AI-driven storage soars on the pay-as-you-go current
It’s hard to remember now, but there was a time not so long ago when storage was perceived as a somewhat stodgy, innovation-challenged zone of the data center. Not anymore. In the past decade, storage has been rocked by one breakthrough after another, from cloud storage services to enterprise all-flash arrays. Businesses wanted storage with tighter connections to networking and compute, and vendors responded with hyperconverged infrastructure. Manufacturers found they were running out of space on the memory chip, so they added more layers vertically, giving us 3D NAND memory.
The topology of the storage market morphed once again with the drive to hybrid cloud infrastructure and the AI wave of the past few years. Such trends propelled the rise of intelligent storage: combinations of storage hardware and services that leverage remote data collection and artificial intelligence to actively manage their environment, whether on premises or in the cloud.
Now intelligent storage is gaining even more impetus from yet another unstoppable trend: consumption-based IT.
What intelligence looks like
More than just a repository for information, intelligent storage proactively provides insights to help businesses ensure that data is available when and where it’s needed. For example, it can identify performance bottlenecks across the environment and detect changes such as security threats or seasonal spikes in usage and issue alerts so that the organization can take timely action.
Intelligent storage is:
- AI-driven. The telemetry monitors data from APIs and management interfaces across the stack, as well as from the storage devices themselves. Advanced analytics and machine learning enable the system to model what’s normal in the environment so it can detect and resolve what’s abnormal. Intelligent storage can apply predictive analytics to resolve many potential issues before admins even know they exist. For other issues, an analysis and recommendation engine provides automated guidance, detailing the problem and suggesting actions to resolve it, even if the source of the problem is outside of the storage device.
- Built for cloud. With native cloud integration, intelligent storage simplifies multicloud management, orchestration, and automation, connecting seamlessly between public and private clouds. You can move any workload from production databases, secondary data, or backup data to the cloud, or repatriate data from public cloud as needed. Open, cloud-native APIs enable integration with products from a range of cloud automation and DevOps ecosystem players, helping businesses develop new services quickly. The advisory expertise and cloud management tools baked into intelligent storage help companies avoid the hidden costs and overspend often associated with public cloud initiatives.
- Delivered as a service. You choose the storage that best fits your business needs—whether it’s flash or hybrid—for block, file, or object storage. The hardware is installed at your site, but ownership remains with the provider. You get a local buffer of capacity based on your forecasted demands, plus the services needed to support, manage, and (optionally) operate it. If you need additional capacity, you can add it on demand and scale it down again when it’s no longer needed. You pay only for what you use, and the hardware, software, and services are all rolled into a single monthly bill.
On-prem storage with cloud economics
That third characteristic, delivered as a service, provides massive momentum for intelligent storage in a hybrid cloud world. According to IDC, 60 percent of enterprises will use flexible, lower-cost IT consumption models by 2023. Intelligent storage is a true consumption-based model that offers companies a cloud-like experience for on-prem storage.
If there’s one big lesson businesses took from public cloud it's this: the power of simplicity, with financial simplicity being a big part of that. It’s a myth, of course, that tapping into a cloud economic model is as simple as sliding a credit card. But companies love the ability to bypass big upfront costs in favor of a monthly bill. The cloud opened up a range of new approaches for budgeting and funding IT. It made IT resources more easily measurable by usage.
But then came technologies that could do essentially the same thing for on-premises infrastructure. New metering systems can track, record, and report on how hardware resources are being used on premises. When combined with a comprehensive set of services to deliver, monitor, and support the infrastructure, these technologies provide a true on-demand solution for the physical data center. With the advent of consumption-based IT, suddenly companies could see their on-prem IT estate in a whole new light.
When exploring an on-premises, consumption-based approach to storage, consider the following criteria to determine if it’s right for your environment and business needs.
Data sovereignty and privacy. Compliance requirements continue to escalate, impacting companies’ storage strategies; the EU’s General Data Protection Regulation is just one example. Many companies prefer to place customer information on premises, and whole industries require special treatment of sensitive data—the healthcare sector, with its mandates to protect patient information, is a good example.
Latency. Compute resources are increasingly moving to the edge, the IT zone outside of the data center, close to where the action is. If you’re running heavy-duty calculations to control production on the factory floor or monitoring safety conditions in a chemical plant, you need to be able to act in real time or close to it. You can’t afford the delays involved in moving information back and forth to the cloud.
Location and connectivity. The movement to the edge is also pushing compute resources into remote or harsh environments, such as oil rigs, where connectivity may be unreliable. But whatever your location, a network issue can quickly turn into a customer service issue if crucial customer data has to be retrieved from the cloud. For many companies, that’s an important consideration when designing backup systems, too. If you’re hit with unplanned downtime, you want to get everything up and running again as quickly as possible. You may not want that capability to hinge on having a good connection to the cloud.
