Through the looking glass
Bolstered by SOAPstation, Actional’s platform helps build, secure, and manage a Web services network
It’s a heady challenge for application management: How to decentralize business processes with Web services?
Difficult enough under traditional circumstances, administering applications with components that span business units or corporate boundaries is proving a stumbling block to broad, real-world adoption of Web services networks. Comprehensive monitoring and management tools for SOAs (service-oriented architectures) have been nearly nonexistent, making it tricky to rein in these interdependent software systems.
Actional is looking to bridge that gap. Its Web services management platform is made up of a centralized command post, the Looking Glass 4.0 server, and two levels of outpost monitoring, Active Agents and SOAPstation 4.0, the latter of which is an XML proxy armed to facilitate Web services management.
The Looking Glass platform sits inside your distributed Web services infrastructure. It tracks and tabulates the fault data and performance metrics reported by in-field agents to provide real-time feedback on the health and availability of your application nodes and network. Additional tools manage the application life cycle, and centralized security and policy planning.
On the downside, we found Looking Glass limited by .Net-only support, and the Active Agents and Looking Glass Console are restricted to the Windows platform. For companies running services on the .Net framework that want to integrate those services with those of business partners running other platforms, SOAPstation becomes the bridge.
Nevertheless, Looking Glass provides a far weightier management capability and toolset for building, securing, and managing a Web services network than its competitors’ offerings. It affords greater control than out-of-house hosted solutions such as those offered by Flamenco Networks, too —and its flexibility certainly beats that of internally instrumented applications.
The platform does its job well: Looking Glass created a useful graphical map of my services infrastructure with clear heartbeat and cumulative availability statistics. Most impressively, the collected results facilitate quick fault trapping and tracing at the operations level, isolating not just the offending node, but the specific transaction. That is a substantial advantage in troubleshooting applications and getting them back online quickly.
The Looking Glass server collects and analyzes information from your services architecture, but its eyes and ears are the Active Agents and SOAPstation server components. Installed on the application servers, these serve as sentries between Web services consumers and your services end-points.
SOAPstation servers can be independently implemented and administered as a stand-alone product, but I feel that the benefits of uniting SOAPstations through a central point of management (as in Looking Glass 4.0) will ultimately make effectively managing large-scale Web services farms easier.
The SOAPstation server improves on Active Agents’ general monitoring capabilities by serving as a Web services proxy, brokering Web services transactions and performing intermediary tasks such as in-flight XSLT (eXtensible Stylesheet Language Transformation), XML message modification, and conversion among different security models. Using SOAPstation, I was able to reduce the intricacy of adapting and routing disparate services and improve processing flexibility.