3 Reasons Why Networking Breaks HCI Scalability

One of the most appealing aspects of hyperconverged Infrastructures is their ability to scale. When IT needs more compute or storage resources to keep up with their rapidly growing enterprise, they are told to “just add a node”. There is of course more to “just adding a node” than just adding a node. Networking that node into the cluster can create a lot of cost and complexity that breaks the HCI scalability pitch.

#1 Network Provisioning

When adding a node, actually after racking that node, it is necessary to connect it to the network. In most cases, it requires making several connections. Most IT planners will have at a minimum, a VLAN dedicated to storage I/O traffic, a VLAN dedicated to east-west cluster traffic, a connection to the corporate network, and a connection dedicated to cluster management. All of these connections need ports from the top of rack switch, provisioned and configured for use. In many cases, the team in charge of the HCI environment does not have the necessary credentials and skills to perform the needed network provisioning. Many data centers report that it takes days to provision networking to the node after its physical installation in the rack.

#2 Network Performance Insights

HCI is a shared everything environment which make troubleshooting more difficult. Isolating which VM is consuming what network resources is very difficult since the network hardware and network management is a separate silo from the management of the HCI environment. IT is typically caught off guard when a performance problem occurs and has limited tools to isolate networking problems down to a virtual machine level.

#3 Networking Costs

One of the advantages of the “just add a node” concept is the makeup of the node itself. The node is a server, typically made from commodity hardware. The storage for the environment is in that node, using server class drives, that are less expensive than the enterprise class drives found in shared SAN/NAS storage systems. As a result, IT can add additional compute, storage performance and storage capacity, to the cluster at very competitive price points.

The problem is that the cost of networking is the same whether the organization uses HCI or traditional SAN/NAS. In fact, to some extent HCI can be more expensive because it typically consumes more ports, which leads to more top of rack switches and eventually more core directors. The overwhelming majority of organizations source these switches from just a handful of vendors that charge a premium for their hardware. The organization has limited options to drive down the cost of networking because they are essentially locked into their network hardware provider.

Can SDN Fix HCI Scalability?

One of the primary benefits of Software Defined Networking (SDN) is that it abstracts the network operating system from the network hardware so it can run across a variety of vendor’s hardware. This abstraction allows organizations to buy commodity switches instead of premium priced hardware. Additionally, SDN brings a scale-out like design to modular switches. Instead of having to keep upgrading to the next larger switch, IT can just keep adding modular switches but manage them as if they were a single entity. Modular switches are generally far less expensive than core, director class switches.

SDN also brings automation and insight. It enables automatic port provisioning when making connections. For example, SDN allows adding an HCI node to the cluster and after completing the physical connections, automatically, through policy, provisioning and configuring those connections. Lastly, SDN provides deep inspection of port and switch traffic. Enabling detailed insight even in virtual environments like HCI.

StorageSwiss Take

HCI is ideal for organizations looking to simplify and speed its ability to scale. But, it must deal with the networking issues to accomplish that objective. Since dedicating network personnel to be on-call as HCI nodes are added is financially unrealistic, using SDN is the only viable option.

To learn more about how SDN can help scale HCI join us for our on demand webinar “How Software Defined Networking Can Simplify Scale-Out HCI”.

Watch On Demand

About George Crump

George Crump is the Chief Marketing Officer at VergeIO, the leader in Ultraconverged Infrastructure. Prior to VergeIO he was Chief Product Strategist at StorONE. Before assuming roles with innovative technology vendors, George spent almost 14 years as the founder and lead analyst at Storage Switzerland. In his spare time, he continues to write blogs on Storage Switzerland to educate IT professionals on all aspects of data center storage. He is the primary contributor to Storage Switzerland and is a heavily sought-after public speaker. With over 30 years of experience designing storage solutions for data centers across the US, he has seen the birth of such technologies as RAID, NAS, SAN, Virtualization, Cloud, and Enterprise Flash. Before founding Storage Switzerland, he was CTO at one of the nation's largest storage integrators, where he was in charge of technology testing, integration, and product selection.