Upfront costs. Business leaders may balk at expenditures for on-prem storage under a straight CapEx model. An approach that matches costs to usage over time is often a more attractive option for the finance side of the house.
Capacity management. Storage investments tend to be underutilized; companies typically buy too much capacity upfront to make sure they can cover workloads at peak periods of demand and avoid running into performance challenges. Overprovisioning is wasteful, and it adds unnecessary costs for electricity, cooling, and maintenance.
Agility. Adapting on-prem resources to new business requirements can be a slow process. Procurement cycles are lengthy, and installation of new hardware takes time. Once a company runs out of capacity, it can take months to add more in the data center. That doesn’t do much to impress internal customers. To get what they need, individual teams and lines of business may decide to head for the public cloud and do an end run around IT policies, resulting in the infamous shadow IT.
An on-prem, consumption-based approach may help you reconcile these complex and often conflicting criteria and quickly make the move to a smarter storage setup.
Things you can do with consumption-based intelligent storage
The pay-as-you-go strategy opens up some exciting new possibilities. You can:
Get productive. You can set up new projects quickly and easily, without the usual big, upfront expenditure. The as-a-service approach encourages experimentation and reduces barriers to innovation. That’s a compelling advantage in today’s business environments, where products and business models are changing at a dizzying pace.
Look for vendors that can offer a wide range of intelligent storage technologies to address the needs of your business, whether it’s a new database management system, big data and analytics, or secondary storage. Some providers may pair complementary software, such as data availability or backup solutions, with the hardware, further simplifying the way you deploy and manage your storage needs so you can innovate faster.
Simplify IT budgeting and planning. Consumption-based storage makes the familiar “use it or lose it” budgeting principle obsolete. Instead, your spending is closely aligned with your business objectives. You gain greater visibility and control over how you are actually consuming your intelligent storage resources on a daily, weekly, or monthly basis.
As with any resource that goes unchecked, it’s easy to consume more than you planned. Old data isn’t always purged systematically and tends to just keep growing, causing waste and impacting performance. Consumption-based solutions provide self-service usage dashboards and budgeting capabilities that alert you when predefined usage or spending thresholds have been reached.
Respond smoothly to changes in capacity demand. The pay-as-you-go model makes it easy to mobilize the resources you need to handle periodic data-heavy tasks like end-of-year reporting, unpredictable spikes in market demand, or new business models. And when you no longer need that capacity, you won’t have to keep paying for it. It’s important to understand how your capacity is managed by the vendor. What is the lead time for delivering additional capacity? Is there a collaborative approach to capacity planning?
Free your admins from the minutiae. Managing storage is labor intensive. The AI capabilities of intelligent storage eliminate many issues that would otherwise tie up IT staffers’ time. The management and support services that are baked into consumption-based IT take that even further. When your admins are not knee-deep in managing storage at a granular level, they can tackle high-value projects and power up your organization’s drive to innovation.
Time to make your move?
The cloud-like economics and agility of consumption-based storage is a natural fit with today’s hybrid cloud environments. To decide whether it’s the right fit for your organization, assess your full range of requirements—including backup, data migration, availability, security, and compliance—and consider how consumption-based storage can help you:
- Ensure that your data is where it needs to be across your hybrid cloud estate
- Manage complex infrastructure from core to cloud to edge, including multi-vendor solutions
- Meet new and existing regulatory, compliance, and data sovereignty challenges
- Free resources for innovation as you navigate a hybrid world
- Bypass outdated business models that put a strain on capital and operations
AI-driven storage, in consumption mode, is coming just at the right time for the industry at large. We’ve all seen the statistics on the corporate data explosion, the march of the zettabytes, and the still-unfolding impact of the Internet of Things.
But here’s the thing: To thrive in the data economy, companies will need not just more storage, but smarter, more flexible, more financially streamlined storage. That’s what consumable intelligent storage delivers, and its future looks bright.
Consumption-based intelligent storage: Lessons for leaders
- Built for the cloud and delivered as a service, AI-driven intelligent storage proactively provides real-time insights on everything from performance bottlenecks to security threats, seasonal spikes, and other changes.
- When evaluating an on-premises, consumption-based approach to storage, companies need to consider the following criteria: data privacy, latency, connectivity, upfront costs, capacity management, and agility.
- Pay-as-you-go intelligent storage opens up new possibilities, including the ability to launch new projects without large upfront investments, simplify IT budgeting and planning, and respond to shifts in capacity demand efficiently.
This article/content was written by the individual writer identified and does not necessarily reflect the view of Hewlett Packard Enterprise Company.