Edited transcript of weekly BriefingsDirect[TM] SOA Insights Edition, recorded Jan. 19, 2007.
Listen to the podcast here. If you'd like to learn more about BriefingsDirect B2B informational podcasts, or to become a sponsor of this or other B2B podcasts, contact Dana Gardner at 603-528-2435.
Dana Gardner: Hello, and welcome to the latest BriefingsDirect SOA Insights Edition, Volume 9. This is a weekly discussion and dissection of Services Oriented Architecture (SOA) related news and events with a panel of IT industry analysts. I’m your host and moderator, Dana Gardner, principal analyst at Interarbor Solutions, ZDNet blogger, and Redmond Developer magazine columnist.
This week, our panel of independent IT analysts includes show regular Steve Garone. Steve is an independent analyst, a former program vice president at IDC and the founder of the AlignIT Group. Welcome back, Steve.
Steve Garone: Hi, Dana. It's great to be here again.
Gardner: Also joining us is Joe McKendrick, an independent research consultant and columnist at Database Trends, as well as a blogger at ZDNet and ebizQ. Welcome back to the show, Joe.
Joe McKendrick: Hi, Dana.
Gardner: Next Neil Ward-Dutton, research director at Macehiter Ward-Dutton in the U.K., joins us once again. Hello, Neil.
Neil Ward-Dutton: Hi, Dana, good to be here.
Gardner: Jim Kobielus, principal analyst at Current Analysis, is also making a return visit. Thanks for coming along, Jim.
Jim Kobielus: Hi, everybody.
Gardner: Neil, you had mentioned some interest in discussing tools. We’ve discussed tools a little bit on the show, but not to any great depth. There have been some recent announcements that highlight some of the directions that SOA tools are taking, devoted toward integration, for the most part.
However, some of the tools are also looking more at the development stage of how to create services and then join up services, perhaps in some sort of event processing. Why don’t you tell us a little bit about some of the recent announcements that captured your attention vis-a-vis SOA tools?
Ward-Dutton: Thanks, Dana. This was really sparked by a discussion I had back in December -- and I think some of the other guys here had similar discussions -- with TIBCO Software around the announcement that they were doing for this thing called ActiveMatrix. The reason I thought it was worth discussing was that I was really kind of taken by surprise. It took me a while to really get my head around it, because what TIBCO is doing with ActiveMatrix is shifting beyond its traditional integration focus and providing a rear container for the development and deployment of services, which is subtly different and not what TIBCO has historically done.
It’s much more of a development infrastructure focus than an integration infrastructure focus. That took me by surprise and it took me a while to understand what was happening, because I was so used to expecting TIBCO to talk about integration. What I started thinking about was, "What is the value of something like ActiveMatrix?" Because at first glance, ActiveMatrix appears to be something with JBI, a Java Business Integration implementation, basically a kind of standards-based plug-and-play ESB on steroids. It's probably a crass way of putting it, but you kind of get the idea.
Let’s look at it from the point of view of a development theme. What is required to help those guys get into building high-quality networks of services? There are loads of tools around to help you take existing Java code, or whatever, right-click on it, and create SOAP and WSDL bindings, and so on. But, there are other issues of quality, consistency of interface definitions, and use of schemas -- more leading-edge thinking around using policies, for example. This would involve using policies at design time, and then having those enforced in the runtime infrastructure to do things like manage security automatically and help to manage performance, availability, and so on.
It seems to me that this is the angle they’re coming from, and I haven’t seen very much of that from a lot of the other players in the area. The people who are making most of the noise around SOA are still approaching it from the point of view: "You’ve got all this stuff already, all these assets, and what you’re really doing is service-enabling and then orchestrating those services." So, I just want to throw that out there. It would be really interesting to hear what everyone else thinks. Is what TIBCO is doing useful? Are they out ahead or are there lots of other people doing similar things?
Gardner: TIBCO’s heritage has been in middleware messaging, which then led them into integration, Enterprise Application Integration (EAI), and now they’ve moved more toward a service bus-SOA capability. Just to clarify, this tooling, is it taking advantage of the service bus as a place to instantiate services, production, and management? And is it the service bus that’s key to the fact that they’re now approaching tooling?
