Blog Host: Ron Schmelzer, Jason Bloomberg - ZapThink READ ENTIRE BIO

This exploding demand for XML traffic capacity is not lost on large chip vendors like Intel, as well as a raft of startups looking to build XML processing capabilities into silicon. As the incumbent vendor, Intel's strategy is to build and promote increasing demand for their chips, and they see SOA as being an important driving force. In fact, Intel has gone so far as to say that corporate networks and the Internet itself might be brought to its knees by SOA-driven XML traffic increases, unless new technologies are brought to bear to solve these problems.

In addition to Goliaths like Intel, startups like Xambala, Tarari, DataPower and Conformative Systems are braving the high capital cost of creating new chips to bring XML performance technology to market. Xambala is building chips that will accelerate several forms of content processing (in addition to XML), while Tarari is also building their own chipset for XML parsing. Datapower delivers a chipset focused on XML parsing and transformation, and Conformative has a unique parallel-processing XML chipset that they are embedding in an XML appliance focused on transformation. As a result, Conformative will be joining the ranks of other XML appliance vendors whose intellectual property primarily rests in software, including Forum Systems, Sarvega and Reactivity, while Xambala and Tarari are looking to take an OEM strategy to the market.

These developments in silicon impact SOA teams in two main ways. First, care must be taken to address any performance or security issues that an SOA rollout introduces, which companies can now address successfully via hardware-only approaches. More importantly, however, is the realization that the technology teams at enterprise clients responsible for SOA initiatives must include network personnel who understand the issues that XML, Web services and SOA introduce to the network. As such, as Web services and SOA continue their inevitable progress towards maturity, so, too, will the solutions that serve those markets, and network-based appliances will surely be in the fore.
Posted by Jason Bloomberg A quick take on the SOA state of the market
04 NOV 2004 06:05 EST (11:05, GMT)
At this point in time, the number of enterprise customers who understand and are implementing the business value of service-oriented architecture (SOA) is increasing dramatically. While the majority of enterprises have Web services implementations, which are one of the foundational building blocks to realizing standards-based SOAs, somewhere in their organization, it has only been during the last six to nine months that IT management and enterprise architecture teams at many enterprises have moved forward with SOA projects. In most cases, such projects are at the planning and design phases or are pilot SOA initiatives. In a few, but notable, cases, ZapThink is seeing widespread and aggressive implementation of SOAs in production systems at some of the largest insurance, financial services, government and health care organizations.

ZapThink is seeing enterprises approach their SOA initiatives in two basic ways. First, many companies are building or have built a relatively small number of mission-critical Web services. By "mission critical," we mean that such services provide sufficient value to the organization that they must be secure, managed and meet certain service-level requirements for availability and scalability. However, companies taking this approach often delay implementing service registries or other approaches to providing location independence for those services, since the services are focused on individual, point-to-point tasks and have not been seen in the context of widespread reuse. As a result, when these organizations begin their SOA planning, they leverage this core set of services that they can subsequently incorporate into their architectural plans.

The second approach to SOA rollout that many enterprises undertake is more top-down, where the efforts begin with an architectural design initiative that focuses on the broader value proposition of SOA, namely providing an agile IT infrastructure that meets changing business requirements in the face of a heterogeneous IT environment. Such design initiatives look to plan the SOA layer of abstraction, often by decomposing key business processes to determine the appropriate services the enterprise must build to maximize the reuse of application functionality across the scope of the SOA initiative.
Posted by Ron Schmelzer Watch out for the SOA hype
03 NOV 2004 05:39 EST (10:39, GMT)
Even though we at ZapThink have been talking about Web services and service-oriented architecture (SOA) for a few years now, we are still quite surprised at how rapidly talk of SOA has come to the lips of vendors and enterprise IT personnel alike. Many organizations are still struggling with how best to use Web services, and now they see discussions of SOA in every magazine and vendor missive. This rush to jump on the SOA bandwagon, however, can be risky -- for several reasons:

• Most end users still don't understand the basics of SOA. SOA is architecture -- in other words, a set of best practices, or a discipline if you will. While you can buy tools that help support an organization's SOA efforts, you can't buy an SOA itself from a vendor, and many end-users don't realize that.

• Many software vendors are too quick to label their products as being part of an SOA solution, without actual evidence of new product functionality or architecture that supports that claim. Many of these vendors presume that if something supports Web services, then it's good for an SOA, as well. Many of these products, however, do nothing to help build or maintain the services abstraction layer at the core of SOA. End users will become disillusioned with SOAs if they don't see any fundamental difference in the products they buy. As a result, vendors should be careful to build and promote SOA products carefully so that they are distinguished from non-SOA products.

