We use cookies and other similar technologies (Cookies) to enhance your experience and to provide you with relevant content and ads. By using our website, you are agreeing to the use of Cookies. You can change your settings at any time. Cookie Policy.

ebizQ's Business Agility Watch


Cloud Will Drive SOA in the Future: Mike Kavis Explains

Vote 0 Votes

What follows is my podcast with Mike Kavis, CTO of M-Dot, a retail technology company. Mike discusses SOA and the cloud and the advantages and difficulties enterprises face moving from SOA to the Cloud. This will also be covered extensively at our SOA in Action Virtual Conference.

Listen to or download the 7:18 podcast below:

Download file


PS: First of all, what are some of the advantages a company already practicing SOA can be gained by moving into the cloud?

MK: Well, the biggest advantage is the loosely coupled nature of SOA. So companies who aren't practicing Service Oriented Architecture will be tightly coupled to their databases, will be tightly coupled to their infrastructure. So it's very hard for them to move, or shift, or change things around whereas companies with a Service Oriented Architecture can look at their entire offering and say, hey, these pieces make sense to move the cloud, these other pieces don't and they can make those moves. Without a Service Oriented Architecture, it almost becomes an all or nothing proposition and that's not a recipe for success.

That makes sense. Now, what are some of the main difficulties a company faces already practicing SOA when they're moving into the cloud?

Well, there are a lot of difficulties and a lot of it comes down to hype and the understanding of the cloud. The definition of cloud computing and especially a definition of private cloud is SOA under dispute that people can kind of miss the message and miss the value of the cloud. So what is really important is that the people on these projects are really doing their due diligence, they're understanding what all the different offerings in the cloud are, they understand the difference between Infrastructure-as-a-Service, Platform-as-a-Service, Software-as-a-Service, those types of things.

They really understand their business requirements and then they kind of marry those two together and deliver something in the cloud that can be secure and compliant. It's a lot harder than it sounds but really comes down, you really have to do your homework, you have to valuate these vendors because some of the vendors, their product offering was born for cloud computing. Others, they've taken existing applications and call them cloud enabled, and you really got to be careful not just to chase your favorite vendor and their cloud but find the right solution for your requirements in the cloud.

A lot of companies looking to the cloud are a little bit afraid of the security issue. Would you say the cloud is less or more secure?

Well, I think this is the biggest myth in cloud computing is that the cloud is not secure. And there will probably be a lot of people who won't agree with me on this. But I'll go back to a conversation that I had when I was sitting in an architecture conference early in the year and a well-known analyst from one of the well-known firms was talking about cloud computing and was basically saying how he doesn't understand how people can move critical data into the cloud.

And his comment was along the line, you know, when you buy Amazon you're not buying security. And so I raised my hand and I said, well, when you buy a rack of servers from IBM, you're not buying security either. So you have to build security, you have to compliance; you have to architect for those things. And I'll also go back to it -- and I'll use Amazon as an example again. That is world-class infrastructure.

This is a company that spends millions and millions in dollars hardening a proven architecture, a proven infrastructure that delivers millions and millions of shopping trips to one of the biggest e-commerce sites in the world. And a start-up like me or a small/medium business like me or even a large company is getting that world-class infrastructure. The difference is now you are launching applications outside your firewall. So that doesn't mean the cloud is less secure, it just means you have more requirements for building applications security than you may have had before.

With a lot of organizations looking to leverage the cloud, we've been basically talking about a lot of the technical issues, but what are some of the non-technical issues that companies need to consider?

Oh, the big thing is resistance to change. So I always go back to organizational change management. I talked a lot of about that when talking about Service Oriented Architecture. The same type of challenge is -- you're asking an organization to do something radically different than they've done in the past. And you have to answer for a lot of people. What's in it for me? Why are we doing this? It requires more rigor, more adherence to good sound architectural practices. It requires a lot of people to not be working the silos, a lot people to work together.

So this type of thing is change, especially for the people who maintain your data centers. A lot of them could see this as a threat. So a lot of companies will focus a lot on the technology even if they figure all that out, if they have an army of people who are fighting this thing tooth and nail, they're going to have a very hard time implementing anything so you really have to evaluate your culture. Can my culture sustain this change? Training. Are there people in here that are so against it that I need to get rid of them? We really have to evaluate your staff and the gaps and address them.