Ward-Dutton: That’s how I believe it, except it extends to service bus in two ways. One is into the tooling, if you think about what Microsoft is doing with Windows Communication Framework. From a developer perspective, they’re abstracting a lot of the glop they need to tie code into an ESB, and TIBCO is trying to do something similar to that.
It’s much more declarative. It’s all about annotations and policies you attach to things, rather than code you have to write. On the other side, what was really surprising to me was that, if I understand it right, [TIBCO] are unlike a lot of the other ESB players. They are trying to natively support .NET, so they actually have a .NET container that you can write .NET components in and hook them into the service bus natively. I haven’t really seen that from anywhere else, apart from Microsoft. Of course, they’re .NET only. I think there’s two ways in which they’re moving beyond the basic ESB proposition.
Gardner: So, the question is about ESB as a platform. Is it an integration platform that now has evolved into a development platform for services, a comprehensive place to manage and produce and, in a sense, direct complex service integration capabilities? Steve Garone, is the definition of ESB, now, much larger than it was?
Garone: I think it is. I agree with Neil. When I looked at this announcement, the first thing that popped into my mind was, "This is JBI." When Sun Microsystems talked about JBI back in 2005, this is what they were envisioning, or was part of what they were envisioning. Basically, as a platform, that raises the level of abstraction to where current ESB thinking was already. At the time was confusing users -- and still is -- because they didn’t quite understand how, or why, or when they should use an ESB?
In my opinion, this raises that level of abstraction to eliminate a lot of the work developers have to do in terms of coding to a specific ESB or to a specific integration standard, and lets them focus on developing the code they need to make their applications work. But, I would pull back a little bit from the notion that this is purely, or at a very high percentage, a developer play. To me, this is a logical extension of what companies like TIBCO have done in the past in terms of integration and messaging. However, it does have advantages for developers who need to develop applications that use those capabilities by abstracting out some of the work that they need to do for that integration.
Gardner: How about you, Joe? Do you see this as a natural evolution of ESB? It makes sense for architects and developers and even business analysts to start devoting logic of process to the ESB and let the plumbing take care of itself, vis-à-vis standards and module connectors.
McKendrick: In terms of ESBs, there’s actually quite a raging debate out there about the definition of an ESB, first of all, and what the purpose of an ESB should be. For example, I quote Ann Thomas Manes . . .
Gardner: From Burton Group, right?
McKendrick: Right. She doesn’t see ESB as a solution that a company should ultimately depend on or focus on as mediation. She does seem to lean toward the notion of an ESB on the development side as a platform-versus-mediation system. I've also been watching the work of Todd Biske, he is over at MomentumSI. Todd also questions whether ESBs can take on such multiple roles in the enterprise as an application platform versus a mediation platform. He questions whether you can divide it up that way and sell it to very two distinct markets and groups of professionals within the enterprise.
Gardner: How about you, Jim Kobielus? Do you see the role of ESB getting too watered down? Or, do you see this notion of directing logic to the ESB as a way of managing complexity amid many other parts and services, regardless of their origins, as the proper new direction and definition of ESB?
Kobielus: First of all, this term came into use a few years back, popularized by Gartner and, of course, by Progress Software as a grand unification acronym for a lot of legacy and new and emerging integration approaches. I step back and look at ESB as simply referring to a level backplane that virtualizes the various platform dependencies. It provides an extremely flexible integration fabric that can support any number of integration messaging patterns, and so forth.
That said, looking at what TIBCO has actually done with ActiveMatrix Service Grid, it's very much to the virtualization side of what an ESB is all about, in the sense that you can take any integration logic that you want, develop it to any language, for any container, and then run it in this virtualized service grid.
One of the great things about the ActiveMatrix service grid is that TIBCO is saying you don’t necessarily have to write it in a particular language like Java or C++, but rather you can compose it to the JBI and Service Component Architecture (SCA) specifications. Then, through the magic of ActiveMatrix service grid, it can get compiled down to the various implementation languages. It can then get automatically deployed out to be executed in a very flexible end-to-end ESB fabric provided by TIBCO. That’s an exciting vision. I haven’t seen it demonstrated, but from what they’ve explained, it’s something that sounds like it’s exactly what enterprises are looking for.
