Business systems encompass a range of components, from infrastructure to applications. All of these components must perform appropriately when assembled in the production environmenteven though it is typically not possible to verify the functional and capacity requirements of an assembled system under production conditions.
Individual component testing and composite application testing in isolated QA environments are not sufficient for verifying changes. Due to the distributed nature of SOA, it is seldom possible for QA staff to precisely replicate the production environment in QA test beds. Some servicessuch as those offered by other groups or departments within an enterprise or those provided by external business partnersmay not be available for testing. Sanitized test data may not present all of the challenges of data exchanged in production.
For these reasons, deploying new services or new versions of existing services into production may be a process fraught with uncertainty. To alleviate the uncertainty, operations teams must verify SOA applications independently before rolling them out into live environments.
SOA deployment procedures can be even more challenging when an application depends on services that are hosted by a different group or an external enterprise. Remote services can change without warning and at damaging timessuch as when the applications are saddled with peak customer traffic. Therefore, production SOA applications must be frequently verified to ensure that they're functioning as required.
This leaves Operations personnel, who are responsible for uptime and availability, with less than optimal assurance of the system’s integrity before it’s taken into production. From a runtime perspective, business systems may not have been tested using operationally realistic baselines. From a design-time perspective, the production systems often do not conform to the various designs and models used to describe them. Using traditional tools it is virtually impossible to validate the conformance of business system before deployment.
To solve these issues, AmberPoint has leveraged its industry-leading runtime governance technology to introduce the industry's first SOA validation system.
AmberPoint’s solution for validating helps deployment teams verify the integrity of services and business transactions as they are assembled piecemeal in the pre-production environment. Using AmberPoint, deployment teams can automatically verify the performance and functionality of a changed service against applications that consume it and the services on which it is dependent. Additionally, they can perform "what-if" runs to see the potential impact of management policies (especially security) as they change within the runtime environment. AmberPoint’s approach to business system validation empowers Operations with the tools and information they need to work effectively and enables application managers to ensure that business systems meet delivery goals.
AmberPoint enables deployment teams to validate both the functional behavior of business systems as well as their capacity requirements. AmberPoint leverages its broad visibility into the business system to enable real-time sampling of consumption patterns, end-to-end transactions and traffic volumes. By tailoring these baselines to the requirements of technical and business stakeholders, deployment teams can ensure on an ongoing basis that business system remain within the expected range of functionality and performance, even before those services are pushed into the production environment.
Successful reuse of services depends on adherence to enterprise-adopted standards. AmberPoint ensures that services and business transactions maintain high levels of reuse by ensuring that implementations do not depart from common standards for interoperability. Automated profile checking and reports quickly provide deployment teams with detailed information on the interoperability of services.
To help complete the validation test bed, AmberPoint provides accurate simulation of services, down to actual production traffic and realistic performance characteristics. Service simulation proves extremely useful where a business application interacts with a large number of external services, all of which must be included for a full transaction to be executed. These include business partners’ services as well software provided as a service over the Internet.
AmberPoint integrates and extends tools from leading SOA testing vendors, providing comprehensive end-to-end validation and testing to support the most complex SOA environments. AmberPoint helps to ensure that operations teams don't fly blindly into production SOA. By supplementing messages and data generated by traditional testing tools with actual messages and data observed in production, AmberPoint enables businesses to validate their services across a broad range of applications and scenarios.