• Even when a software product is able to help a company build or run an SOA, there is typically no clear roadmap as to how to accomplish this task. For example, UDDI is a critical piece of a Web services-based SOA, but many companies that buy UDDI-compliant registries don't know how to achieve the location independence that services require in an SOA by using the registry. End users need not only technology, but also guidance in how to apply that technology to realize their SOA goals.

• Fundamentally, SOAs are difficult. They require careful planning, deep change management and buy-in from both business and IT. If companies jump in too quickly they are likely to get disillusioned.

When we talk about user interfaces, we generally talk about clients. A client, in effect, is really the application that resides on an individual's machine that controls the interaction with business logic. In the past, when we spoke about these UI clients, we sorted the kinds of technology approaches into two relative buckets: thick clients and thin clients. However, in the context of standards-based, loosely coupled, distributed computing, the terms "thin" and "thick" are not particularly meaningful in any case. What we really care about is whether or not the consumer of a given client can value-add the service with a rich set of user interactivity capabilities or is limited to a poor set of capabilities. In this vein, we are talking about rich clients that define an increasing set of capabilities that thin, thick and anywhere in between can take advantage of.

In addition to operating systems, desktop applications themselves are increasingly becoming rich clients for Web services. In the past few years, applications such as the Microsoft Office suite, Adobe Acrobat, FileMaker and a wide range of other applications have increasingly added support for querying remote data sources from Web sites, FTP sources and other destinations. The latest versions of these products now add support for retrieving information from Web services-based applications.

While certainly these applications are built for a specific task such as spreadsheet, word processor, document viewing or database development, they are increasingly finding potency as general-purpose rich clients for distributed applications. In particular, people use Microsoft Excel as a general client for business intelligence, analytic and information integration needs. For many users, applications like Excel and Acrobat represent their ultimate client capable of accessing and manipulating any data they need.

Furthermore, new offerings such as Microsoft InfoPath and the latest versions of the Acrobat product blur the lines between document editing activities and interaction with live sources of distributed data. As such, it will be harder to classify the set of desktop productivity applications as being purely desktop, client/server or rich client applications. Finally, people are increasingly using mobile devices and appliances of a wide range of form factors as general-purpose rich clients for distributed applications. In these scenarios, devices such as mobile phones, TiVo appliances and tablet form-factor PCs would all present credible rich clients.

Throw into the mix products like Macromedia's Flex and JackBe's NQ Suite, and the market will definitely heat up.

This ZapFlash is an excerpt of ZapThink's recent report entitled Rich Clients for SOAs. To read more detail about rich clients and the impact they are having on Web services and SOAs, please read the rest of the report.
Posted by Ron Schmelzer Adoption of SOA growing among Indian professional services firms
01 NOV 2004 04:50 EST (09:50, GMT)
ZapThink has recently seen an upsurge in interest in Web services and service-oriented architecture by India-based professional services and systems integration (SI) firms. Firms like Mumbai-based MphasiS and Bangalore-based Infosys are increasingly seeking advisory assistance in building their SOA practices as well as increasing their visibility to U.S. markets for their emerging SOA and Web services consulting services.

At first, it might seem odd that Indian SI firms are first to aggressively add SOA capabilities to their product lines, as their growth has been largely fueled by outsourced IT initiatives that aim to primarily place low-cost developers in place of their more expensive counterparts in routine IT tasks. In essence, why should Indian firms adopt strategies that maximize reuse and minimize the cost of integration when their own costs are significantly lower than those in the United States' or Europe's? If anything, common wisdom might imply that Indian firms should lag behind their foreign brethren in SOA adoption, since the economic imperatives are not as strong.

However, ZapThink has found this situation not to be the case. The move to business process outsourcing (BPO) and offshoring of IT development has actually accelerated the need for Indian firms to consider SOA as a means to simplify development and provide the necessary layer of abstraction to isolate their clients from IT implementation details. In fact, the trends towards SOA and BPO are mutually reinforcing -- the more companies seek to offload their IT capabilities, the more such companies will need to define those capabilities using application-neutral metadata in order to compose applications as needed to meet changing business process needs. As a result, ZapThink expects Indian firms such as MphasiS, Infosys, Wipro, Satyam and Tata Consulting Services to be leading indicators on how SI and professional services firms will adopt SOA methodologies and practices.
Posted by Jason Bloomberg So, what about these new-fangled ESBs?
29 OCT 2004 06:05 EDT (10:05, GMT)
One approach that contributes to an optimal SOA implementation is the use of a so-called "Enterprise Service Bus" (ESB) to provide an infrastructural element to distributed services on the network. The ESB approach towards integration considers systems as discrete services that connect to each other via an asynchronous, message-oriented communications infrastructure. The message-oriented infrastructure allows loosely coupled, document-oriented exchanges between independent systems.