It’s a virtualized development environment. It’s a virtualized integration environment. And, really, it’s a virtualized policy management environment for end-to-end ESB lifecycle governance. So, yeah, it is very much an approach for overcoming and taming the server complexity of an SOA in this level backplane. It sounds like it’s the way to go. Essentially, it sounds very similar to what Sonic Software has been doing for some time. But TIBCO is notable, because they’re playing according to open standards that they have helped to catalyze -- especially the SCA specifications.
Gardner: Now, TIBCO isn’t alone in some releases since the first of the year. We recently had webMethods with its Fabric 7.0. Has anyone on the call taken a briefing with webMethods and can you explain what this is and how it relates to this trend on ESB?
Kobielus: I've taken the briefing on Fabric 7.0, and it’s really like TIBCO with ActiveMatrix in many ways. It's a strong development story there and it’s a strong virtualization story. In the case of webMethods Fabric 7.0, you can develop complex-end-to-end integration process logic in a high-level abstraction. In their case, they’re implementing the Business Process Modeling Notation (BPMN) specification. Then, you can, within their tooling, take that BPMN definition, compile it down to implementation languages like BPEL that can then get executed by the process containers or process logic containers within the Fabric 7.0 environment.
It’s a very virtualized ESB/SOA development environment with a strong BPMN angle to it and a very strong metadata infrastructure. WebMethods recently acquired Infravio, and so webMethods is very deep now both on the UDDI registry side and providing the plumbing for a federated metadata infrastructure that’s necessary for truly platform agnostic ESB and SOA applications.
Gardner: And, I believe BEA has come out through its Liquid campaign with the components that amount to a lot of this as well. I'm not sure there are standards in interoperability, based on TIBCO's announcement, but clearly I think they have the same vision. In the past several weeks, we’ve discussed how complexity has been thrown at complexity in SOA, and that’s been one of the complaints, one of the negative aspects.
It seems to me that this move might actually help reduce some of that by, as you point out, virtualizing to the level where an analyst, an architect, a business process-focused individual or team can focus in on this level of process to an ESB, not down to application servers or Java and C++, and take advantage of this abstraction.
Before we move on to our next topic, I want to go back to the panel. Steve Garone, do you see this as a possible way of reducing the complexity being thrown at complexity issue?
Garone: Yes, I do. A lot of it's going to depend on how well this particular offering -- if you're talking about TIBCO or webMethods, but I think we were sort of focusing mostly on TIBCO this morning.
Gardner: I think I’d like to extend to the larger trend. Elements that IBM is doing relates to this. Many of the players are trying to work toward this notion of abstracting up, perhaps using ESB as a platform to do so. Let's leave it on more general level.
Garone: That’s fine a good point. You’re right. IBM is doing some work in this area, and logically so, although they come at this even though they have a lot of integration products. I consider them a platform vendor, which means their viewpoint is a little more about the software stack than a specific integration paradigm.
Gardner: Joe McKendrick, how about you in this notion of simplicity being thrown at complexity? Are we going to retain that? Is this the right direction?
McKendrick: Ah, ha. Well, I actually have fairly close ties with SHARE, the mainframe user group, and put out a weekly newsletter for them. The interesting point about SOA in general is that TIBCO, webMethods and everybody are moving to SOA. They have no choice. They have to begin to subscribe to the standards they agree upon. What else would they do?
When we talk about what was traditionally known as the Enterprise Application Integration (EAI) market, it’s been associated with large-scale, expensive integration projects. What I have seen in the mainframe market is that there is interest in SOA, and there is a lot of experimentation and pilot projects. There are some very clear benefits, but there is also a line of thinking that says, "The application we have running on the mainframe, our CICS application transaction system, works fine. Why do we need to SOA-enable this platform? Why do we need to throw in another layer, an abstraction of service layer, over something that works fine, as-is?"
It may seem archaic or legacy. You may even have green-screen terminals, but it runs. It’s got mainframe power behind it. It’s usually a two-tier type of application. The question organizations have to ask themselves is, Do we really need to add another layer to an operation that runs fine as-is?
Gardner: If they only have isolated operations, and they don’t need to move beyond them, I suppose it's pretty clear for them from cost-benefit analysis to stay with what works. However, it seems that more companies, particularly as they merge and become engaged in partnerships, or as they ally with other organizations and go global, want to bring in more of their assets into a business process-focused benefit. So, that's the larger evolution of where we’re going. It's not islands of individual applications churning away, doing their thing, but associating those islands for a higher productivity benefit.
