Hyperconverged technology for server and storage networking is not brand-spanking new. The technology has been out on the market for about five years now and there are many vendors offering their version of it.
What is noteworthy is a relatively new player in the game who is turning heads with a new approach. Cisco HyperFlex is quickly gaining market share, recently surpassing some of the first movers in this new technology class.
What is hyperconverged technology?
To fully understand what makes this product different from the rest of the pack, it is first important to understand what hyperconverged technology is. For non-admins or those otherwise unfamiliar with the concept, hyperconverged refers to a prepackaged platform that handles server and storage needs.
Back in the ‘olden days’ (i.e. maybe seven years ago) the traditional infrastructure consisted of servers and numerous storage arrays cobbled together with multiple operating environments/interfaces. Then along came ‘converged infrastructure’ which took traditional infrastructure and pre-configured it into a rack at the factory, so you simply had to turn it on and you were done. It offered rapid deployment with a pre-designed, pre-architected solution.
“By 2022, more than 50% of enterprise-generated data will be created and processed outside [of] the core data center or cloud.” – Gartner
Hyperconverged infrastructure takes it to the next step so that instead of a 3-tier design, comprised of servers, networking and storage, the entire environment is collapsed to just servers, or HCI (Hyper-Converged Infrastructure) Nodes. These servers typically have significant internal storage, and software-defined storage is utilized to present shared capacity to the virtual environment. This is part of a greater movement towards virtualization of the entire data centre, where applications are decoupled from the constraints of physical hardware. The end result is an environment that is more flexible, maneuverable and scalable than traditional 3 tier converged deployments
Cisco’s HyperFlex is taking hyperconverged technology and putting a new spin on it. Here are just three reasons why network admins are taking notice.
Reason # 1 – Single support contract
One of the key differentiators of Cisco HyperFlex is that with the addition of Cisco networking equipment, Cisco is able to bring the entire HCI stack into a single-support contract, serviced through a single support team and contract to manage.
Most of the competitive offerings on the market have separate support contracts for the server hardware, hyperconverged software, and networking. These competing offerings will attempt to make the transition between support contracts as painless as possible but the risk of the dreaded finger-pointing between vendors remains.
Reason # 2 – Ability to attach to external storage
One of the challenges that many organizations have with hyperconverged solutions is that typically you add hyperconverged nodes to the HCI cluster to add additional storage. While this “building block” or scaling approach can appear attractive, the reality is that for many clients, storage requirements grow faster than their compute requirements. Resulting in the need to add less necessary compute, RAM, and software (HCI software and Hypervisor) to add the required additional storage to an environment.
With Cisco, you have the option of attaching a secondary storage array, block based (iSCSI or fiber channel) or NAS, and it’s all supported within the HyperFlex platform. Some of the other HCI vendors are addressing storage scaling by allowing for storage heavy nodes or storage only nodes. Cisco’s allowance and support of external storage is a huge market differentiator and a real game-changer. This helps eliminate one of the major obstacles that many organizations struggle with when considering HCI infrastructure.
Reason # 3 – Ease of Operations
A common assumption is that any system you put in place simply requires a flip of the ‘easy switch’ when it comes not only to installation but also to regular upgrades. Quite often, this is not the case. With many of the hyperconverged systems in the market, updates can be a bit of a nightmare to implement. The multi-stage process requires first upgrading your servers, then your HCI software, and finally your hypervisor – so it’s neither simple nor quick. I’ve seen this become a disaster when people don’t realize that not all versions of software play well with all versions of hypervisors.
However, Cisco HyperFlex, really is easy to operate and upgrade. With unified control from the network down, upgrades can be completed through a downloadable module that includes everything required. It’s very clean and that is an important factor for many organizations.