Anyone involved in software product creation should be familiar with DevOps, a set of practices that bring software development and IT operations together with the goal of shortening the development life cycle and providing continuous delivery and high-quality products.
As enterprises increasingly move their application development and workloads to the cloud, and as spending on those clouds becomes more complex, a related concept, CloudOps, or "cloud operations and maintenance," has emerged.
Here, we'll look at what CloudOps is, how it will benefit your enterprise, and what key things to keep in mind when implementing CloudOps in your enterprise.
What is a CloudOps?
CloudOps is an operational practice for managing the delivery, optimization, and performance of IT services and workloads running in a cloud environment.
Whether an enterprise adopts a multi-cloud, hybrid cloud, or private cloud strategy, CloudOps aims to establish processes and best practices for cloud-based processes, just as DevOps does for application development and delivery.
CloudOps: A multi-level framework for cloud operations and maintenance
"Holistic CloudOps is a multi-level framework that can be used to help organizations manage all aspects of their cloud ecosystem," says consulting firm Capgemini
Said Jason Hatch, vice president and head of Americas' Cloud Center of Excellence.
One is the governance layer, which includes activities such as financial operations (also known as FinOps) to control costs and manage cloud budgets. "The governance layer should also contain architectural standards for how to deploy what in the cloud, and have a way to implement those standards programmatically." Hatch said.
Other framework layers include the cloud application layer, which covers how the enterprise will deploy and manage/monitor applications and application-specific services in the cloud; The cloud operations layer, used to deploy, manage, monitor, and operate cloud services; And the cloud-based layer, which includes core services such as identity, network management, logging, central backup management, infrastructure as code, and central monitoring capabilities.
"Straddling all of these layers is the 'security layer,' which includes vulnerability and threat management, workload protection, and integration with the company's larger cybersecurity management capabilities." Hatch said.
Application of CloudOps in the enterprise
The CloudOps model has particular relevance to the delivery of applications, which many enterprises are focusing on through digital initiatives aimed at increasing sales and enhancing the customer experience.
"CloudOps is able to bring together five major responsibilities for building, deploying, running, monitoring and managing application delivery capabilities in the cloud," said Suresh, executive vice president of engineering and Operations at Replicon, a cloud service provider
Networking, computing, security and storage are four key components that must be kept in mind during the initial construction and design phases, Kuppahally said. "This way, companies can deploy their applications automatically, or they can have continuous integration and continuous delivery." He said.
An enterprise's CloudOps team should be clearly separated and operated independently from the engineering or product teams, which allows CloudOps to achieve "transparency and quality of service [QoS] accountability" in the enterprise, Kuppahally said.
The benefits of CloudOps
Starting with an enterprise's overall deployment of cloud services, the business benefits of CloudOps are considerable, says Capgemini's Hatch.
CloudOps helps drive further enterprise adoption and use of cloud computing. If enterprises can effectively deploy, manage and protect their cloud environments, they should increase their use of the cloud and provide the ability to experiment and innovate with services and technologies, he said. "This in turn can make them more flexible, provide faster time-to-market and help drive innovation."
Businesses using CloudOps can also better manage the growing number of cloud services they use, Hatch added.
"We keep hearing from customers that they are exceeding their cloud budget, but they either don't know why or can't put in place controls to manage it," Hatch said. "An effective CloudOps would help alleviate this problem. At the governance level, we can implement better budgeting and financial tracking and optimization. This also facilitates better automated deployment and management at the operational level."
Will, managing director of consulting firm Protiviti
Another of the biggest benefits cited by Protiviti's customers is the ability to automatically release authorized resources in the cloud, which can help enterprises manage increasingly complex cloud environments, Thomas said.
Enhanced security is another key benefit of CloudOps, Thomas said, as the model can "ensure conformance to security controls, standards, or frameworks, and develop strategies to limit noncompliance while reporting health and activity within the cloud."
Thomas also believes that companies that implement CloudOps will be better able to optimize their cloud environments because "CloudOps engineers will focus on modernizing applications with the latest and greatest services by leveraging licensed resources in the cloud." He said.
In addition, enterprises deploying CloudOps can establish reasonable resource allocation schedules based on performance and cost considerations; Ongoing reporting and review of cloud health indicators; And supports proactive allocation of resources while maintaining regulatory compliance within the cloud, he said.
Replicon's Kuppahally points out that CloudOps can scale cloud services cost-effectively without affecting QoS. "It makes a lot of strategic sense to combine QoS objectives with CloudOps investments," he said, "because a dedicated CloudOps team can be motivated to manage operating costs and therefore have a vested interest in reducing operating costs."