Kobielus: The notion of what organizations have to examine is right on the money, but I think that’s more of a fundamental issue around SOA in general. I think the question you asked was how does something like this affect the ease with which one can do that, and will it figure into the cost-benefit analysis that an organization does to see if in fact that's the right way to go.
Gardner: Neil, this was your topic. How do you see it? Does this larger notion strike you as moving in a direction of starting to solve this issue of complexity being thrown a complexity? That is to say, there’s not enough clear advantage and reduced risk as an organization for me to embrace SOA. Do you think what you’re seeing now from such organizations as TIBCO and webMethods is ameliorating that concern?
Ward-Dutton: Yes and no. And I think most of my answers on these podcasts end up like that, which is quite a shame. The "no" part of my answer is really the cynical part, which is that, at the end of the day, too much simplicity is bad for business. It’s not really in any vendor’s interest to make things too easy. If you make things too easy, no one’s going to buy any more stuff. And the easiest thing to do, of course, for the company is to say, "You know what? Let’s just put everything on one platform. We’ll throw out everything we’ve got, and rebuild everything from the ground up, using one operating system, one hardware manufacturer, one hardware architecture, and so on."
If the skills problem would go away overnight, that would be fantastic. Of course, it’s not about technology. It’s all of our responsibility to keep reminding everyone that, while this stuff can, in theory, make things simpler, you can’t just consider an end-state. You've got to consider the journey as well, and the complexity and the risk associated with the journey. That’s why so many organizations have difficulties, and that's why the whole world isn't painted Microsoft, IBM, Oracle, or webMethods. We’re in a messy, messy world because the journey is itself a risky thing to do.
So, I think that what's happening with IBM around SCA, and what TIBCO is doing around ActiveMatrix, and what webMethods is doing, have the capability for people with the right skills and the right organizational attributes. They have the ability to create this domain, where change can be made pretty rapidly and in a pretty manageable way. That's much more than just being about technology. It’s actually an organizational, cultural process, an IT process, in terms of how we go about doing things. It's those issues, as well as a matter of buying something from TIBCO. Everything’s bound up together.
Gardner: To pick up on your slightly cynical outlook on vendors who don’t want to make it too simple, they do seem to want to make things simpler from the tooling perspective, as long as that requires the need for their run time, their servers, their infrastructure, and so on.
TIBCO has also recently announced BusinessWorks 5.4, which is a bit more complex-turnkey-platform approach that a very simplified approach to tools might then engender an organization to move into. I guess I see your point, but I do think that the tooling and the simplification is a necessary step for people and process to be the focus and the priority, and that the technology needs to help to bring that about?
Ward-Dutton: You’re absolutely right, Dana, but I think the part of the point you made when you were asking your question a few minutes ago was around do we see less technical communities getting more heavily involved in development work. This is the kind of the mythical use of programming thing I remember from Oracle 4GL and Ingress 4GL. That was going to be user-programming, and, of course, that didn’t happen either. I do see the potential for a domain where it’s easier to change things and it’s more manageable, but I don’t see that suddenly enabling this big shift to business analysts doing the work -- just like we didn’t do with UML or 4GLs.
Gardner: We’re not yet at the silver-bullet level here.
Kobielus: Neil nailed it on the head here. Everybody thinks of simplicity in terms of, "Well, rather than write low-level code, people will draw high-level pictures of the actual business process, not that technical plumbing." And, voila! the infrastructure will make it happen, and will be beautiful and the business analysts will drive it.
Neil alluded to the fact that these high-level business processes, though they can be drawn and developed in BPMN, or using flow charting and all kinds of visual tools, are still ferociously complex. Business process logic is quite complex in it’s own right, and it doesn’t simply get written by the business analyst. Rather, it gets written by teams of business and IT analysts, working hand in hand, in an iterative, painful process to iron out the kinks and then to govern or control changes, over time, to various iterations of these business processes.
This isn’t getting any simpler. In fact, the whole SOA governance -- the development side of the governance process -- is just an ongoing committee exercise of the IT geeks and the business analyst geeks getting together regularly and fighting it out, defining and redefining these complex flow charts.
