Conformance testing of peer-to-peer systems using message traffic analysis

Peer-to-Peer architectures are used by a large number of distributed systems; however, the challenges such as maintaining a reliable and stable peer-to-peer network can make such networks undesirable for distributed systems. Peer-to-peer architectures are designed to be executed on systems with diverse hardware configurations, distant geographic locations, and varied, unpredictable Internet connectivity that make the software testing process difficult. This research defines a method for conformance testing peer-to-peer content distribution systems called “Method for Conformance Testing by Analyzing Message Activity” (MCTAMA). MCTAMA uses a common representation for describing the behavior of nodes during both design and deployment. ATAMA generates, evaluates and filters test cases that help determine variation between the expected and observed behaviors. The focus on message traffic allows MCTAMA to be used at multiple stages of development and deployment while not being affected by the variations in the operating environment, availability of source code or the capabilities of a monitoring mechanism. As a part of MCTAMA, this research includes a method for combining sequence diagrams to create a description of the expected behavior of nodes in the system.

Name thesis:degree_name

Doctor of Philosophy (PhD)

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Computer Sciences

Grantor

Computer Science

Year dc:date.available

2012

Author dc:creator

• Burris, John Wesley

• distributed systems
• dynamic testing
• peer-to-peer
• Conformance testing
• test case generation

Statement dc:rights

• unrestricted
• Release the entire work immediately for access worldwide.

Identifier

etd-04132012-164527
https://repository.lsu.edu/gradschool_dissertations/436

OAI identifier oai:identifier

oai:repository.lsu.edu:gradschool_dissertations-1435