CloudOps in practice
One company that benefited from adopting CloudOps was Stretto. The bankruptcy services and technology company, which serves corporate and consumer bankruptcy departments, identified the need for CloudOps practices early on and incorporated key principles into its applications and systems that run in the cloud, says Its CTO, George
"For example, we have a strict, fast rule that we only use infrastructure-as-code [IaC] practices for deployment," Tsounis said. "We decided to achieve redundancy by saying that all applications/systems will always run across two availability zones, so we leverage the high availability capabilities built into the cloud provider."
A key part of Stretto's strategy is leveraging CloudOps' practices to ensure it has a more proactive approach to its technology operations, Tsounis said. "We prefer to empower our architects and engineers to create high-performance, self-healing and resilient cloud-on-premises solutions for our internal and external customers, rather than continue to operate in a passive manner." He said.
The introduction of cloud services, and even the transition to serverless capabilities, presents unique challenges, Tsounis said. "CloudOps is a strategy to help us meet these challenges." He said.
The benefits that CloudOps ultimately brings to Stretto include reduced cost, scalability, automation, simplified disaster recovery, and seamless integration of the infrastructure as part of the application.
"Our team has benefited from overall improvements to the application, and these CloudOps concepts have been fully adopted," Tsounis said. "CloudOps practices also improve software quality. This is achieved by using the IaC approach to make the deployment and configuration of the cloud infrastructure repeatable. We reduced configuration errors because IaC enabled a consistent infrastructure configuration as we rolled out applications across our various environments."
By eliminating manual configuration of cloud infrastructure, Stretto has reduced quality issues by about 20 percent, Tsounis says.
"Leveraging the CloupOps practice provides engineers with the confidence they need to know that application/system behavior in a pre-production environment will be the same when released to production," he said. "In addition, we also saw an improvement in our overall IT operations due to improved quality of our applications and fewer help desks and internal tickets."
Keep up with evolving methodologies
Nothing is set in stone when it comes to cloud services and how they are used, so businesses using CloudOps need to adjust their approach regularly to keep up with changes.
This is still new territory for many companies and there is a learning curve to overcome. "As more and more enterprises adopt true multi-cloud deployments, their CloudOps implementations need to mature and scale," Capgemini's Hatch said. "Many customers manage their cloud native maps in silos, using different tools and processes, and with little ability to view their entire cloud native map as a whole."
To improve efficiency and effectiveness, "companies need to develop their own CloudOps frameworks so that they can easily integrate new cloud providers and services while still providing the right level of management, monitoring, and operational rigor." Hatch said.
The way companies handle incident management in the cloud could also improve, Kuppahally said.
"This is something that most CloudOps teams struggle with," he said. "They are overwhelmed by internal and external events and lose the means to manage them effectively. Establishing a dedicated project management process to simplify filtering and prioritizing event management is one way to reduce risk."
At the same time, companies also need to reduce the rate of false positives of events. "When CloudOps teams can't keep up with high false positive rates, they get stuck," Kuppahally said. "Having an effective strategy and plan to reduce or eliminate false positives is a critical success factor."
CloudOps could benefit from technologies such as AI and machine learning, said Aref Matin, executive vice president and chief technology officer of Wiley, a research and education services provider.
"Through machine learning, CloudOps tools can help define enterprise-wide policies, detect and report exceptions, and be able to take corrective action in an automated manner to maintain best practice policies in the cloud environment." Matin said.
Like DevOps, the success of CloudOps depends heavily on developing a culture that makes full use of frameworks and tools. As more enterprises move more of their work and processes to the cloud, they will also need to focus on building CloudOps expertise.
"When dealing with cloud computing, most customers are in a reactive state, unable to respond to events, changes or requests for new services," says Protiviti's Thomas. "CloudOps establishes a deployment structure through automation that allows monitoring, review, and optimization of existing resources, as well as reviewing corporate policies for alignment with the cloud environment."
Stretto's Tsounis agrees that businesses need "a broader understanding of the proper organizational structure, expertise and appropriate adjustments to collaboration that CloudOps really works."
"CloudOps is not a separate team or department. IT, security, architecture, and application teams all need to collaborate and align on common CloudOps practices, "the CTO said. "CloudOps doesn't work very well if these teams work in isolated islands."
Based on his experience putting CloudOps into practice, Tsounis believes that in order to succeed, businesses also need to better define the basic skills required for CloudOps, rather than reinventing the wheel.
"Technology teams need to understand cloud-based architecture, networking, security and automation," he said. "Without basic skills, teams can risk implementing solutions where cloud services already exist."