Gardner: One of the points here is around how the plumbing relates to the process, and so it’s time and experience that ultimately will determine how well this process is defined. As you say, it’s iterative. It’s incremental. No one’s just going to sit there, write up the requirements, and it’s going to happen. But it’s the ability to take iterations and experience in real time and get the technology to keep up with you as you make those improvements that's part of the “promise” of SOA.
McKendrick: The collaboration is messy. You’re dealing with a situation where you’ve got collaboration among primarily two major groups of people who have not really worked a lot together in the past and don’t work that well together now.
Gardner: Well, that probably could be said about most activities from last 150,000 years. All right, moving onto our next topic: IBM came out with its financials this week, we’re talking about the week of January 15, 2007, and once again, they had a strong showing on their software growth. They had 14 percent growth in software revenues, compared to the year-ago period. This would be for the fourth quarter of 2006, and that's compared to the total income growth for the company of 11 percent -- services growing 6 percent, and hardware growing only 3 percent.
So, suddenly, software, which does include a lot at IBM, but certainly a large contribution form WebSphere and middleware and mainframes. Mainframes are still growing, but not great -- 5 percent. Wow. The poster child at IBM is software. Who'd have thunk it? Anybody have a reaction to that?
Ward-Dutton: Of course, one of the things that's been driving IBM software growth has been acquisitions. I know I’m a bit behind the curve on this one, but the FileNet acquisition was due to close in the fourth quarter. If that did happen, then that probably had quite a big impact. I don’t know. Does anyone else know?
Gardner: I guess we’d have to do a bit more fine-tuning to see what contribution the new acquisition’s made on a revenue basis, but the total income growth being a certain percentage and then the software, as a portion of that, I suppose, is the trend. Even so, if they’re buying their way into growth, software is becoming the differentiator in the growth opportunity for IT companies, not hardware, not necessarily even professional services.
That does point out that where companies are investing, where enterprises are investing, and where they're willing to pay for a high-margin and not fall into a commodization pattern, which we might see in hardware, is in software.
Kobielus: Keep in mind, though, in the fourth quarter of 2006, IBM had some major product enhancements. Those happened both in the third and the fourth quarter in the software space, and those were driving much of this revenue growth. In July, they released a DB2 Version 9, formerly code-named Viper, and clearly they were making a lot of sales of new licenses for DB2 V9. Then, in the beginning of the fourth quarter, they released their new Data Integration Suite. That's not so new, but rather enhancements to a variety of point integration tools that they’ve had for a long time, including a lot of these software products that they'd acquired with Ascential.
Gardner: That’s the ETL stuff, right?
Kobielus: Not only that, it's everything, Dana. It’s the ETL, the EII, the metadata, the data quality tools, and the data governance tools. It’s a lot of different things. Of course, they also acquired FileNet during that time. But also in the late third quarter IBM released at least a dozen linked solo-product upgrades. In the late third quarter, they were clearly behind much of the revenue growth, and in the fourth quarter for the software group. In other words, the third and fourth quarters of this past year had announcements that IBM had primed the pump for in terms of the customers’ expectations. And, clearly, there were a lot of pent-up orders in hand from customers who were screaming for those products.
Gardner: So you're saying that this might be a cyclical effect, that we shouldn't interpret the third and fourth quarter software growth as a long-term trend but perhaps as beneficial but nonetheless a "bump in the road" for IBM.
Kobielus: Oh, yeah. Just like Microsoft is finally having a bump, now that it’s got Vista and all those other new products coming downstream. These few quarters are going to be a major bump for Microsoft, just like the last two were a major bump for IBM.
Gardner: Let’s take that emphasis that you have pointed out, and I think is correct, on the issue of data -- the lifecycle of data, and how to free it and expose it to wider uses and productivity in enterprise. IBM has invested quite a bit in that. We just also heard an announcement this week from Hewlett-Packard that they are going to be moving more aggressively into business intelligence (BI) and data warehouse activities, not necessarily trying to sell databases to people, but to show them how to extract, and associate, and make more relevant data that they already have -- a metadata-focused set of announcements. Anyone have reaction to that?
Garone: I don’t know too much about this announcement, but from what I’ve read it seems as if this is largely a services play. HP sees this as a professional services opportunity to work with customers to build these kinds of solutions, and there's certainly demand for it across the board. I’m not so sure this is as much products as it is services.
