The Cloud and Cybersecurity

This post sponsored by the Enterprise CIO Forum and HP.

As part of federal CIO Vivek Kundra’s 25-point plan to reform federal IT management announced last December, federal agencies must adopt a “cloud-first” policy that requires them to move three applications to the “cloud” over the next 12 to 18 months. Agencies must identify the three “must move” services within three months, move one of those services to the cloud within 12 months and the remaining two within 10 months.

This cloud-first policy and the incremental approach to cloud adoption will make IT reform real and should result in huge (30-50%) savings in federal IT budgets. One specific and measurable goal laid out in the plan calls for a reduction in government data centers from the current 2,094 number to fewer than 800 by 2015. Already 50 percent of government agencies are moving to private clouds but to realize the full potential of the cloud, the government needs to move from many small clouds to fewer large, shared clouds. Of course, federal acquisition policies and authorities must be modified before agencies can fully embrace this strategy. The Federal Risk and Authorization Management Program (FedRAMP) begins to address this issue.

FedRAMP allows joint authorizations and continuous security monitoring services for Government and Commercial cloud computing systems intended for multi-agency use. Joint authorization of cloud providers results in a common security risk model that can be leveraged across the Federal Government. The use of this common security risk model provides a consistent baseline for Cloud based technologies. This common baseline ensures that the benefits of cloud-based technologies are effectively integrated across the various cloud computing solutions currently proposed within the government. The risk model will also enable the government to “approve once, and use often” by ensuring multiple agencies gain the benefit and insight of the FedRAMP’s Authorization and access to service provider’s authorization packages.

There are still a lot of challenges that federal agencies need to work out with the cloud–data sovereignty, privacy and security, funding models, etc–but it is clear that the cloud model will allow government to operate more efficiently and effectively. Nonetheless, there persists the nagging perception that the cloud is inherently unsafe. Government agencies are uncomfortable handing over control of their data to other agencies, vendors or third parties. They are right to be concerned; reported cyber attacks against federal systems increased by 39 percent during the last fiscal year when compared to the year before, says an annual report on agency implementation of the Federal Information Security Management Act (FISMA). The report–posted online last month by the Office of Management and Budget (FY2010 FISMA Report)–finds that Federal agencies reported 41,776 cyber incidents during fiscal 2010. In 2009, agencies reported close to 30,000 incidents.

Despite the grim outlook, we believe the security of the federal enterprise, as well as its functionality, can be significantly enhanced by smartly implementing cloud computing. The following are some key principles that can facilitate this:

• The importance of mission-focused engineering. Private clouds inside the federal enterprise can enhance mission support, but mission-focused engineering should be a first step in this pursuit.

• The continual need for security, including data confidentiality, integrity and availability. All federal computing approaches must be engineered to be in total consonance with IA guidelines to assure federal information, information systems and information infrastructure. Cloud Computing, when engineered right, makes dramatic, positive changes to the mission assurance posture of the federal enterprise. Cloud computing enables stronger end point security and better data protection. It also enables the use of thin clients and the many security benefits they provide. Identity management and encryption remain of critical importance.

• The need for always instantaneously available backup of data in the cloud. Ensured availability under all circumstances is a key benefit of smart cloud computing approaches.

• The continual need for open source and open standards. Most cloud infrastructure today is based on open source (Linux, Solaris, MySQL, Glassfish, Hadoop) and this positive trend will help in net centric approaches. According to the IDC Group, open source software (OSS) is “the most significant, all-encompassing and long-term trend that the software industry has seen since the early 1980′s” Gartner projects that by 2012, 90 percent of the world’s companies will be using open source software. This all indicates open source and open standards should be a key principle for federal cloud computing and other net centric approaches.

• The continual need to evaluate both low barrier to entry and low barrier to exit. As approaches to cloud computing are evaluated, too frequently the cost of exiting an approach is not considered, resulting in lock-in into a capability that may soon be inefficient. Cloud computing capabilities should be adopted that do not result in lock-in.

• The need for open standards. Cloud computing contributions to enhanced functionality for the federal workforce and increase interoperability as the code, API’s and interfaces for cloud computing are secure but are widely published for all participants to interface with. Federal involvement in open source and open standards communities should continue and be accelerated, since increasingly cloud computing open standards are being discussed and designed by open standards bodies like W3C, OASIS, IETF and the Liberty Alliance. Document and other formats used by federal cloud computing activities will be open and available for all authorized users on all devices.

• The need to understand the cost of “private clouds”. For at least the near term, the federal government will remain a provider of “private cloud” capabilities where security dictates ownership levels of control over compute power. This fact means the federal enterprise must continually engineer for change and technology insertion, which underscores the need for low barriers to exist in design criteria.

Regarding security, cloud computing holds the potential to dramatically change the continuous loosing game of continual workstation patching and IT device remediation by reducing the amount of applications on desktops and changing the nature of the desktop device from fat client to thin client. Devices can now have their entire memory and operating system flashed out to the device from private clouds and can have the power of the cloud presented to users as if the user is on an old fashioned desktop. This can be done in a way that never requires IT departments to visit the workstation to patch and configure it. And since all data is stored on private clouds it can be encrypted and access only provided to authorized users. No data can ever be lost when laptops are stolen and no data can ever be lost when desktops are attacked by unauthorized users. Security by well engineered use or cloud computing and thin clients or cloud computing and smart fat clients is dramatically enhanced.

This all leads to a key conclusion for the federal enterprise: as we move forward in cloud computing for support to the mission, the federal enterprise should continue to strengthen formal processes to ensure that lessons learned from both industry and the government’s own successful cloud computing initiatives are continually examined and broadly adopted across the enterprise

Crucial Point associates Dillon Behr, Alex Olesker, Bob Gourley and Chris Barnes contributed to this post.

This post sponsored by the Enterprise CIO Forum and HP.