How to Avoid 3 Common Hyperconvergence Pitfalls
When weighing the pros and cons of Hyperconverged Infrastructure (HCI), the scale is usually tipped in favor of going hyperconverged. The idea of standing your ground with traditional or even converged infrastructure is getting harder to sell to both the management and technical types.
Just because it’s often the best solution, however, doesn’t mean you don’t need to understand its potential issues and downsides. To that end, here are a few common pitfalls you need to be aware of before making the decision to get on board the HCI bus. This article will point out three of the more common ones to take into account.
1. Underestimating Capacity Needs
Capacity planning is crucial to architecting an HCI deployment that will provide you with what you need currently, while at the same time providing a buffer for growth. HCI vendors will sell you on the fact that you can scale on demand by adding a new block or unit to your current rack configuration.
But remember that while HCI allows you to scale on demand, you can’t scale by individual component need: the new block or unit will contain compute, storage and memory. This adds to the capacity as a whole, not individually. Underestimating capacity needs mean you will need to purchase a new block or unit when it’s discovered that capacity is near 100% in your infrastructure, so your bottom-line ROI will suffer.
It’s great news that you can simply scale out by adding a new node, but take into account the procurement process in your organization. Does it take one day to request, procure, deliver and rack/cable a new block? Typically, the answer to this is “No Way!” It’s far more realistic to think the procurement process will take closer to several weeks, if not months. If you’ve underprovisioned, you may well be out of capacity and facing service degradation for your customers.
Avoid this pitfall by planning for needed capacity plus a two-year growth buffer. Know your procurement process. How long does it take from request to deployment? This will create a situation where you can plan for scale in the event you need to.
2. HCI Sticker Shock
There can be significant sticker shock when pricing and procuring a new HCI deployment. That’s not uncommon; but keep in mind that the costs over time will usually end up in your organization’s favor.
If you let the upfront CAPEX (capital expenditures) scare you off of HCI, you’ll be missing out on the cost savings benefits that HCI actually brings to your total cost of ownership (TCO). A good analogy for this sticker shock is implementing a virtual desktop infrastructure (VDI) solution. The initial outlay for VDI is really high, and may seem like a money sink. Over time, though, the cost of not spending the money upfront means having to do hardware refreshes on desktop machines and laptops. That will likely end up costing you more – maybe a lot more – in the long run.
The same idea applies to HCI. The initial cost is outweighed by the overall savings benefit and TCO big picture. HCI bundles compute, storage and memory in one chassis, while traditional infrastructure is fragmented by separating each component into its own silo. Traditional infrastructure doesn’t allow for scale on demand by simply adding a new block or appliance. Traditional infrastructure generally requires a “forklift upgrade” to add capacity. The TCO of HCI is much lower overall than that of a traditional infrastructure, if you look at it from a five-year perspective.
3. Fear of Change
While not directly related to HCI, this pitfall will keep you from realizing all the benefits and potential HCI can bring to your organization. Fear of change is a real thing, and very prevalent in many organizations around the world. This is especially true when it comes to embracing new technology that challenges the way things have been done for decades. It’s very easy to get stuck in the rut of feeling comfortable with the way things have worked in your IT department, but this frame of mind can end up setting you back.
HCI brings a robust set of features to organizations looking to lower TCO, make support and troubleshooting easier, and allow for on-demand scale of resources. HCI accomplishes those objectives by bundling all components into a single chassis.
Consider, for example, that instead of having three different vendors pointing the finger at each other when troubleshooting an issue, you’ll only be dealing with a single vendor. The logs used for that troubleshooting come from the appliance as a whole, and will cover all possible issues as opposed to collecting logs from each piece of hardware involved.
Additionally, the adoption of HCI effectively kills the “siloed teams” approach to IT management. Each administrator will be adept at all aspects of the infrastructure, and it will all be managed from a single pane of glass. If you’re afraid of change, though, and stay with your traditional data center model, you’ll miss all these benefits.