Home
 > Research and Reports > TEC Blog > Taming the SOA Beast - Part 2

Taming the SOA Beast - Part 2

Written By: Predrag Jakovljevic
Published On: September 23 2008

Part 1 of this blog topic introduced the notion of how complex and tricky it can be to manage and govern enterprise applications' service oriented architecture (SOA). That blog post also tackled Progress Software’s recent acquisition of Mindreef in order to round out its SOA governance solution for distributed information technology (IT) environments.

Mindreef joined the Progress Actional SOA Management product family that provides policy-based visibility, security, and control for services, middleware, and business processes. This acquisition continues Progress’ expansion of its burgeoning SOA portfolio and strengthens the company’s position as a leader in independent, standards-based, heterogeneous, distributed SOA enterprise infrastructures.

Prior to being acquired, Mindreef decoupled some plug-in features from its previously all-in-one SOAPscope Server suite.

One capability was SOAPscope Policy Rules Manager that tests compliance with rules such as whether the Simple Object Access Protocol (SOAP) and Web Services Description Language (WSDL) headers comply with the WS-I Basic Profile for Web services interoperability. Also, the feature checks whether the extensible markup language (XML) schema was formed properly, and whether the “contracts” between Web services are valid so that companies can ensure they won't break at run-time because of faulty logic.

Another plug-in, called Load Check, provides a pre-test simulation of the system's performance. The underlying idea was to mitigate the bad practice that, when developing Web services-based applications, the load or performance testing tends to be an afterthought that is often compensated for by purchasing extra hardware after the fact and at a hefty price.

Progress Actional + Mindreef

Like its parent, Mindreef has always designed its products as a good fit for third-party IT governance solutions, with the ability to check on whether Web services are well formed and remain consistent with business policies.

Progress does not release the number of customers it has for specific products or as a corporation, although it admits to gaining access to more than 3,000 of Mindreef's customers at more than 1,200 organizations worldwide. The ideal customers for the combination of Progress Actional and Mindreef SOAPscope are those seeking full life-cycle quality management of their SOA environments, ranging from design through operational deployment.

Mindreef SOAPscope is a recognized testing and validation software product for SOA services at the design stage, while Actional is the market leading SOA management, validation and monitoring software for operational SOA. Thus, the combination of the two provides a solution that is likely to be the first in the market to address the entire SOA lifecycle with SOA quality, validation, and runtime governance.

Progress Actional and Mindreef provide a deep level of SOA management, testing, validation and run-time governance functionality, but not all organizations that have begun implementing SOA environments recognize the need to implement that functionality as yet.  As a result, those companies that have felt the significant pain of having to diagnose why SOA composite applications have failed in order to get them rapidly back up and running, or who have discovered rogue Web services within their environments into which they have no visibility, should see the benefit of deploying Progress Actional and Mindreef.

Progress Actional and Mindreef are sold worldwide from offices in North America, Latin America, Europe, and Asia. A complete list of Progress Software offices is available here.

While hardly any player in the market currently has equal lifecycle SOA quality capabilities as the combination of Actional and Mindreef provides, traditional competitors for Actional include Amberpoint, SOA Software, IBM, Hewlett-Packard (HP), Layer 7 Technologies  and Computer Associates (CA).

As for Mindreef, while it can also be hard to find a single product that functionally competes head to head with SOAPScope, some other vendors’ functionality is comparable to that found in SOAPScope. Namely, in sales situations, Mindreef sometimes runs across IBM Rational Software  and HP/Mercury, and occasionally some of the smaller niche players like Parasoft Solutions, iTKO LISAPushToTest, and Crosscheck Networks.

Forget Not about Oracle Fusion Either

The recent acquisition of the former middleware competitor, BEA Systems, has promoted Oracle into the middleware market leader, at least in the Java world. The idea behind the ambitiously broad Oracle Fusion Middleware (OFM) suite is the following:

  • to enable the enterprise applications’ architecture shift to SOA

  • to become a comprehensive platform for developing and deploying service-oriented enterprise applications

  • to form the foundation for modernizing and integrating the burgeoning Oracle Applications portfolio


Oracle’s middleware product strategy is foremost to provide a complete (unified) and pre-integrated middleware suite that is also modular, standards-based, open, and thus “hot pluggable.” Furthermore, the strategy is to develop and deploy enterprise applications on the Internet via unifying SOA Management, business process management (BPM), business intelligence (BI), enterprise content management (ECM), and enterprise 2.0 capabilities.

The third part of the strategy, the lowest total cost of ownership (TCO), by managing systems, applications, and user identities on low cost hardware and storage systems, has been too overplayed by virtually all vendors to really ring as differentiating, but is certainly a worthwhile attempt by Oracle.

