known and used ones are (in random order):

  • Nutanix Cloud Infrastructure
  • VMware vSAN
  • Dell EMC vXRail
  • HPE SimpliVity
  • Cisco HyperFlex

Which one to pick depends on the requirements that you have as an organization.

Conclusion

It is impossible to provide you with all the right tools to select an infrastructure that suits your needs. It requires thorough knowledge of the different options, but more importantly, the organization itself.

As with any solution, both need to align to gain the most benefit.

One thing is for sure, the rise of HCI won’t end soon as hybrid cloud scenarios are becoming more common, and extending your on-premises infrastructure is a valuable asset.

Related Posts:
Hyperconverged Infrastructure Pros, Cons, Benefits, and Drawbacks

Hyper-V Storage Spaces Direct Disaggregated vs Hyperconverged Deployments

Follow our Twitter and Facebook feeds for new releases, updates, insightful posts and more.

5/5 - (1 vote)