Skilling up the cloud: What it means for infosecurity pros
In its 2011 Global Information Security Workforce Study, Frost and Sullivan argues that cloud computing “illustrates a serious gap between technology implementation and the skills necessary to provide security”.
The analyst firm’s survey of more than 10,000 information security professionals worldwide found widespread use of cloud technologies, especially private clouds and software as a service, yet 70 per cent of respondents said they need new skills to properly secure these systems.
Vladimir Jirasek, a spokesman for the freshly minted UK and Ireland arm of the Cloud Security Alliance, says it would be easy to say that F&S is wrong; that no new skills are needed. But the devil is always in the details.
“Cloud is a different type of outsourcing, if you like it is outsourcing on steroids. You are giving away the responsibility for running a service. You might not always know where your data is. So yes, new skills are needed,” Jirasek says.
These new skills are both technical and non-technical. On the non technical side, it is all about information risk management: information security pros must work out how to map the standards of the cloud provider to their own requirements.
“You need to look at your requirements in terms of certification, policies, guidelines and check that they are being met by the provider. If you have 500 requirements, it is worth checking to see if they all need to apply to the cloud service provider,” Jirasek notes.
Martin Ferguson, head of policy at the UK’s Society of IT Managers (SOCITM), agrees that new skills are needed, (“In a nutshell? Strategic commissioning, data management, risk assessment and risk management”), but says there are great opportunities as well
“There should be greater scope for interoperable systems built around common standards and APIs (application product interfaces), but these still need to adhere to the same data protection and privacy regimes as non-cloud systems,” he says.
According to Ferguson, the bigger change is that cloud computing means information managers have to surrender a degree of control and visibility, especially regarding security and privacy of data, and compliance. Major areas of concern are:
- Secure storage, processing and disclosure to third parties
- identity and authentication of users
- application security; privacy for critical, personal data
- physical security of hardware, software and networks
- personnel management
- audit trails for different compliance regimes
- business continuity and data recovery; maintaining electronic records
Jirasek agrees that this is a potentially tense area: “You can’t audit Amazon. You can’t show up at Google’s offices and ask to see their data centres. This is the old model, and it doesn’t work anymore,” he says.
Communication and a willingness to compromise are key to managing this transition well - useful skills for dealing with the technical aspects, too.
“They [IT security professionals] need to have a good discussion with the security architects at the cloud provider; need to understand the technology to ensure their own requirements are fulfilled,” Jirasek adds.
"Cloud is outsourcing on steroids"
A lot will depend on how much of your technology stack you are going to be handing over to your cloud provider. While Amazon is mainly offering infrastructure as a service, Google and Microsoft can take over the whole stack. In those cases, Jirasek says, it is vital that clients ask how their data will be segregated, and authentication be managed?
“Cloud providers are finding that the best way of providing a seamless service is by using identity federation. This means that once you log on to your own corporate network, you are pretty much authenticated to use services that are run in the cloud
“This is increasingly the way things are done. And information security professionals need to get to grips with it,” Jirasek says.
Ferguson is a little more optimistic, pointing out that security may be easier if it means that applications and services are provided in a “bubble” on the client device.
“[This includes] strong passwords, a move towards two factor authentication and tight lock down and encryption of the client device. Then there are skills related to data management. These skills are not new, but new people need to get a handle on them. Security people have to start talking to business people about how they classify information, how they manage the information security lifecycle.
“This is very hard. Business people typically don’t think of themselves as owning the data, so it will be down to the security people to translate the business requirements into technical protocols, and then analyse the various cloud offerings and see which one is the best fit."
Jos Creese, president of SOCITM, offers the following caution: “It’s too easy to play safe and allow security to 'trump' everything. When imbalance happens, security is, ironically, weakened. People share passwords tacitly, use their own devices anyway, or simply don't allow data to be shared electronically with the consequences that follow. The latter results in more paper (less secure) or people getting hurt because key data is not shared between agencies.” ®
Sponsored: DevOps and continuous delivery