Many IT organizations assume that security risks increase with a shift to cloud computing. The reality, however, is not so clear-cut. In fact, many of these same organizations will be surprised to learn that adopting cloud operating models with appropriate governance and security controls can actually reduce the level of risk relative to their current IT environments. Here's why:

IT professionals frequently develop unwarranted security concerns regarding cloud computing primarily because cloud environments are dynamic and enable new levels of workload portability that are very different from what they're familiar with. In cloud environments, application workloads can be moved to totally different physical infrastructure or service providers from one deployment to the next. The underlying application data can move even more frequently, depending on the type of instance and persistent storage options you've selected.

This means your security boundaries have to be dynamic too. They have to move with the workload and the data, and self-configure themselves in new environments in a consistent and automated manner.

Taking Cloud Security for a Spin
A simple analogy can be made between securing cloud workloads and securing a car. When you park your car in your home garage, typically you just close the garage door and that's it. You assume your car is safe inside your garage along with your other belongings, so you typically don't worry about locking your car doors or taking other precautions.

However, when you park your car somewhere else, you typically lock the doors to secure it. There are several ways you can do this. The door locks could be activated by a remote, a keypad on the door, or the proximity of an RFID tag in the key fob. You may decide to upgrade your security by adding a factory alarm system, steering wheel lock, LOJACK tracking system, or other security system depending on the car's value. Finally, you can also decide where to park your car depending upon your risk tolerance. For example, you may accept your favorite restaurant's offer of valet parking in a monitored lot instead parking down a secluded street.

The point is that you can create a portable security boundary around your car that can be equal to or even more secure than your garage. Cloud security is similar in concept where portable cloud workloads offer a wide range of options to establish a very effective portable security boundary. In fact, cloud workload security has an additional important benefit over the car analogy, which is that security configurations can be completely automated and policy-driven. Using the car analogy, this means you no longer have to worry about forgetting to lock your door or arming your alarm system in the parking lot, because the car will automatically do it for you.

Under the Hood: Cloud Security Options
This new approach to securing a moving workload is a big departure for many IT groups that are used to working in more static and controlled environments (similar to the home garage). These IT groups are used to working with physical data center infrastructure, traditional firewalls, mostly static networks, and familiar resources that they own and control. The idea of moving workloads in and out of new environments they don't control is a big concern, especially knowing they've expended tremendous time and attention manually configuring their own environment.

However, today a broad range of proven technologies can deliver consistent, automated security for portable cloud workloads. They include virtual private networks, encrypted data storage, host intrusion detection systems, hypervisor-based firewalls, and federated identity management systems. These systems can complement each other to provide an end-to-end security solution that encompasses instances, data, network, and role-based access as desired.

Importantly, these systems can be automatically enforced through security policies that essentially eliminate the risk of human configuration errors (e.g., forgetting to lock the car doors). In other words, you can design a customized level of security into each of your cloud workloads from the beginning, so that they consistently and automatically establish the desired security boundary conditions each and every time they are deployed. These security policies are designed to be abstracted above the cloud implementation layer, and can be enforced across multiple heterogeneous cloud deployment environments.

Benefits That Extend Beyond Security
This leads to important business benefits that extend beyond the typical risk mitigation aspects of providing security. By automating end-to-end security configuration, you can make significant gains in reducing workload provisioning and deployment time. Depending on your current security processes, this can make a very significant impact on improving business agility. Typically, the upfront investment to initially create automated security policies gets paid back in dividends many times over by automating enforcement and enabling more rapid deployment. You can also significantly reduce the complexity of managing and changing your security rules over time since policies can be enforced broadly or fine-grain across your organization, and modifications take effect right away.

Although securing cloud workloads is important, it doesn't stop there. You'll also need to revisit the people, processes, and management systems used to govern workload deployments. Insufficient control over who can provision a workload, where it can be deployed, for how long, and at what cost or capacity can be a recipe for disaster, even if the workload itself is secure. Beyond obvious cost and capacity management concerns are additional risks associated with regulatory violations and inadvertently deploying assets to the wrong environment.

Governance Rides Shotgun
For these reasons, cloud security and cloud governance need to work hand-in-hand. Enterprise cloud governance often gets complicated very quickly due to the many-to-many relationships that exist among workloads, user groups, deployment environments, departmental usage policies, industry regulations, geographic restrictions, and other attributes that exist in a large enterprise. As a result, policy-driven governance platforms are critical to managing and controlling all these permutations in a transparent and automated manner.

Effective governance requires several key capabilities, including a policy framework that is extensible, so that organizations can customize their own industry- or company-specific policies when needed. It also requires a governance platform integrated with your cloud provisioning and implementation layer, so that governance is consistent and enforceable across all of your workloads and all of your internal and external cloud environments.

Tackling the Security Challenge
Addressing cloud security is an important milestone you'll face as you embark on your cloud strategy. At times, security challenges may seem too complex or insurmountable, so here are a few tips to help make those initial projects more manageable.

First, realize that cloud security represents a spectrum of capabilities that you can tailor to your needs. You probably don't want to implement all of it right away, so screen initial projects and workloads based on security sensitivity and their target cloud deployment environments to lower the initial security risks and requirements.

Also realize that there's a broad range of cloud tradeoffs you can make to lower your risk/security exposure in exchange for partial sacrifices in economic and agility benefits in the short term. For example, some organizations set up their first external private clouds at their current hosting providers with dedicated hardware in a locked cage using an existing dedicated network link. This approach may cost a bit more, but you can start with this conservative security approach and still prove out aspects of your cloud operating model while building trust and security experience over time at a comfortable pace.