Managing multi cloud through the ‘by design’ approach

Agencies are in year 12 of the cloud strategy. The Office of Management and Budget launched its cloud first strategy back in December 2010, and June will be the four year anniversary of the shift from cloud first to cloud smart.

Over the last 12 plus years, agencies have made a lot of progress to take advantage of cloud services.

For example, the FedRAMP cloud security authorization program reached a milestone of having more than 300 approved cloud services. A majority of them are in the software-as-a-service category, meaning agencies are moving toward a more advanced stage of cloud services where infrastructure is mostly in the cloud and now they are looking at applications and workloads.

The challenge now, of course, is most agencies are living in a multi-cloud environment. This is where agencies are integrating services from multiple cloud providers to meet their mission needs.

Kelsey Monaghan, the lead for federal strategic programs and partnerships for cloud and edge at Dell Technologies, said while cloud services aren’t new for agencies, what is changing is the definition of what cloud is continues to expand as the operating models evolves.

“We’re seeing that definition and that use case really expand to the core and out to the edge so that continuum of cloud is really expanding,” Monaghan said on the discussion Innovation in Government. “As we look at that, the need to have it by design approach to avoid some of those silos, and improve that operational benefit is absolutely where we’re seeing not only a focus of the federal government, but also of industry in that deep partnership.”

Portability, flexibility are key

The concept of “by design” related directly to the use of more than one cloud provider whether public or private or hybrid instantiation.

Monaghan said agencies need to understand their workloads and which cloud makes sense to optimize those efforts. She said key factors include portability between cloud providers and flexibility within those vendors.

“When we talk about multi cloud by design, we’re really talking about an ability to govern and provide workload flexibility across all those individual cloud deployments so that consistency is really the underpinning of that by design methodology,” she said. “When we talk about workload optimization, or when we talk about by design, what we really mean is understanding those with our agency partners and bringing the best of bread from an industry perspective to place those workloads in either the public, hosted or co-located cloud deployments or even edge deployments that would best fit that mission or that outcome.”

Agencies have come to realize over the last decade that not all workloads are created equal and the cloud isn’t always the only answer based on consumption trends.

Monaghan said understanding your workloads and applications becomes more important as agencies push data and services to the edge.

She said more agencies are looking at a distributed model for services to the edge.

“There’s analytics use cases and those time-to-value discussions are continuing to drive cloud discussions, but in a slightly different fashion,” Monaghan said. “What we’re really seeing is that in some cases, there are repatriation of workloads or the movement of workloads from the public cloud back to on premise or in many cases to co-located or to different hosted environments. Some of those are because they weren’t necessarily entirely cloud ready. They weren’t re-platformed to really enhance and utilize the benefits that the public cloud providers and those tool sets really provided to them day one. So to get those better cost and operational efficiencies that we have seen some of that movement.”

Multiple clouds vs. multi cloud

Another factor that agencies are becoming smarter about is the different between using multiple clouds and taking a multi-cloud approach.

Multi-cloud is about portability of workloads, consistent governance and management of cloud deployments and creating a continuum of services from the core to the edge.

“When we say multi-cloud, we’re talking about whether you have applications consistency in the data layer, in the fabric and the connection strategy across these cloud environments. That really is an ecosystem approach,” Monaghan said. “When we look at these discussions with our agency partners, we’re looking at industry and what each provider really brings from a best of breed perspective. We believe it’s not a public cloud or edge cloud or co-location or public, it really isn’t a discussion, and because of that we’ve partnered not only from a cloud native perspective around bringing those tool sets right across that continuum, but also from a data layer perspective to provide that consistency.”

She added by relying on a broker of brokers approach to manage a multi-cloud approach, agencies can improve their experience and ensure operational efficiencies.

“What we look at is understanding if that workload today is really supporting the outcome or the mission of the agencies. What we hear a lot about today is access to data or collaboration experiences as different areas of the organization are using IT,” Monaghan said. “Some of the challenges that we’re hearing are absolutely around proximity, access and collaboration tool sets, but also around latency and the time to value discussion. Some of those can be indicators if a workload is really running the best place. Others can be around the application itself. Is it legacy? Has it been re-platformed? And is it running an environment where it’s taking advantage of the best capability that an agency is paying for? You really can look at both that operational and financial discussion as well as that operational right and technical capability. Is the workload running in a place where it can utilize and really integrate with the necessary back end resources that it needs.”

Listen to the full interview:

Copyright © 2024 Federal News Network. All rights reserved. This website is not intended for users located within the European Economic Area.