Asserting SOA Governance Competitiveness

As for the product strategy for the Oracle SOA Governance suite, as a subset of OFM, it starts with offering an integrated and complete lifecycle SOA governance platform entailing tools, service registry and repository, policy manager, monitoring console, and so on.

Additionally, the goal is to enable visibility into an organization’s service portfolio via the ability to discover, categorize, manage change, audit usage, and monitor Web services. Last but not least, as discussed in Part 1, the ultimate goal is to provide better control over the lifecycle of services by enforcing policy compliance from software development to operations.

But what really impressed me post-acquisition was Oracle’s due diligence and even (atypical) humility in admitting BEA’s advantages (e.g., in terms of Enterprise System Bus [ESB] and service mediation capabilities) and bundling it with Oracle’s established capabilities of workflow management and Web services orchestration. Other specific areas where BEA had superior technologies were Java virtual machines, transaction processing monitors and certain security products. Conversely, Oracle has products like BI, ECM and identity management, where BEA did not have products.

Accordingly, Oracle has stratified the combined Oracle and BEA middleware products into the following three groups:

  1. Strategic products -- BEA products that are being adopted immediately with limited re-design into OFM, since no corresponding Oracle products exist in a majority of cases. Where corresponding Oracle products exist, they will converge with BEA products with rapid integration over next 12-18 months;

  2. Continued and converged products – BEA products that are being incrementally re-designed to integrate with OFM. There is gradual integration with existing OFM technology to broaden features with automated upgrades. Oracle hereby grants continued development and maintenance for at least nine years; and

  3. Maintenance (a.k.a., “stabilized”) products – those products that even former independent BEA had marked as the end-of-life (EOL) ones due to limited adoption prior to Oracle’s acquisition. Oracle hereby promises continued maintenance with appropriate fixes for five years.


Translating this into the product offerings for Oracle SOA Governance, most of the Oracle and BEA products will end up in the strategic category, starting with BEA AquaLogic Enterprise Repository at the core. It is a repository to capture, share, and change manage SOA artifacts across the lifecycle, with capabilities like audit trail and metrics, service level agreement (SLAs) and policies management, rules and standards definition, WSDL and XML Schema Definition (XSD) schemas, capturing and modeling business requirements, and dependency management.

For its part, Oracle offers Oracle Service Registry, which is a standards-based Universal Description Discovery and Integration (UDDI) v3.0 registry to publish and discover Web services. Furthermore, Oracle Web Services Manager is a policy manager to define and manage security, auditing, and the quality of services (QoS) policies on Web services.

Moreover, Oracle Enterprise Manager (EM) takes care of services deployment, staging and approval, change management, retirement and removal. To that end, the Oracle EM Service Level management pack is a management console to monitor service level response times and availability, while the Oracle EM SOA management pack is a management console to monitor, trace and change manage SOA.

The above-mentioned strategic Oracle SOA Governance suite can be bolstered optionally with the business process analysis (BPA), design and modeling capabilities via the partnering IDS Scheer's ARIS tool, and Oracle’s JDeveloper or open-source Eclipse tool. The latter two tools can also be leveraged for Web services implementation purposes.

The only product slated for maintenance is BEA AquaLogic Services Manager, which is an embedded product (i.e., AmberPoint) and is deemed redundant with Oracle EM.

At this stage, Oracle cannot really compete with Progress Actional in highly distributed and diverse IT environments, and thus must continue to develop and market its slated SOA governance technology as heterogeneous. Based on the availability of Oracle database for Microsoft Windows, I could imagine the similar decoupling of OMF products to also include non-Oracle/Java technologies in the future.

It remains to be seen how the impending acquisition of ClearApp and its planned rolling into Oracle EM will help in this regard. In addition to not limiting the sale of middleware to Oracle or BEA Java-oriented environments, the giant can gain further credibility for interoperability in this market by expanding its SOA governance partner ecosystem.

For its part, Actional's openness and scalability were recently given an endorsement by another SOA powerhouse. Namely, Software AG, announced in early September that it will embed, as an original manufacturing product (OEM), Progress Actional and re-name it as WebMethods Insight. This means that Software AG will now be selling Progress Actional to its own substantial SOA customer base. This is a huge opportunity for Progress, which it did not have prior to this partnership.

What are your thoughts in this regard, and your experiences with the SOA management solutions like Progress Actional/Mindreef and Oracle/BEA? Again, your comments, thoughts, suggestions or individual experiences with SOA governance and testing tools are more than welcome.
 
comments powered by Disqus

Recent Searches
Others A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

©2014 Technology Evaluation Centers Inc. All rights reserved.