Architecting, developing and testing for performance of tiered collaboration products

ABSTRACT

The collaboration industry has seen an unimaginable explosion since the basic e-mail clients were introduced decades back. The products are now loaded with powerful features and have fierce competition in the market. To win in the marketplace, the offerings have to meet user expectations and demonstrate a high level of performance. The collaboration products aim at personal productivity, group networking and basic communication flow in the enterprise. They form the nerve-system of an organization, with processes of companies constructed around them, therefore ensuring performance-centric development of such products would be a key to their success. In this paper, we discuss the performance-oriented design, development and testing considerations that can find its application in large-scale multi-tiered collaboration products. We present a methodology termed as "Performance in Each Tier" (PET), which encompasses performance throughout the entire development process. PET concentrates on individual and holistic transactions. The paper describes strategies for dealing with performance issues at early as well as later stages of product development. The approach is applicable to products that have evolved over time with growing customer needs and changing business realities. Finding root causes of performance issues in a large product base and regressing functions for performance fixes, would prove more expensive as compared to prioritizing performance considerations along with feature development. This affirms the Performance Engineering concept that performance shall have priority as the functional features do from the commencement of the product development.

REFERENCES

• 1: Connie U. Smith, Performance engineering for software architectures, Proceedings of the 21st International Computer Software and Applications Conference, p.166, August 11-15, 1997
• 2: Gregory Fox, Performance engineering as a part of the development life cycle for large-scale software systems, Proceedings of the 11th international conference on Software engineering, p.85-94, May 1989, Pittsburgh, Pennsylvania, United States [doi>10.1145/74587.74596]
• 3: Zimran, E.; Butchart, D. Performance Engineering throughout the product life cycle. CompEuro?93. 'Computers in Design, Manufacturing, and Production' Proceedings, 24-27 May 1993 Page(s): 344--349
• 4: De Pasquale, C.J. Measure it, model it, manage it (MI³) ? a Java performance-engineering framework. SoutheastCon, 2005, Proceedings IEEE, 8-10 April 2005, Page(s) 555--560
• 5: Vibhu Saujanya Sharma , Pankaj Jalote , Kishor S. Trivedi, A Performance Engineering Tool for Tiered Software Systems, Proceedings of the 30th Annual International Computer Software and Applications Conference (COMPSAC'06), p.63-70, September 17-21, 2006 [doi>10.1109/COMPSAC.2006.15]
• 6: Java Technology. http://java.sun.com/
• 7: Eclipse Open Source project. http://www.eclipse.org/
• 8: Eclipse test and performance tools project. http://www.eclipse.org/tptp
• 9: IBM Rational PurifyPlus. http://www-306.ibm.com/software/awdtools/purifyplus/win/

INDEX TERMS

Primary Classification:
  C. Computer Systems Organization
  C.4 PERFORMANCE OF SYSTEMS
      Subjects: Modeling techniques

Additional Classification:
  C. Computer Systems Organization
  C.4 PERFORMANCE OF SYSTEMS
      Subjects: Measurement techniques; Reliability, availability, and serviceability

General Terms:
Performance

Keywords:
benchmarking, collaboration, eclipse, groupware, measurement, multi-tier, performance, user-interface