As organizations continue to send more services into the cloud, many IT leaders are beginning to discover that their patchwork approach is leading to inefficient and sometimes even counterproductive results. Cloud experts, such as Ken Englund, technology advisory leader at business consulting firm EY Americas, understand that cloud optimization has become essential to tapping into the cloud's full value. "Organizations that optimize their cloud operations move beyond being 'on-the-cloud' to exploiting 'in-the-cloud' ways of working," he said.
Plan before deploying
A well-planned cloud deployment strategy can save money and dramatically improve time to market for new solutions, observed Mark Cressey, executive vice president and general manager, IT hosting services, at Liberty Mutual Insurance. "Best practices for maximizing the success of cloud operations include centrally managing cloud resources and establishing standards and best practices for cloud environments, such as reserved instances to maximize value," he noted.
Cressey pointed out that it's also important to embrace DevOps and a pipeline-driven approach to leverage the cloud's inherent capabilities, such as autoscaling, a self-healing infrastructure, and multi-AZ-aware services. "A well-automated pipeline ensures controls are repeatable throughout the organization and enforces the centralized management of the infrastructure," he explained.
Since small changes can lead to major costs, engineer education is critical to ensure that teams understand the impact of design and deployment decisions as they build and manage solutions. "Establishing best practices and auditing cloud usage will keep the cloud operation running smoothly and efficiently," Cressey stated.
Deep visibility is essential for cloud optimization, and organizations can turn to various tools to monitor and measure cloud usage and associated costs. "These tools help organizations reduce waste by alerting users to lowered demand or automatically scaling usage to optimal rates," said Mehdi Aftahi, CTO at Technology Evaluation Centers, an impartial software selection and advisory firm. Organizations can also use visibility tools to increase the efficiency of their cloud applications. "With built-in reporting features, IT can easily identify the waste and redundancies in cloud operations," he noted.
Orchestrate and automate
Orchestration and automation technologies are a proven way to simplify complex cloud environments. Over time, cloud deployments can become individual compute platform islands that fail to fully integrate into operational or business workflows, degrading both cost and efficiency, noted Bob Lamendola, vice president, infrastructure and engineering services, at printing and office systems provider Ricoh USA. Due to cloud architectures' dynamic nature and versatility, equal attention and investment must be made to platforms to ensure that they're integrated properly, guaranteeing that maximum value is achieved, he said. "Cloud operations are not solely about the cloud technology, they're also about the surrounding ecosystem in every organization that enables them to engage with customers, provide services, and efficiently transact business," he added.
By embracing a holistic approach to cloud operations, organizations can not only deliver new services to their customers, but also ensure that those services drive business results, Lamendola explained. "The underlying orchestration and automation technology has evolved to a point that it's within reach of most organizations," he said. "Specifically, investments in automation drive down the cost of delivery while improving the speed to market lifecycle."
Conduct application and cost reviews
As they strive to keep their cloud operations optimized, organizations need to periodically review their application infrastructure as well as the costs associated with maintaining applications. "As companies grow or shift their operations in different directions, the applications, and the costs associated with them, may not necessarily align with the new direction of the operations," Aftahi said. "It's crucial for organizations to review their application footprint and compare it with business process requirements on an on-going basis," he noted. Doing so will allow them to make any necessary adjustments and keep costs under control.
Build product operating models
Product operating models allow organizations to organize themselves around their software products, many of which are cloud based, and to streamline their product feature and support workflows through de-matrixed, end-to-end teams. "The end-to-end teams [will] increase customer outcome ownership and empower an increased usage of cloud native services that improve speed to innovation," Englund said.
Product operating models expand on DevOps models, enabling greater end-to-end ownership by aligning identical resources to support both feature development and production support, Englund noted. "These end-to-end teams feel the impact of customer challenges and have the ability to resolve them," he added. "These teams are also well positioned to leverage working [in-cloud] methods, tools, services, and more to make rapid and meaningful customer impacts."
Many organizations unintentionally hinder their cloud operations by relying on incomplete strategy plans and deploying a poor cloud security architecture. This typically happens when IT departments fail to understanding their role in cloud security, or carelessly migrate existing on-premises applications to the cloud without first adapting them to the new security environment, observed Eric McGee, senior network engineer at colocation provider TRGDatacenters.
When viewed from a management perspective, cloud services don't differ very much from traditional IT operations. "Although a more efficient compute model that yields infinite possibilities and unlocks opportunities, [the cloud] needs to be managed and invested in like any other operational structure, using sound business principles," Lamendola said. "The challenge is not to allow those sound business principles to disrupt innovation."
The migration to the public cloud should be viewed as a journey, Cressey added. "Fail fast, learn faster and adjust your cloud operations as needed to achieve success."
For more on cloud optimizations and strategy, follow up with these articles: