Fiorano Delves Deeper into SOA Integration
Type: Competitive Intelligence Brief
Analyst: S. Willett
Report Date: November 14, 2005
Module: Application Infrastructure
ID: CIR20942 |
Current Perspective: Positive/Neutral
Vendor Importance: Moderate
Market Impact: Low |
|
Summary
Event Summary
November 10, 2005 -- Fiorano Software, Inc., a provider of business process integration and enterprise messaging solutions, announced the general availability of its next-generation SOA platform: Fiorano SOA 2006. Built on Fiorano's enterprise service bus, Fiorano ESB, the SOA 2006 platform incorporates several advanced features, including peer-to-peer messaging, distributed BPEL-compliant business process management, and visual orchestration tools, together with a powerful model for coarse-grained business services.
Analytical Summary
• Current Perspective: Slightly positive on Fiorano’s new SOA 2006 platform, because the firm is adding more features for building business logic in Java to its overall JMS-based ESB offering.
• Vendor Importance: Moderate to Fiorano, since it needed to package its functionality together and add more differentiation through its “business component” architecture in order to gain new customers.
• Market Impact: Low on the integration and Web services market, since the move reinforces the need to offer some linkage between an ESB and application development platforms and for vendors to have a strong JMS offering.
Perspective
Current Perspective: Positive/Neutral
We are taking a slightly positive stance on Fiorano’s release of SOA 2006, a packaging of its entire suite, including its flagship JMS engine (FioranoMQ), its ESB, a BPEL orchestration engine and editor, and its adapter and J2EE-oriented development kit, dubbed Business Components and Adapters 2006. The company is offering a credible platform for SOA-type integration, anchored by its JMS engine and supporting standards such as BPEL and XSLT. Although Fiorano is small (estimated revenues are probably under $12 million), it gained a notable business in JMS at a time when larger vendors had weak JMS products. Since that time, larger vendors have moved into the JMS market, lessening Fiorano’s differentiators. Fiorano has since followed a similar pattern, with an ESB product that has preceded larger vendors’ offerings, and it is following it up with an overall “SOA” product. The company seems to be closely shadowing moves by Sonic, without the marketing, mindshare, or support of a larger organization (in Sonic’s case, Progress).
However, the company does have a few differentiators. For example, it touts its “business components,” essentially a way to build Java-based adapters and some limited business logic fronted by Web services. In this release, it is including a business component development kit with support for JCA, JMS, and JMX, Java libraries, and an “ant-based’ compilation framework. There is support for multi-threaded execution of these business components. Another technical innovation in SOA 2006 is resource sharing among multiple business components that could be distributed. Components can be strung together into composite business components, through the firm’s BPEL tool. These composite business components can be re-usable bits of application/process logic. With 2006, users can build transaction support inside or between business components more easily through the support of ACID transaction capabilities. In general, business components let Fiorano users build adapters into existing backend systems with more logic than some third-party tools. In addition, an updated version of FioranoMQ is supported, with more options for directing JMS routing, and there is more integration with third-party services and adapters.
Fiorano, however, is small and does not seem to have a big marketing effort. It will suffer from viability concerns as the inevitable shakeout occurs in this market. In terms of the SOA 2006 features, these are mainly technical and incremental improvements; there is nothing major here to tip the balance Fiorano’s way. The company is not delving into any new areas such as registries, service management, and service governance. It is unclear what new WS-* standards the firm will support. It should consider differentiating itself in one of these areas, perhaps event management. With business components, it is also duplicating functionality found in application server platforms. This is not all bad, though, but the firm should consider bundling with an open source application server (JBoss or Apache) and more tightly integrating with it.
The firms’ tooling is not yet Eclipse-compliant and it is not clear if this is the direction Fiorano is heading. In addition, it is not clear if it will conform to the JBI and/or open source definition of an ESB. It should make some commitments in this area.
Positives and Concerns
Competitive Positives
• Fiorano released SOA 2006, a packaging of its entire suite, including its flagship JMS engine (FioranoMQ), its ESB, a BPEL orchestration engine and editor, and its adapter and J2EE-oriented development kit, dubbed Business Components and Adapters 2006. The company is offering a credible platform for SOA-type integration, anchored by its JMS engine and supporting standards such as BPEL and XSLT. As the ESB and SOA integration market gains more credibility (due to larger players jumping in), more firms will evaluate Fiorano, making it plausible for the company to break out of its niche markets.
