Testing
The AgentScape Operating System provides a framework to support the development of
large-scale, secure, heterogeneous, distributed multi-agent systems.. To verify and
validate the functionality of the AgentScape Operating System, a test suite
environment has been developed.
The test suite environment tests three types of functionality.
First, primitive functionality of an agent platform are tested and analyzed. This
functionality includes: management of agents, communication between agents on the same
host and between agents on different hosts, migration of agents between agent
platforms. For an agent platform to operate correctly, at least this basic
functionality should be present and functional.
Secondly, the performance of these primitive behaviours is examined. The extent to
which these behaviours scale is tested and visualised.
Thirdly, a set of prototypical multi-agent system implementations test the overal
functionality of an agent system. Implementations of such complex system scenarios are
a valuable tool for testing a system's real-life performance and stability.
The test suite has been designed to be a separate service that can be used to test the
Agent Operating System, but also to test other agent platforms. This makes it possible
to compare the performance of different agent platforms. The test suite has been used
to evaluate a JADE platform. it is currently being transformed into a full-fledged
service for both JADE and AgentScape platforms.
Challenges:
Implementing tests for the current implementation of the AgentScape Operating System.
Comparing results between AgentScape OS and other agent platforms.
