vege - Fotolia
A hyper-converged system has a modular structure based around the use of nodes, each containing its own compute, storage and network resources. While this modular approach can be handy, it has also been criticized because it can result in wasted resources.
Imagine an organization is using a hyper-converged system and needs to add compute resources. Because scalability occurs at the node level, the only way to add extra compute resources is to purchase an additional node that also contains storage and network resources. In this case, the organization is paying for storage, network resources and software licenses it does not need.
This raises the question of whether there are ever advantages to node-level scalability, which makes it impossible to add individual resources without adding an entire node. The answer is that you can reap the benefits of hyper-convergence in these situations, but it all comes down to the use case.
Hyper-converged infrastructures have a number of different use cases, but the one that seems to get the most attention is virtual desktop infrastructure (VDI). One of the benefits of hyper-convergence that makes it well-suited to VDI is the way in which virtual desktops consume resources. Each virtual desktop requires CPU and network resources, as well as storage capacity and storage IOPS.
The nice thing about hyper-converged systems is that each node contains its own disks. The IOPS produced by the virtual desktops can be spread across the nodes, with each node handling the storage IOPS for the virtual desktops running on that node. This not only eliminates the need for an expensive, high-performance storage appliance, it means that scalability is completely predictable because each node will be able to handle a specific number of virtual desktops.
VDI is not the only case in which the use of modular, hyper-converged systems makes sense. The benefits of hyper-convergence often outweigh the drawbacks for organizations with limited IT staff resources because the modularity simplifies the deployment process and eliminates many (but not all) of the capacity planning tasks required to deploy more traditional systems.
Hyper-convergence brings scalability to the data center
Where do hyper-converged workloads work best?
Eight questions to ask hyper-converged vendors
Dig Deeper on Hyper-Converged Infrastructure Management
Related Q&A from Brien Posey
Organizational resilience encompasses everything a company needs to run in times of crisis. These examples show how businesses handle tough ... Continue Reading
There are several different methods to cloud storage encryption. These best practices for encryption can help improve security of important cloud ... Continue Reading
While only a small number of hardware vendors offer DPUs, the technology has significant implications for IT storage systems -- and the admins who ... Continue Reading