BACKGROUND IMAGE: iSTOCK/GETTY IMAGES
A hyper-converged data center provides many benefits over traditional setups servers, storage arrays and networks running between them.
Businesses depend on IT innovation to meet their strategic objectives. This can lead to a steady stream new demands that stretch the limitations existing data center resources, and the challenges become even more apparent when we consider the dynamic nature demands on data center managers.
While hyper-convergence, the packaging compute and storage into networked nodes with virtualization and management software natively installed, is still far from ubiquitous, products are quickly maturing to meet the scale and manageability enterprise administrators demand.
There are a lot of different reasons why IT organizations choose a hyper-converged data center setup. Depending on your situation, one or more of these reasons may apply to you.
1. The in vogue technology
One of the top reasons IT organizations look at hyper-converged data center setups is because other customers are looking at them. Hyper-convergence is a very fashionable phrase that is getting attention. This isn't in itself a reason to check it out, but it is a sign that you should look at the rest of this list to see whether hyper-convergence suits your enterprise needs.
2. Scale-out workloads
Hyper-convergence is a scale-out platform, so it suits scale-out workloads. Virtual desktop infrastructure (VDI) is the classic scale-out workload, with hundreds or thousands of desktop VMs. Each virtual machine is a fairly small load, but the total VDI load is large. Web server farms and some database workloads can also benefit from a hyper-converged data center. The scale-out platform delivers consistent performance for applications designed to grow in this manner.
3. Growing workloads
Hyper-convergence makes it easy to expand the capacity of your data center by adding new nodes over time. If you have a phased rollout or simply experience long-term IT resource demand growth, hyper-convergence allows you to purchase the needed capacity as the workload grows, aligning infrastructure spending with the value that the workload delivers. It is easy to increase node purchases if growth occurs faster than planned and decrease them if growth lags behind your forecast.
A significant feature of a hyper-converged data center is simplified management. Most hyper-converged platforms include policy-based virtualization management, which reduces the effort required to manage a collection of VMs. This simplification allows IT teams to spend less time adjusting and optimizing the platform and focus more on the VMs and applications that deliver value to the business.
5. Eliminate technology silos
A hyper-converged data center eliminates the storage silo and its dedicated network. This relieves businesses from the need to maintain expertise in the storage array and network. In addition, hyper-convergence usually removes the need to manage the logical unit numbers, presentation and storage paths that come with a storage area network (SAN). This removes the need for dedicated expertise and allows IT teams to focus on other parts of data center and application stacks.
6. Infrastructure resources
Small projects that require modest resources may not suit deployment onto a large converged infrastructure platform or dedicated fleet of servers and storage arrays. Hyper-convergence enables data center staff to deploy a much smaller unit of infrastructure for smaller requirements. Adding a few nodes to your hyper-converged data center setup can serve the needs of a small project without making a large purchase. If a series of projects each requires a few nodes, the IT team can combine the nodes into a single, larger hyper-converged cluster. Combining these smaller projects into a larger infrastructure keeps the operational cost of running the projects from increasing as each project grows.
7. ROBO support and management
Some organizations have a vast number of remote offices or branch offices (ROBO). Many of these locations need enough IT that they will benefit from a robust virtualization platform. It gets pricey deploying SAN and servers to each site, and the complexity of operating these systems makes the cost of supporting ROBO high.
Space is at a premium for some branch offices. Combining compute and storage into one box can make hyper-convergence much more compact than conventional infrastructure. If the branch office is a vehicle, ship, truck, aircraft or even submarine, then the space and power savings may prove critical. If there are many ROBO locations, then managing the platform at these sites becomes a challenge. Hyper-convergence simplifies management, keeping operational costs from spiraling out of control.
8. Hardware refresh
IT organizations often purchase, and replace, entire virtualization platforms together. This is an ideal time to evaluate hyper-converged products for your next infrastructure. If you deploy a hyper-converged data center, it is likely that the future process of refreshing the hardware will be far simpler than replacing virtualization platforms en masse. Hyper-convergence generally allows node-by-node replacement to update a cluster, meaning there are no large outages or long VM migrations to get a new infrastructure in place.
Finding competent, specialized staff is a challenge for any IT department. The simplicity of a hyper-converged data center reduces the variety of skill sets you will need on the team.
Managing individual VMs eats up a lot of time. Using policies to manage groups of VMs and only having to think about noncompliant VMs reduces this workload. The same number of staff can do a better job of managing even more VMs than they currently handle when traditional processes are replaced by good policy-based management.
10. No public cloud
The ultimate simplification of IT is to not run IT, which is what makes cloud services so attractive. However, many organizations have legal or compliance requirements, or simply business rules, that prevent them from using cloud services. A hyper-converged data center allows for reduced infrastructure management of cloud services while keeping all IT on premises. Hyper-converged vendors design the platforms for programmability, making it easier to build private or hybrid cloud platforms on top.
How to set up hyper-converged systems alongside existing servers
The differences between converged- and hyper-converged infrastructures
Hyper-converged storage starts to offer greater choice