Now what do you see for the future of the cloud and SOA?

Yeah, I see a gradual acceptance of SOA enabled cloud services in the future. One of the big challenges right now, we all know that SOA has had its troubles catching on and I think the movement to the cloud is one of the best things that could happen to SOA because what you'll find is its very hard to move to the cloud when you're tightly coupled to your architecture. And then I also see a lot more requirements for businesses to integrate with partners, to leverage mashups, to do those types of things, connect into other Software-as-a-Service providers.

And if you don't take a loosely coupled approach to that, it's going to be hard. So I think the cloud's going to drive SOA, and what's going to drive the cloud is success stories, and that's what's kind of lacking now. There's a lot of little success stories for people doing ad hoc work here and there but we need those enterprise success stories and those are going to come because the cloud computing vendors are getting better, the technology's getting better, there's groups working on standards.

Its' all going to come, but it's going to be slow and early on its going to be the small/medium companies, the start-ups who out of necessity have to use the benefits of the elasticity and cheap computing resources where the bigger companies it's going to be harder because they the huge investments in data centers. They have huge amounts of people they have to get to shift from one mode of thinking to the other. I also think a lot of the work the government's doing, this administration's investing a lot in this area so there's a lot of research going into that area and eventually they'll be some success stories in that area as well and that's really what's going to move this.

I think ten years from now this will be -- today's on-premise will be like yesterday's 8-track player. It'll be common about ten years from now.

Thanks much, Mike. Everyone listening, if you're interesting in the future of SOA and of course the cloud, I urge you to attend the SOA in Action Virtual Conference where Mike will be one of the speakers.


It appears I quite disagree with Mike Kavis.

First, organisations hesitate to go with Cloud Computing absolutely not because of the “resistance to change� but resistance to the reckless change. Companies do not go with public Cloud for production because of non-mitigated business risks. Technology may be great but if it gets disconnected from the business, the business value of such technology is minimal.

Second, “Cloud Will Drive SOA in the Future� might be true not for SOA but for the technical SOA implementation. In the future, SOA will be driven by Business and Cloud will be (and is now) the same – the HOW part of the business task implementation.

SOA is not about technology any more (though it started as technical initiative and must be proud of this), it is about the style of conducting Business.


Good comments,

Resistance to change is one of many reasons. There must also be valid business reasons to leverage cloud computing resources. What I often see is a strong resistance to cloud due to concerns about security and compliance based on word of mouth rather than due diligence. In other words, I hear the cloud "can't be secure", "can't be compliant", or "we have been doing this for years" which are all generalized statements and/or myths. That is why the enterprise success stories that we will start seeing and hearing about over the next several years will get many companies past their fears.

I agree that SOA will or at least should be driven by the business. I also believe that companies that start using the cloud as a platform to deliver applications and services will be more likely to take service-oriented approach because they may only want to put certain pieces of their architecture in the cloud, not everything.

I never intended to imply that SOA was about technology. If you get a chance to read any of my posts on CIO.com or on my blog you will see that I am a big believer that SOA is all about business optimization.

Software has always driven everything in technology. Think VisiCalc, Lotus 123, Excel, desktop publishing/MAC, etc. The star of any technology is the user, not the hardware or infrastructure. This has always has been true and will of course be the case with the cloud. The application running in the cloud will make or break the cloud, not the cloud itself. For example, we saw a shift from Software as a Service (SaaS) to Platform as a Service (PaaS (e.g., force.com) and now we see a shift from PaaS to Open Platform as a Service (OPaaS) www.openplatformasaservice.com The cloud will be another choice, but the user layer on top of the cloud will make or break the underlying technology.

ebizQ’s expert blog team covers a broad range of BPM, business integration, business analytics/monitoring, collaboration, content and related issues.

Peter Schooff

Peter Schooff is Contributing Editor at ebizQ, and manager of the ebizQ Forum. Contact him at pschooff@techtarget.com

Recently Commented On

Monthly Archives