Over the last decade, hyper converged Infrastructure (HCI) systems, such as Dell’s VxRail, have become established proven technologies.

The 2018 IDC report figures suggest Dells VxRail provides 56% Faster IT service delivery, 489% Five-year ROI and 60% More efficient IT teams.

However, in a crowded marketplace, filled with a growing mix of technologies that are fighting for IT departments’ attention for on-premise, hybrid and cloud solutions, procurement decisions to refresh, or build-on, current technology stacks are rapidly becoming is vastly extended.

No longer is there one easily identifiable, fit-for purpose solution. Boards and IT Departments are working though huge reams of technical information, webinars, demos and proof of concepts (POC) to identify what would be best solution for the business – often over 12 months ahead of time.

As with any purchase, there are various stakeholders’ voices also competing for airtime. The CFO requires the right cost and cost model, the CTO wants the optimal tech, the CIO wants the best security with lowest risk, the IT manager wants the best technology they understand, and End Users want the best experience for their day-to-day needs.

With all of this complexity, in an ever-changing business environment, when does the adoption of Hyper-Converged Infrastructure make business sense?

Defining factors for total cost of ownership (TCO) of Hyper-converged infrastructure (HCI)

When are you implementing HCI?:

  • HCI usually replaces traditional hosts, networking and shared storage solutions – often referred to as a Converged Infrastructure or three-tier architecture (storage, networking, compute). Therefore, the ideal solution is to implement HCI when the current infrastructure is all at the end of their useable lifecycles, or when implementing HCI as an additional infrastructure for a requirement your current system can’t deliver. 
  • However, if your current infrastructure is not due for a refresh at the same time, the first consideration is, does HCI replace all technology which is no longer required, can it deliver tangible value to the business and does it still have a leasing overhead, warranty cost, locked in support contract or similar.
  • If they are all at end of life, or if it is for a new project, you start on a like-for-like TCO basis. If not, there is a cost to consider for what you are leaving behind and their associated costs.

Support costs – a single point of support for all software and hardware?

  • Organisations usually have hardware support through one or more providers, covering the multiple technologies HCI could replace.  Each one of these will have their own cost and will be backed by skilled support engineers. This could reduce as you now only have one technology and less hardware to support.

Warranty costs – single vendor, single technology, single warranty?

  • Quite simply, you could be replacing multiple warranties all with different SLA’s – for example, 24/7/365 4-hour to 9-5 NBD_ to a single consolidated warranty included as part of the initial purchase. This will reduce time receiving multiple quotes, processing different renewals and managing the support when required.

Timing-based factors – can you really free up time?

Many of the costs of Total Cost of Ownership are created by the potentially free time, which can be utilised elsewhere – but why and what does it really mean?

Time for management:

As HCI reduces your ‘stack’, all with their own management platforms, upgrade requirements and unique characteristics to a single infrastructure with a single platform. This results in less passwords, less multi system knowledge and one single platform to use in order to optimize infrastructure operations and use actionable insights and analytics.

Edge Deployments to local branch offices can be centrally-managed under the same platform. This could reduce need for on-site IT personnel with more direct control from the central team.

Time for upgrades:

VxRail HCI Automates upgrades from one known good state to the next, simplifying operations while reducing risk. This minimises downtime with bad patches or patching multiple hosts and storage running through downtime on each.

Time for implementation

Professional services for implementing a converged infrastructure or three-tier architecture (storage, networking, compute), each with their own unique install approach, migration process and timing, could cost more than a single VxRail HCI device.

Space and power – ‘to infinity and a bill?’

Space is at a premium in the data centre, especially when you may pay for rack space through a provider. A migration to HCI may reduce your datacentre footprint. Ultimately, this costs less in space and power. This is, of course, subject to what HCI is replacing or being compared to, how much space it takes and how much power it needs.

Cheaper graphics with better performance?

Dell VxRail can be a good fit for Virtual Desktop infrastructure (VDI). Under none VDI models, you could be paying to roll out high end graphics cards to each end user device to run AutoCad or similar.  A product like Dell VxRail for VDI can provide high performance scaling and virtual graphics and compute, with 2x graphics acceleration performance and 50% more users per board, centralising GPU cost which may be cheaper and have betterer performance.

Are all of the Total Cost of Ownerships positive?

Does it really replace all of the tech?

Yes, HCI could replace all of your shared storage and servers. However, for many organisations, the existing stack won’t just disappear. They may also choose to keep their hosts and storage as a backup targets or as a disaster recovery solution. Of course, some of these are additional value to what your business already has, but they may not remove the devices or their associated cost base.

Licencing:

Don’t always believe your licencing costs will go down just because you are consolidating equipment.

With traditional converged infrastructure, you may have previously operated with two or three hosts running your virtualised environment. However, with HCI, you may actually need additional hyper converged nodes and higher capacity processors as they are now doing more to give you the required performance.

Performance upgrades (RAM and Processing power):

Getting the right size device early on is critical, ensuring it delivers the performance and enables recommended growth overhead based on planned projects.

Ram and processors usually mirror across all nodes in the cluster to manage your virtual machines (VM’s) and storage. What previously may have RAM upgrade for one host, which is was delivering a RAM intensive VM, quickly becomes an across the board RAM upgrade on all nodes.

Alternatively, you can add another node to upgrade which will load balance, which will incur cost and licencing etc.

In summary, it really depends on each individual business case as to if HCI offers better total cost of ownership. Things to ask yourself are:

  • Are you ready to refresh all EOL kit or taking on a new project?
  • Are you paying for a big stack of technology?
  • Are you burning time on management?
  • Are you paying for staff to manage at branch offices?
  • Are you paying multiple support vendors?
  • Are you losing time dealing with warranty renewals?
  • Is your current stack delivering poor performance?

For an open and trusted conversation about whether or not VxRail HCI is right for you, then get in touch today

Reference sources: