Peak Performance: Key Pillars for Enterprise Cloud Operations

When government agencies migrate into cloud environments, they often do so with an eye on anticipated benefits. Whether those include a more flexible IT infrastructure, solutions that can easily scale to meet evolving mission needs, or lower IT infrastructure costs — we know that cloud can enhance the way organizations manage and deliver capabilities. 

But while a benefits-driven approach helps jumpstart cloud transformation and assess which tools to adopt, many organizations start their journey without a critical piece in place: operations. 

“The operating playbook for a traditional on-premises approach is thrown out the door when an agency embraces cloud," says Brad Beaulieu, principal at Booz Allen Hamilton. “You don’t want to discover operational challenges at the point of a production or security issue, when it’s too late.”

Incidents like the SolarWinds hack or, more recently, the Log4j flaw, which left millions of servers vulnerable to potential exploitation, require clear visibility into network activity, operating agreements in place with service providers, and the ability to mount a real-time response to enterprise threats.

“Whether it’s a zero-day vulnerability, and there’s a need to quickly understand and address threats — or it’s ineffective processes, policies, or resource models that may have worked on-premises — traditional management and operations lead to problems in the cloud,” Beaulieu explains. “And at a minimum, legacy operations make it difficult or impossible to realize the benefits that cloud has to offer.” 

Fortunately, inefficiencies and vulnerabilities can be avoided when agencies start to transform operations early in their cloud modernization journey.    

Start Right With Automation 

While automating manual processes is something available on premises, organizations in the cloud have access to new opportunities for "out-of-box" automation through service providers. Prioritizing automation is critical in order to take full advantage of efficient and flexible cloud operations. 

“Things that used to take an engineer hours to configure can all be done in a matter of minutes or even seconds,” says Osama Malik, vice president at Booz Allen Hamilton. “Security and network configurations, enterprise policies and controls can all be built into the way that assets are deployed through automation.” 

The result? A highly efficient workflow where engineering staff can focus more on critical and core mission services. The more mature organizations, Malik adds, develop applications in cloud-native formats and can then start to automate cloud resources to scale up or down based on demand — enabling heightened flexibility that simply isn’t available on-premises.

Still, Beaulieu says, automation requires significant upfront time which can be a barrier to IT leaders. 

“Organizations may say they don't have time to automate today, because they forgot to do it last year. But it’s a vicious cycle because automation saves a ton of time later on,” he adds. “And it doesn’t just impact one person’s workflow — it can empower groups and divisions to innovate faster in the long term.” 

For example, agencies that have automated their DevSecOps pipeline can update, secure, govern and test through an automated workflow. Of course, there’s a trade-off that needs to happen between automating processes — so that the workforce can take advantage of new products and won’t be inhibited by repetition or manual error — and allowing for flexibility between projects. 

Create the Conditions for Innovation and Scale

More government organizations are starting to rethink the way they manage and standardize cloud across the enterprise. A consolidated operating model offers benefits such as streamlined processes, technical governance and shared resources to innovate more quickly. It also brings down the cost of ownership — including the gain from shifting the workforce to focus more on the core mission and less on IT operations.

But Malik cautions it takes a cultural shift to ensure cloud adoption and buy-in from stakeholders across the organization. 

“The biggest barrier for a successful cloud transformation is still primarily people and culture,” he says. “If the enterprise services don’t clearly add value for teams, or if the structure isn’t set up to empower teams to innovate and test new ideas, operations will continue to be a challenge.”

A deliberate approach is required to identify the common needs of an organization’s major programs, and then clearly communicate that teams are still able to customize and innovate, with the added benefit of core services at their fingertips. The next cultural shift then involves changing the way capabilities are developed, by moving to a “product mindset.”

“A project has a start and an end date, but a product is dynamic,” Beaulieu explains. “We’re starting to see the benefit of product teams oriented around customers, responsible across services and applications used by those customers. This focus on the customer changes the operating mindset so teams are not just delivering a new application or feature for today’s requirements but are looking forward at changing user needs into the future.”

Ultimately, successful cloud operations enable an investment in innovation and continuous improvements. Changing the model with a focus on enterprise services and product development helps facilitate experimentation through sandbox environments where teams can be nimble and rapidly test prototypes for cloud transformation without risk or red tape.

Right-Size Services for the Mission Needs 

There’s a common fallacy that having more control of IT is the way to best manage it, secure it and respond to change. And especially in the cloud, there is a tendency to opt in for all the bells and whistles that are now available for operations and performance. But the inclination to over-engineer — and over-control — can quickly balloon the budget or create an unmanageable cloud enterprise. 

“There's a big misnomer out there, and people will look at cloud and say, ‘we thought it would be cheaper than it is,’ but they're not thinking about the true costs associated with running the cloud, including all the security services required to prevent and address increasing threats,” Malik explains. 

More organizations are considering how to effectively ‘right-size’ their cloud operations so they can take advantage of cloud commodities, engineer at the level require for the mission, and continue to pay down technical debt. 

“It’s really easy to architect your cloud beyond the capabilities required by your business — and the costs of maintaining that accelerate in an exponential fashion,” says Malik. “Agencies that don’t actually need all of the various services or particular aspects of a service level agreement should take that money to better deliver the core mission and innovate for future needs.”

Commit to the Journey Without Cutting Corners 

Most agency leaders can agree that cutting corners with cloud operations and implementation isn’t an option. Taking extra measures upfront to set up the most effective organizational and IT operation will pay dividends and avoid new technical debt for the enterprise. 

“Technical debt is a lot like real debt,” Beaulieu says. “When you buy something using a credit card, you end up paying a lot more over time because of all the interest payments. And when you cut corners for IT modernization, and don't set things up properly, you pay for that over the long term.” 

To do things the right way, it may mean a longer and more extensive start to cloud transformation — but then you can start to find efficiencies after the first year, Malik adds. One of his clients, he shares, migrated their first application by using the same processes and manual configurations they used on-premises. While it was inefficient, they were able to create an initial baseline of what could be improved through best-in-class operations. Through automation, they subsequently decreased configuration from three months to two days.

But this level of operational change requires the commitment of diverse stakeholders — from leadership and tenants to external customers of an agency’s cloud services. In some places, there may be resistance to transformation, and it’s essential to enhance trust and transparency so stakeholders can understand what’s happening under the hood.

“As you move to the cloud, you’re outsourcing critical aspects of IT operations to other providers. Making sure there’s visibility into performance and security, and how the system is responding to issues, allows business and application owners to feel confident in the cloud environment,” Malik says.

Open communication will be key for agencies looking to enhance their cloud operations moving forward. And it’s not always the success stories that move the needle.

Beaulieu emphasizes the importance of sharing mistakes — not just wins. 

“To truly advance cloud operations — automating, building around customers, and continually investing in a cycle of improvements — isn’t always a straight path and there will be issues that come up,” Beaulieu adds. “Being transparent when something goes wrong and showing how any issues were addressed or what can improve in the future, can turn people who were once resistant into advocates and evangelists.” 

Osama Malik is a vice president at Booz Allen, leading digital transformation projects across civil government; Brad Beaulieu is a senior technologist and cloud architect, leading the firm’s cloud security offering. 

Find out how Booz Allen Hamilton’s team of experts can help you optimize your agency’s cloud operations strategy. 

This is part of Booz Allen Hamilton's "Government Cloud" series. Click the links below for other articles in the series: 

What to Expect for Federal Cloud Innovation in 2022 and Beyond

Mission Critical: Leveraging Cloud Resilience for High-Stakes Delivery