image_pdfimage_print

The greatest existential threat is the lack of adaptation. Once upon a time, giant apes roamed Southeast Asia. Believed to have stood 10 feet tall and weighed around 1,000 pounds, Gigantopithecus is gone now. Scientists suspect that as their habitat changed, they didn’t adapt to their new environment.

Today, we still see the proverbial 800-pound gorillas— a legacy storage vendor. These companies have grown to massive size by servicing the historic on-premises storage market via capital purchases. But… that environment is changing.

Businesses want to move workloads to the cloud, and rightly so. The economic benefits, operational improvements, and simplified overall experience are very attractive. To put it another way, the habitat in which a storage vendor lives is changing. Customers are now demanding on-premises models that look more like the new world of cloud, not the old world of multiyear capital purchasing, operational overhead, and the responsibility of delivering SLAs to stakeholders. Savvy CIOs want to get away from these legacy issues, along with the associated vendor lock-in, risk forecasting, and unpredictable costs.

The gorillas are trying to adapt to this new environment. One way legacy storage vendors have attempted to adapt is by introducing flexible consumption models. In theory, these models mimic cloud resources with pay-as-you-go models, pricing tiers, and some level of management. When considering these offerings, you need to ask the following question:

 Are these legacy vendors adapting to the cloud environment, or are they trying to force you into infrastructure that looks like a flexible consumption model on the surface—but in reality retains the same issues of long-term commitments, unpredictable costs, and disruptive upgrades?

For example, Hewlett Packard Enterprise (HPE) has introduced GreenLake Flex Capacity. Billed as pay-per-use storage, it looks like a cloud-style consumption model from the outside. I read the documentation¹ and I was surprised by what I discovered. Let’s say, for example, you use around 50TB of storage a month. As I understand it, you can enter a contract and commit to pay for that 50TB every month for the duration of the term. On day one, there’s a 50TB Minimum Contractual Capacity (or minimum monthly charge). For what HPE calls a “Basic Deal” in its documentation, if you use 55TB the following month, your minimum monthly charge for the duration of the contract will now be 55TB. “Pay-per-use” appears to become “pay-monthly-for-that-one-time-you-exceeded-your-commitment.” In the documentation explainer video, you can clearly see a customer using less than 40TB during the month but being charged for 80TB of usage when their committed capacity was only 60TB. Search the doc for “shrink rules” and you’ll see what I’m talking about.

Does that sound like a true cloud pay-as-you-go financial model, or more like a storage vendor trying to lock in as much capacity for as long as possible—whether you need it or not? This overprovisioning is a hallmark of legacy capital storage purchasing. From our perspective, it isn’t adaptation; it’s another way to make sure overprovisioned assets drain your wallet every month.

And that’s just one of the big gorillas. When digging into the specifics of the “consumption models” offered by other players like Dell, Hitachi, and NetApp, you’ll see some strong similarities to a typical legacy vendor selling behavior. Keep an eye out for things like:

  • Long-term contracts
  • High expectations for usage commitments
  • Traditional issues around disruptive upgrades instead of true elasticity

Is that adaptation, or more like a capital lease? Forcing you to commit to large amounts of “pay-as-you-go” storage up-front and then billing those costs over a long period of time with the potential for rebuys offers little benefit over an operationalized lease. But, at least with a lease, your payments are fixed and not subject to ongoing increases based on a couple of months of higher than expected usage.

When evaluating these services, ask all of the usual questions and then ask some more:

  • What’s the experience like if you need more capacity?
  • Do you tell the vendor you need additional capacity? Or do they tell you?
  • What’s the process to expand the gear in your data center as needs change?
  • How does expansion or additional usage change the contractual costs?
  • What downtime is expected for upgrades or expansion?
  • Are they willing to put “zero-planned downtime during upgrades and expansions” in an SLA for you as a contract term?
  • If your needs change, do you have options to reduce your committed spend?
  • Is the service linked to an identified asset? Will your accounting department be able to recognize the spend as Opex?
  • Hypothetically, if you had this service on January 1, how would your bills and rate structure have changed as a result of an unexpected increase in remote workers?
  • If you sign a contract today, what would your experience be like if you suddenly exceed your committed capacity, or are trending to exceed available capacity in the next 90 days? Are the expansions disruptive? What cost is attached? If hardware needs to be upgraded, are those costs included in the service or additional charges? Is a new contract required?

You’ll likely find that these “services” are delivered in a very similar way to a lease, and that shouldn’t be a surprise. Legacy vendors have spent years building out financing relationships and capital lease organizations to help you pay for expensive hardware over longer periods of time. For some, adaptation is trying to pretend the world isn’t changing and continuing with business as usual—the usual overprovisioning, usual overcommitments, and usual storage vendor and revenue lock-in, with the new buzzword of “consumption” layered in.

Back to Gigantopithecus. They’ve disappeared, but a distant relative of theirs is still among us and shares more similarities to modern humans than other primates —the orange orangutan. Orangutans are a great example of adaptation, and they’re not the only orange-hued thing that has adapted to a new environment. Pure Storage has taken adaptation to the next evolutionary level.

Two years ago, Pure Storage® launched Evergreen//One™ to provide a flexible way for enterprises to consume storage. A usage-based model should be simple, and similar to the way cloud infrastructure is consumed—with options for service tiers, short-term contracts, and the ability to start small and grow over time.

Pure remains focused on adapting to the changing market, and as a result, we’ve lowered our minimum commitments—even on 12-month contracts. Signing a usage contract shouldn’t feel like signing a lease, so with Pure, you won’t have to worry about any of the overcommitments or long minimum contracts you’re likely to see with competitive models.

With Pure, you get not only a hassle-free experience but also the highest quality of service. Built on Evergreen™ Architecture, our non-disruptive operational model allows us to grow and expand your Evergreen//One environment as you need it—without large commitments. Pure1® helps us track and predict your usage so you don’t need to worry about growth. Evergreen//One maintains additional headroom so there’s one less thing for you to worry about.

We love our products, our customers love our products as proven by our Net Promoter Scores, and we hope you will, too. Our goal is to earn your business with our service every single day, and punishing someone for loving their storage is not very… Pure.

Read the IDC reportDigital Infrastructure Resiliency Drives Interest in New As-a-Service Cloud Operating Models, to learn more about STaaS and why enterprise organizations are prioritizing digital infrastructure resiliency as a foundational element of their IT strategy. 


  1. https://docs.consumption.support.hpe.com/CCS/030Viewing_data/Monthly_charges#Understanding_Minimum_Contractual_Capacity