Hyperconverged Infrastructure

Hyperconverged Infrastructure Checklist: 5 Must-Ask Questions

Every pilot knows about the pre-flight checklist: the items that must be checked and meet a standard of satisfaction before the pilot turns over the engine. Without it, the pilot could be in a world of hurt; if they haven’t checked the fuel for water, the engine could fail. If the control surfaces (ailerons, rudder, etc.) aren’t freely moving, controlling the plane could become difficult or impossible.

Those checklists are there to ensure the best possible flight with the fewest chances that something will go wrong.

What does this have to do with hyperconverged infrastructure (HCI)? Well, having a checklist ready for vendors will help you find the right one, since you’ll know what you’re looking for at the start of the process. That will lead to fewer surprises and fewer potential disasters as you implement your solution.

Ready for takeoff? Then read on for these items that should be on your HCI vendor checklist; these are some of the most important questions to ask a supplier as the first step toward finding one that meets your needs and fits your budget.

1. Is your focus on the enterprise or on small/medium businesses (SMB)?

There is a huge range of HCI equipment and price points available, for every kind of business. Some vendors are set up to sell and service huge environments with thousands or tens of thousands (or more) end-users and applications. Other vendors are more focused on the SMB space, and their solutions are a perfect fit for those settings.

Make sure you ask questions about the total cost of a node, including hardware and software. Also find out how many nodes a vendor recommends for your business, balancing performance and high availability with those costs.

2. How simple is setup and management?

One of the supposed advantages of HCI is the ease of deploying nodes. But it isn’t always so easy. You should ask how long it takes, on average, to go from zero to a functioning environment that is managing workloads. Some environments can be up and running in an hour or less; others will take longer.

Also ask how long it typically takes to add a node. Is it a matter of minutes to plug in and integrate a new node, or hours or days?

In terms of management, you should find out if there’s a “single pane of glass” to manage the systems. If not, the time and manpower it will take to administer the environment could increase exponentially.

3. How well do your products scale?

Do you anticipate needing to add more nodes quickly, or are your future needs more modest? Also, when that scale-up or scale-out happens, will management needs increase as well?

Note that with some HCI solutions, you can’t simply add a node with more storage if that’s what you need; instead, you have to buy a node that also increases your compute power, which you may not want. In effect, you’re paying for compute you won’t use. Does the vendor offer the ability to scale up only what’s necessary? Can you mix and match nodes?

4. How robust are your high availability and fault tolerance features?

You need to know how well the data in an HCI node is protected. For instance, is data striped and mirrored across all drives in the cluster? If the answer is yes, that eliminates any single points of failure.

Is there automatic failover for crashed nodes? If it requires hands-on administration to move virtual machines from one node to another, problems loom in your future.

5. How experienced is your company in creating HCI products?

You need to have confidence in the company you choose. Start by making sure that the company has been building and testing its HCI systems for more than a month. Then go deeper: have they been in this specific business awhile? Do they have recognition within the industry (awards, solid reviews, etc.)? Do customers think highly of their service?

Using this checklist will help your HCI implementation or transition feel more like smooth sailing in clear skies than flying into stormy weather.