Becoming responsive to service usage and performance changes by applying service feedback metrics to software maintenance

Abstract

Software vendors are unaware of how their software performs in the field. They do not know what parts of their software are used and appreciated most and have little knowledge about the behavior of the software and its environment. In this paper we present a metrics-based approach that is used by software vendors to create real-time usage reports, based on data gathered by leveraging aspect-oriented programming techniques. This approach enables software vendors to respond quickly to performance and usage changes in their service software, both at specific customers and concerning the service software in general. We show that by using this approach, vendors can make informed decisions with respect to software requirements management and maintenance. The metrics and usage reports are validated by way of a case study at a Dutch software vendor. While validation shows high potential of the approach, a successful implementation will require change management at the software vendor.

Service Level Achievements -- Distributed Knowledge for Optimal Service Selection

“…The first approach [18] aims at the support of the software life-cycle process. By means of aspect-oriented programming, a service provider can observe quality aspects of a software component such as reliability, throughput, latency (response time), accuracy, availability, and usability.…”

A Reference Framework for Utilization of Software Operation Knowledge

“…For example, software measurement, monitoring and feedback techniques have been proposed [5]- [7] and software operation data acquisition techniques and tools have been developed [2], [8]. Little research has been initiated to incorporate all processes in one framework, however.…”

“…Like logging, software operation data acquisition can be considered as a typical cross-cutting concern and can thus be implemented by using aspect-oriented programming (AOP) techniques [8]. Note that the amount of software operation data that is acquired depends on both an end-user's software usage behavior as well as the type of software operation knowledge k ∈ κ that is eventually extracted from the operation data.…”

Reducing Maintenance Effort through Software Operation Knowledge: An Eclectic Empirical Evaluation

“…Aspect-oriented programming (AOP) is effectively deployed in the domains of software monitoring and tracing [9], [15]. However, we encounter in industry that product software vendors have to overcome time-consuming obstacles when they leverage AOP repeatedly and separately for each of their software products: vendors write product-specific SOK acquisition aspects and extend their AOP libraries to new (versions of the same) software products.…”