An ESB is the combination of standards-based messaging middleware, distributed service containers that use Web services, XML transformation and rules-based routing of documents. We have also seen ESBs defined as "a new kind of middleware that combines features from several previous types of middleware into one package." Previous types of middleware presumably include message-oriented middleware (MOM), Enterprise Application Integration (EAI) and business-to-business (B2B) integration solutions. ESBs also often provide some value-added services beyond those found in those previous middleware approaches, such as message validation, transformation, content-based routing, security, service discovery for a service-oriented architecture (SOA), load balancing, failover and logging.

Yet, despite these attempts at definition, the term ESB is being used in increasingly more vague and unhelpful ways. For example, Microsoft's Indigo has been described as an ESB. However, Indigo, which forms the basis of Microsoft's next-generation development environment combining SOA as well as advanced code management and code portability capabilities, is not by itself any more an ESB than any other development or runtime environment.

The term ESB is helpful in the short term as companies seek to understand how to move from their proprietary, tightly coupled integration approaches of today to the standards-based, loosely coupled, agile infrastructures of tomorrow, which solve integration problems through architectural approaches. In addition, ESBs represent a viable implementation approach for SOAs that relegate the application server to a role that is one of executing particular synchronous services, while the ESB forms the backbone for asynchronous inter-service communication. While ESBs by themselves don't offer the entire SOA picture that companies will require, they provide a necessary stepping stone.

Read more about ESBs in ZapThink's ZapFlash ESB: Just another TLA?
Posted by Ron Schmelzer Web services for small and medium-sized businesses
28 OCT 2004 05:46 EDT (09:46, GMT)
ZapThink frequently talks about how Web services and SOAs provide significant benefits to enterprises by virtue of their ability to dramatically reduce the cost and complexity of integration and enable business agility. However, much of the short-term promise of Web services applies to companies where the integration cost and complexity is high because their IT infrastructure is heterogeneous -- that is, consisting of a wide range of disparate technologies that are difficult to interoperate. Certainly, a heterogeneous IT environment is the norm in companies of any significant size, but the smaller a company gets, the more homogeneous its IT infrastructure becomes. And so what do Web services and SOAs offer the medium-sized company that has, until now, simplified its IT universe by standardizing on a single platform or environment?

It's still about integration
At its core, every business shares the same elements: products or services that a business sells, customers that buy those products or services, suppliers that provide the components of a product or service and the operations that make producing, buying and selling possible. For large enterprises, each of the above operations can be extremely complex, involving hundreds or thousands of different systems, individuals and processes. However, the smaller a business gets, the more external its integration issues become. Rather than dealing with internal heterogeneous infrastructures, small businesses must integrate with their suppliers', customers', outsourced vendors' and partners' systems. In some industries, such as insurance, healthcare and government, external integration is the primary integration issue since there are a large number of disparate, disconnected parties to integrate.

For such companies, Web services can make just as a significant dent in reducing the cost and complexity of external integration as it can with internal integration for larger companies. The challenge these small businesses face is that they must be able to implement secure, reliable, process-driven Web services across enterprise boundaries while staying within their small company IT budgets.

Looking to others for service
Medium-sized businesses that depend on technology, like their larger peers, can be also become bogged down if their IT infrastructures are not capable of rapidly changing in response to business requirements. Therefore, technology-centric medium-sized businesses have much to gain by implementing process-driven Web services and SOAs that allow them to outsource critical aspects of their operations and make changes to their businesses with minimal IT impact. Web services and SOAs are location independent, which means that the underlying systems can exist anywhere behind or in front of the corporate firewall. By taking advantage of location independence, medium-sized businesses can continue to be agile -- innovating and expanding their markets while reducing their cost of change.
Posted by Jason Bloomberg A not-so-well known spec: WS-Discovery
27 OCT 2004 06:07 EDT (10:07, GMT)
Specs like UDDI are good for when the various Web services participants are all connected to the network, but what do you do about mobile devices, phones and even things like printers that will increasingly take advantage of Web services, which systems aren't aware of when the devices are available or not?

In this vein, Microsoft recently introduced an under-covered spec called WS-Discovery protocol, which is focused on two key things: notifying systems through a multicast protocol when a device or service is available and providing a location "bootstrap" so that UDDI systems and event-driven systems can continue to locate and communicate with the device.

The protocol works by providing a means by which a device can "announce" itself to a local network and then "chat" with interested systems and devices that are looking to communicate with it.