Kobielus: HP, in the fourth quarter of 2006, acquired a services company in the data warehousing and BI arena called Knightsbridge, and Knightsbridge has been driving HP's foray into the data warehousing market. But, also HP sees that it’s a major hardware vendor, just as Teradata and IBM are, and wants to get into that space. If you look at the growth in data warehousing and BI, these are practically the Number 1 software niches right now.
For HP it’s not so much a software play. They are partnering with a lot of software vendors to provide the various piece parts, such as overall Master Data Management (MDM), data warehousing, and business intelligence product sets. But, very clearly, HP sees this as a services play first and foremost. If you look at IBM, 50 percent of their revenues are now from the global services group, and a lot of the projects they are working on are data warehousing, and master data management, and data integration. HP covets all that.
They want to get into that space, and there’s definitely a lot of room for major powerhouse players like them to get into it. Also, very interestingly, NCR has announced in the past week or so that it’s going to spin off Teradata, which has been operating more or less on an arms-length basis for some time. Teradata has been, without a doubt, the fastest growing product group within NCR for a long time. They're probably Number 1 or a close Number 2 in the data warehousing arena. This whole data warehousing space is so lucrative, and clearly HP has been coveting it for a while. They’ve got a very good competency center in the form of Knightsbridge.
They have got a good platform, this Neoview product that they are just beginning to discuss with the analyst community. I’m trying to get some time on their schedule, because they really haven't made a formal announcement of Neoview. It’s something that’s been trickling out. I’ve taken various informal briefings for the last six months, and they let me in on a few things that they are doing in that regard, but HP has not really formally declared what its product road map is for data warehousing. I expect that will be imminent, because, among other things, there is a trade show in February in Las Vegas, the Data Warehousing Institute, and I’m assuming that they -- just like Teradata and the others -- will have major announcements to share with all of us at that time.
Gardner: Well, thanks for that overview. Anyone else have anything to offer on the role of data warehousing?
McKendrick: Something I always found kind of fascinating is that the purpose and challenges of data warehousing are very much parallel to those of SOA. The goal of data warehousing is to abstract data from various sources or silos across the enterprise and bring it all into one place. And the goal of SOA is to take these siloed applications, abstract them and make them available across the enterprise to users in a single place. The ROI formula interestingly is the same as well.
When you start a data warehouse, you’re pumping in a lot of money. Data warehouses aren't cheap. You need to take a single data source, apply the data warehouse to that, and as that begins to generate some success, you can then expand the warehouse to a second data source, and so forth. It’s very much the same as SOA.
Kobielus: I agree wholeheartedly with that. Data warehouses are a platform for what’s called master data management. That's the term in the data-management arena that refers to a governance infrastructure to maintain control over the master reference data that you run your business on -- be it your customer data, your finance data, your product data, your supply chain data and so forth.
If you look at master data management, it’s very much SOA but in the data management arena. In other words, SOA is a paradigm about sharing and re-using critical corporate resources and governing all that. Well, what's the most critical corporate resource -- just about the most critical that everybody has? It's that gospel, that master reference data, that single version of the truth.
MDM needs data warehousing, and data warehousing very much depends on extremely scalable and reliable and robust platforms. That’s why you have these hardware vendors like HP, IBM, Teradata, and so forth, that are either major players already in data warehousing or realizing that they can take their scalable, parallel processing platforms, position them into this data warehousing and MDM market, and make great forays.
I don’t think HP, though, will become a major software player in its own right. It’s going to rely on third-party partners to provide much of the data integration fabric, much of the BI fabric, and much of the governance tooling that is needed for full blown MDM and data warehousing.
Gardner: Great. I'd like to thank our panel for another BriefingsDirect SOA Insights Edition, Volume 9. Steve Garone, Joe McKendrick, Neil Ward-Dutton, Jim Kobielus and myself, your moderator and host Dana Gardner. Thanks for joining, and come back next week.
If any of our listeners are interested in learning more about BriefingsDirect B2B informational podcasts or to become a sponsor of this or other B2B podcasts, please fill free to contact me, Dana Gardner at 603-528-2435.
Listen to the podcast here.
Transcript of Dana Gardner’s BriefingsDirect SOA Insights Edition, Vol. 9. Copyright Interarbor Solutions, LLC, 2005-2007. All rights reserved.