Enterprise Test Automation Platform

The Enterprise Test Automation Platform from Arsin, a SemanticSpace company, is an agent-based test framework designed to test the functionality of integration scenarios.

The need for Enterprise Application Integration is becoming more and more vital for firms. The two main reasons for this are - enterprise systems and their implementation in firms is getting to a maturity level where their day to day business is increasingly dependent on them and secondly most firms find value in implementing packaged application (such as SAP as an ERP system and Siebel as a CRM system) versus developing custom solutions to meet their computing needs.

While these best of breed packaged applications, do their function best; business processes are not stove piped within the boundaries of these applications. A customer order, by the time it gets delivered, has to pass through multiple systems, participating in that business process. While human operators can manually “integrate” this process across multiple systems, firms see tremendous ROI in integrating the process across systems.

Testing EAI manually is a daunting task. Automating EAI testing not only makes testing more manageable but also repetitive, thereby making the testing organization more productive at more or less the same cost.

Effecta for EAI (ETAP) has a distributed architecture of software components comprising of a test manager and multiple agents. They communicate with each other using the HTTP or HTTPS protocol. Using HTTP protocol allows the test manager to communicate with the agents across firewalls.

The test manager and agents are implemented as extensions to a web server. Manager communicates with the System under Test (SUT) via the agent using an adapter. Adapter is a component that plugs in to an agent and enables connectivity to an SUT.

The test manager coordinates execution of the test case. When a test case execute command is received, it reads the test case from the test case repository and parses it. After parsing the test case, the test manager notifies different agents and adapters to execute parts of the test case. It keeps the context of the test case alive until the test case is done executing. As the test case executes on the SUT, the adapters capture different events on SUT and send the data associated with them back to the test manager. Test manager evaluates this data to determine whether the test case passed or failed.

Test Manager also has a web based user interface that the tester uses to create, manage, execute and report test cases.

Test agent acts as a conduit between the test manager and the adapter. Test agent manages different commands it is receiving from the test manager and routes them to the adapter. Similarly it accepts responses from the adapter and routes them to the manager. It also queues and caches the data such that the data does not over run.