Now WS-Discovery is really meant for limited resource devices and systems and is to be used on local networks, since it's not meant to replace UDDI and other protocols that can be used for discovery of services across the firewall and through other networks. The reason for this is that multicast protocols are very "chatty," which means they can consume a lot of bandwidth as they continue to notify systems that they are available for interaction. As such, Microsoft et. al. are intending this to be used as a way to "bootstrap" a device onto the network and then will leverage other protocols (like WS-Eventing, WS-Addressing, WS-Security and WS-ReliableMessaging) for ongoing system-to-system communication.

Also, this protocol is not meant to replace (at least in the short term) the device-specific Universal Plug-and-Play (UPnP) specification that Microsoft has successfully championed. UPnP will be used first and foremost to let systems know when devices are available for direct communication, while WS-Discovery will be a way to let an entire network know about the services that are available on that device. So, one can imagine a scenario where you plug a network printer into the network, it configures itself on your machine with UPnP and then communicates with other servers on the network using WS-Discovery to let them know that it can print digital color photos.
Posted by Ron Schmelzer 'Tis the season of the merger and acquisition
26 OCT 2004 14:08 EDT (18:08, GMT)
A surprising number of vendors have announced or completed acquisitions or mergers in the Web services/SOA marketplace in the last month or so: Digital Evolution acquired Flamenco Networks; Computer Associates is acquiring Netegrity; Actional and Westbridge Technology have merged; and TIBCO acquired General Interface. Of course, such mergers and acquisitions (M&A) happen all the time, but there does seem to be a marked increase in M&A activity this quarter. Is there a pattern in these events? What does this recent activity say about the Web services and SOA marketplace?

Two of these events -- the Digital Evolution/Flamenco and Actional/Westbridge deals -- are squarely in the Web services management space. This market category is essentially transitional, as the management of Web services, and more broadly SOA implementations, are not truly separate from the management of the rest of the IT infrastructure. Over time then ZapThink expects the incumbent systems management market (IBM Tivoli, CA Unicenter, HP OpenView, etc.) to subsume the Web services/SOA management market in the long term. In the short term, ZapThink has long been predicting consolidation within the space, as market forces drive the vendors to offer more comprehensive solutions.

General Interface, on the other hand, is a member of the relatively nascent rich client marketplace, which includes such startups as Nexaweb, JackBe and Altio and incumbents Macromedia and Adobe. The incumbents are building momentum based on their existing offerings and customer base, but it's still not clear how startups fit into the enterprise picture. Rich clients, after all, provide a richer user experience than browser-based thin clients can and often offer special functionality like offline/mobile support and local data storage. But are these solutions viable standalone products, or are they better positioned as interface tools for middleware suites? TIBCO clearly believes the latter.

The CA/Netegrity acquisition clearly focuses on the security/identity and access management (IAM) markets, but there's an important SOA/Web services story to tell here, as well. CA has been building their Web services management offering since the 2003 acquisition of Adjoin in their Unicenter WSDM product, and Netegrity has thrown their hat into the Web services ring with TransactionMinder. CA is now in the process of moving their broad product line to a service-oriented model with Unicenter WSDM leading the charge, while Netegrity had been struggling to build a market for TransactionMinder. The combination of the two efforts, however, promises a comprehensive SOA/IAM offering that will give the other systems management vendors a run for their money.
Posted by Jason Bloomberg Consolidating Web services security specs?
25 OCT 2004 09:19 EDT (13:19, GMT)
IBM recently announced that they will be supporting the Liberty Alliance specs. At seemingly the same time, Microsoft is pulling back from full-throttle support of Passport -- or at least as far as a global, competing spec from Liberty Alliance is concerned.

Any move towards consolidation of the standards is a good move. IBM took the opportunity of responding to a customer demand for Liberty Alliance adoption, and rather than resist that, they took the initiative and made a bold move to support the standard outright, giving their customers some relief in the area of evolving specs.

I think the challenge still will be in getting the federated identity specs agreed upon so that they can be broadly supported by the WS-I. At some point, one of these specs will win out. Also announced recently is the news that Monster.com dumped Passport recently, so maybe there's a shift in momentum to the Liberty side? In any case, a consolidated stack of security standards for Web Services is great for the industry. Maybe Microsoft is next to support the Liberty Alliance?

Now that WS-I has proved that they can bring the major standards creators together into a cohesive organization to address interoperability, they face their greatest challenge as they approach the more difficult areas of operability: security, management and process. The question is: Can they hold together this group of highly competitive vendors as they start to decide on which security, management and process standards to adopt, when the vendors themselves back different options? If they can, they'll prove their worth even more. If not, it can pose some challenges to WS-I's long-term success. The next year or two will be the real proof years for the organization's long-term value.
Posted by Ron Schmelzer