Cloud computing, like most industry trends, has had a full share of hyperbole. Its promise is heralded, but its...
potential flaws are glossed over. Security is often cited as an inhibitor, but data interoperability and other traits may prove to be trickier issues.
In June 2011, SearchSOA.com presented a full-day virtual trade show addressing cloud computing integration and architecture issues. Included on an expert panel was Melvin Greer, Senior Fellow/Chief Strategist of SOA & Cloud Computing at Lockheed Martin. In the first of a series based on SearchSOA.com's 2011 SOA in Action VTS, cloud computing and SOA strategist Melvin Greer discusses government issues in moving to the cloud platform.
On the heels of the U.S. government's effort to better train enterprise architects has come a somewhat similar mandate to drive cost efficiencies with distributed cloud computing. Greer is playing a central role in this effort. Here is Greer responding to Site Editor Jack Vaughan's question about things one should plan for when approaching a cloud implementation.
SearchSOA.com: Melvin, we titled today's panel ''What Can Go Wrong in the Cloud'' - let's get right to that. The question is - what can go wrong in the cloud?
Melvin Greer: So, of course, every implementation of IT delivery has some concerns, and cloud is no different. Our focus area has been on three primary areas where cloud can go wrong.
The first, and the one that has been identified as a primary inhibitor to cloud adoption has been security. And without belaboring the point I’m sure your audience has been focused on cloud security as well. There is a significant concern on the part of federal government agencies when it comes to the ability to put data in a cloud and whether it’s going to meet the federal security requirements laid out by the [Federal Information Security Management Act] first, and now the 230 security recommendations that are part of the continuous monitoring in the [Federal Risk and Authorization Management Program] activity that’s sponsored by the federal government.
But, once we get beyond security - and, by the way, I do believe we will get beyond it, there is such an intense focus on it on the part of both cloud providers and cloud adapters - we will have no more of a sensitivity to security in the cloud then we have today in the enterprise.
Once we get beyond it we will be focusing on the other areas that are right now my primary concern, which are data interoperability, or cloud interoperability - that is, the ability of multiple clouds [to] operate together - and data portability, meaning the ability to move data from one cloud to another. It’s borne on the premise that hybrid cloud computing is probably going to be in vogue in a very short period of time.
And so because of that we’re going to see a situation where one cloud won’t fit all, we’ll have a number of clouds - public, private, and community - that make up this hybrid cloud model. And so, going forward, what we will expect to see is a need to have multiple clouds integrated and we’ll also need to have the ability to move data from one cloud to another as we go through the missions that we’re trying to build.
So, what can go wrong is, of course, you can have security that is less than what you’re looking for. You can have situations where cloud implementations are basically siloed and inoperable. And you can also have situations where you aren’t sure whether the data you moving from one cloud to another has actually been removed and is no longer there for anyone else to see, or actually has been accurately moved so that all your data has moved with you. So, these are the focus areas that I believe are common across all cloud adopters.