Optimize Storage With Multicloud By Design

Sponsored Feature The challenge of storing data in the most efficient and cost effective manner has become complicated, with CIOs facing a seemingly endless array of conflicting choices.

What data, applications and services should they retain in an on-prem environment and what could be safely committed to a public cloud platform? Under what circumstances does it make sense to repatriate data that it once seemed best to send offsite, and how can this be done? If public cloud is to be part of the solution then which platform is best for which workload? Should the choice be single cloud or multi-cloud? Is a hybrid of the various options the right way forward, or a recipe for confusion?

Success may depend on deciding which model represents the correct balance of value, performance, security and compliance for a particular organization and its individual business needs. This inevitably means the factoring of many variables. And business priorities, of course, may alter over time while perspectives can change as new ways of doing things become available.

Those tasked with navigating this stormy ocean of possibilities should at least be glad that the debate has moved on from the orthodoxy of 'cloud first' which held that all new applications and services belonged in the cloud, no questions asked. It is now apparent that the economics of data storage demand a more nuanced and flexible approach than this.

What CIOs need is fewer obligations and commitments and more choice. A multi-cloud or hybrid approach is often the best solution, but not if it leads to friction and loss of freedom in moving workloads around. No organization should be deterred from repatriating data from cloud to home when circumstances dictate that this is the right way forward simply by the complexity or cost of doing so. Quite simply CIOs need a way to manage data wherever it is stored, helping them to handle hybrid and multi-cloud storage architectures while resolving worries over data repatriation.

When public cloud services first hit the mainstream, it seemed to many IT chiefs like a no-brainer to offload the headache of managing internal infrastructure to someone else, and many were seduced into paying what seemed like a modest initial cost for the privilege of doing so. The idea gained ground that public cloud is inherently less expensive than keeping IT in house, delivering greater and greater economies of scale the more stuff you migrate.

But the reality is that multiple calculations play a part in assessing the balance between storing data on-prem versus in the cloud. It might be wise, for example, to weigh up the impact on data mobility when moving workloads between on and off-prem sites. Use of cloud services will naturally create a degree of latency which could impact the effectiveness and indeed viability of certain applications. Plus, some cloud providers will impose egress charges each time data is moved off their platform, affecting the value proposition in ways that might not at first have been obvious. Every aspect of cost must be assessed when the economics of data are calibrated.

Time for an objective look

The advent of new technologies, or new ways of applying existing technologies, can precipitate a rethink of storage strategy, altering the on-premise versus off-premise value paradigm. We've seen this happen over the last few years in the form of innovations that have boosted the performance and reduced the cost of in-house capacity. We're talking here about software-defined storage architectures and object storage.

Object storage is often a good choice for managing the kind of unstructured data that a modern digitally-focused organization usually has in vast quantities. When deployed on today's all-flash storage arrays (known also as solid-state arrays), it can add even more software-driven speed, performance and agility.

Given good built-in management features, CIOs can use object technology to host their own on-premise storage solution at a lower overall cost than with public cloud offerings. When deciding whether to repatriate workloads from cloud back to in house, object storage has the power to clinch the argument, taking stress and uncertainty out of the equation.

"Traditionally object was used for backup and archive use cases, replacing tape," explains Anahad Dhillon, Director of Object Storage Product Management with Dell Technologies. "Now, in addition to cost and reliability, people are asking us about performance and functionality. This is because object storage is used for business-critical applications like analytics, AI, ML and modern workloads like streamed data, with a shift towards primary storage use cases."

The Enterprise Strategy Group (ESG) recently conducted a study investigating object storage, paying particular attention to Dell's ECS solution. The analysis revealed that on-premise storage with ECS can provide a cost advantage over public cloud, ranging from 50 percent for cold archive storage up to 76 percent for active storage, cutting data access latency at the same time. ESG said its findings were based on conservative assumptions offering the possibility that more aggressive configurations could yield even greater savings.

ECS delivers by being an eminently scalable and highly efficient global object storage solution, says Dell. Its software-based design means CIOs can deploy industry-standard hardware and ensures they get all the low-cost, ease-of-use and scalability benefits associated with public cloud storage without the security risks.

"A large customer of ours is saving over $2M every month and getting 3-4x the performance by repatriating their data on a scale-out solution built on the ECS EX5000." says Anahad.

Multi-cloud solutions provider Rackspace Technology is another company that has put Dell ECS to work as the basis of its Object Storage service. Michael Levy, Director of Product, Private Cloud with Rackspace, explains: "We have a datacenter footprint all over the world and solutions covering the full IT stack, both virtualized and non-virtualized," he says. "Customers need to extract the maximum business value from the huge amount of intellectual property they are storing. Object's flexibility and ability to tag with metadata allows you to extract that value."

Levy points out that 'born in the cloud' workloads have relied on object for the past 10 years or so, because that's the storage of choice in the hyperscale sector: "Now you're seeing more traditional enterprise applications writing to object," he notes.

Cheap and deep no longer the object

To a certain extent, the old assumptions about public cloud's cost benefits have been turned on their head with the rise of inexpensive object-led solutions. Quite simply, the days of object being used for just 'cheap and deep' storage are fading away, according to Levy.

"In order to achieve a multi-cloud reality, you must deploy your storage in an egress-neutral zone – data egress makes the public cloud economics untenable," he says. "CIOs have gotten fired when a large egress bill landed on their CFO's desk. You need a way for your data to bounce around to where it's going to be most useful, without financial penalties, either on prem or in the cloud."

Levy says Rackspace is excited about moving to the next level with a new globally distributed multi-tenant ECS platform: "ECS inherently supports multi-tenancy which is incredibly valuable. Our aim is to fill a space in the market for this type of high-performance object storage."

As more and more workloads get repatriated and as data volumes soar, both inside and outside the public cloud, driven by hunger for better business intelligence and analytics and the need to incorporate AI, it will become ever more essential to provide a highly performing infrastructure that supports timely decision-making, further driving the shift toward object storage.

Built upon an object storage architecture, Dell ECS can tie it all together, delivering a number of advantages that significantly reduce the cost of storing, protecting, and accessing high volumes of data. It minimizes data retrieval costs, egress charges and additional cloud fees while reducing risk and cost of moving workloads to where they belong.

For many organizations, it's truly a storage solution that can meet all current data needs.

You can sign up to watch an in-depth discussion on how to optimize storage with multi-cloud by design featuring storage solutions experts Anahad Dhillon, Dell's Director of Object Storage Product Management, and Michael Levy, Director of Product, Private Cloud, at Rackspace by clicking here.

Sponsored by Dell and Rackspace.