• Fiorano is increasing its support for building some business logic associated with services as well as adapters and Java code. It calls these “business components,” and it is including a business component development kit with support for Java libraries, JCA, JMS, and JMX, and an “ant-based’ compilation framework. There is support for multi-threaded execution of these business components. Another technical innovation in SOA 2006 is resource sharing among multiple business components.
• There is the possibility of building transaction support inside or between business components through the support of ACID transaction capabilities. In fact, components can be strung together into composite business components, through the firm’s BPEL tool. These composite business components can be re-usable bits of application/process logic. In general, this allows Fiorano users to build adapters into existing backend systems with more logic than some third-party tools.
• SOA 2006 embeds Version 8.2 of the firm’s JMS engine and allows more direct tuning of JMS event routing. In general, FioranoMQ is a sound platform for the entire suite, allowing for distributed, highly available message delivery. The company also has an XSLT transformation product.
• The new version includes support for third-party JCA adapters, as well as seamless integration of Web services running on third-party Web servers as part of BPEL processes. There is a JMX management interface. This will allow it to run better with third-party application server platforms. The inclusion of JBoss and/or Geronimo could be a powerful combination.
Competitive Concerns
• Many of the new features in SOA 2006 are rather technical and incremental. The company is not delving into any new areas such as registries, service management, and service governance. It is unclear what new WS-* standards the firm will support.
• With business components, Fiorano seems to be duplicating functionality available on Java application server platforms (including open source platforms) and lower-level programming tools. While it is intrinsically linking business components with its BPEL process product, theoretically, any BPEL product should be able to take advantage of Web services programming models.
• The firms’ tooling is not yet Eclipse-compliant and it is not clear if this is the direction Fiorano is heading. In addition, it is not clear if it will conform to the JBI and/or open source definition of an ESB.
• The firm is small and it will suffer from viability concerns. Its marketing and channel efforts are generally lower than that of competitors such as BEA, IBM, Sonic, and even Cape Clear. In fact, the ESB market will probably have a shakeout eventually.
• The firm advertises an event-driven architecture, but the fact is that Web services are not inherently event-driven, nor are database adapters (at least not from the database side). While WS-Eventing and other common invocation layer standards have been proposed to make a true event-driven architecture, it is unclear how or when Fiorano will support these.
Recommended Actions
Recommended Vendor Actions
• The firm should continue to make improvements to its BPEL orchestration tool, adding more graphical ease-of-programming features and more links to its adapters or business components as well as its JMS engine.
• The firm should bundle the product with JBoss and/or Geronimo and put more links between Fiorano and these platforms.
• Fiorano should attempt to differentiate itself from Sonic, IBM, and others, perhaps through SOA management/governance and registries. It may want to consider an acquisition in this area.
Recommended Competitor Actions
• Competitors in the ESB space (e.g., Sonic, BEA, IBM, Cape Clear, Cordys etc.) should consider BPE for lower level service orchestration due to its intrinsic support for WSDL and its status as an accepted standard. This should be separate from higher-level BPM products, which include human workflow and handle business-level operations.
• ESB vendors without a JMS engine (e.g., Cape Clear, etc.) should attempt to partner with an open source provider to offer this functionality.
• ESB vendors without an application server or Java application development tools should consider bundling an open source offering (e.g., JBoss, Apache AXIS, Geronimo, etc.) into their product and offering plug-ins to leading commercial J2EE servers.
• Integration competitors without an explicit ESB product should consider this as a subset of their integration offerings. The product should be standards-based with the assumption of Web services endpoints, along with transformation and services orchestration.
Recommended End User/Customer Actions
• End users should consider Fiorano’s bundled product, SOA 2006, due to strong JMS capabilities, a framework for services-based integration, BPEL process tools, and the ability to create services out of Java and/or JCA adapters.
• Users in larger organizations should get more details about the global professional services support available from Fiorano.
• In general, users should welcome ESB products, and they should distinguish them from regular integration offerings due to lower cost and strong standards support. Characteristics of an ESB include assumption of service end points, some kind of “bus,” transformation, run times for services, and low-level service orchestration. Optional features should include backend adapters, BPM, and an application server.
|
