Sara Sprenkle

Dynamic virtual clusters in a grid site manager

This paper presents new mechanisms for dynamic resource management in a cluster manager called Cluster-on-Demand (COD). COD allocates servers from a common pool to multiple virtual clusters (vclusters), with independently configured software environments, name spaces, user access controls, and network storage volumes. We present experiments using the popular Sun GridEngine batch scheduler to demonstrate that dynamic virtual clusters are an enabling abstraction for advanced resource management in computing utilities and grids. In particular, they support dynamic, policy-based cluster sharing between local users and hosted Grid services, resource reservation and adaptive provisioning, scavenging of the idle resources, and dynamic instantiation of Grid services. These goals are achieved in a direct and general way through a new set of fundamental cluster management functions, with minimal impact on the Grid middleware itself.

Automated replay and failure detection for web applications

User-session-based testing of web applications gathers user sessions to create and continually update test suites based on real user input in the field. To support this approach during maintenance and beta testing phases, we have built an automated framework for testing web-based software that focuses on scalability and evolving the test suite automatically as the applications operational profile changes. This paper reports on the automation of the replay and oracle components for web applications, which pose issues beyond those in the equivalent testing steps for traditional, stand-alone applications. Concurrency, nondeterminism, dependence on persistent state and previous user sessions, a complex application infrastructure, and a large number of output formats necessitate developing different replay and oracle comparator operators, which have tradeoffs in fault detection effectiveness, precision of analysis, and efficiency. We have designed, implemented, and evaluated a set of automated replay techniques and oracle comparators for user-session-based testing of web applications. This paper describes the issues, algorithms, heuristics, and an experimental case study with user sessions for two web applications. From our results, we conclude that testers performing user-session-based testing should consider their expectations for program coverage and fault detection when choosing a replay and oracle technique.

An empirical comparison of test suite reduction techniques for user-session-based testing of Web applications

Automated cost-effective test strategies are needed to provide reliable, secure, and usable Web applications. As a software maintainer updates an application, test cases must accurately reflect usage to expose faults that users are most likely to encounter. User-session-based testing is an automated approach to enhancing an initial test suite with real user data, enabling additional testing during maintenance as well as adding test data that represents usage as operational profiles evolve. Test suite reduction techniques are critical to the cost effectiveness of user-session-based testing because a key issue is the cost of collecting, analyzing, and replaying the large number of test cases generated from user-session data. We performed an empirical study comparing the test suite size, program coverage, fault detection capability, and costs of three requirements-based reduction techniques and three variations of concept analysis reduction applied to two Web applications. The statistical analysis of our results indicates that concept analysis-based reduction is a cost-effective alternative to requirements-based approaches.

A case study of automatically creating test suites from web application field data

Creating effective test cases is a difficult problem, especially for web applications. To comprehensively test a web applications functionality, test cases must test complex application state dependencies and concurrent user interactions. Rather than creating test cases manually or from a static model, field data provides an inexpensive alternative to creating such sophisticated test cases. An existing approach to using field data in testing web applications is user-session-based testing. Previous user-session-based testing approaches ignore state dependences from multi-user interactions. In this paper, we propose strategies for leveraging web application field data to automatically create test cases that test various levels of multi-user interaction and state dependencies. Results from out preliminary case study of a publicly deployed web application show that these test case creation mechanisms are a promising testing strategy for web applications.

A Study of Usage-Based Navigation Models and Generated Abstract Test Cases for Web Applications

While web applications expand in usage and complexity, testing demands are growing without corresponding automated support. One promising approach to automatic test generation is statistical model-based testing, where logged user behavior is used to build a usage-based model of web application navigation, from which abstract test cases are generated. Executable test cases are then created by adding parameter values to the abstract test cases. Several researchers have proposed variations of this approach, however, no one has empirically examined the tradeoffs and implications of the different ways to represent user behavior in a navigation model and the characteristics of the automatically generated test cases from different models. We report on our exploratory study of automatically generated abstract test cases and the underlying usage-based navigation models constructed from over 3500 user sessions across five publicly deployed web applications. Our results suggest how web testers can easily tune statistical model-based automatic test case generators for web applications toward generating tests closely related to user behavior or toward new navigations without using large additional test resources.

Integrating customized test requirements with traditional requirements in web application testing

Existing test suite reduction techniques employed for testing web applications have either used traditional program coverage-based requirements or usage-based requirements. In this paper, we explore three different strategies to integrate the use of program coverage-based requirements and usage-based requirements in relation to test suite reduction for web applications. We investigate the use of usage-based test requirements for comparison of test suites that have been reduced based on program coverage-based test requirements. We examine the effectiveness of a test suite reduction process based on a combination of both usage-based and program coverage-based requirements. Finally, we modify a popular test suite reduction algorithm to replace part of its test selection process with selection based on usage-based test requirements. Our case study suggests that integrating program coverage-based and usage-based test requirements has a positive impact on the effectiveness of the resulting test suites.

WebVizOr: A Visualization Tool for Applying Automated Oracles and Analyzing Test Results of Web Applications

Web applications are used extensively for a variety of critical purposes and, therefore, must be reliable. Since Web applications often contain large amounts of code and frequently undergo maintenance, testers need automated tools to execute large numbers of test cases to determine if an application is behaving correctly. Evaluating the voluminous output-typically Web pages full of content-is tedious and error-prone. To ease the difficulty, testers can apply automated oracles, which have tradeoffs in false positives and false negatives. In this paper, we present the design, implementation, and evaluation of WebVizOr, a tool that aids testers byapplying a set of oracles to the output from test cases and highlighting the symptoms of possible faults. Using WebVizOr, a tester can compare the test results from several executions of a test case and can more easily determine if a test case exposes a fault.

Configuring effective navigation models and abstract test cases for web applications by analysing user behaviour

As web applications become more complex and are used more pervasively, testing demands are increasing without corresponding automated support. One promising approach to automatic test generation is statistical model‐based testing, where logged user behaviour is used to build a usage‐based model of web application navigation, from which abstract test cases are generated. Executable test cases are then created by adding parameter values to the abstract test cases. Several researchers have proposed variations of this approach; however, no one has empirically examined the tradeoffs and implications of the different ways to represent user behaviour in a navigation model and the characteristics of the test cases automatically generated from different models. This paper reports on our exploratory study of automatically generated abstract test cases and the underlying usage‐based navigation models constructed from over 19,000 user sessions across five publicly deployed web applications. Our results suggest how web testers can easily configure statistical model‐based automatic test case generators for web applications toward generating tests closely related to user behaviour or toward new navigations without using large additional test resources. Copyright

Reshaping the image of computer science in only fifteen minutes (of class) a week

Low undergraduate enrollments in computer science will not meet the future demand of employers. Some reasons for the low enrollments are computer sciences nerdy image, lack of understanding of the field, and low motivation for learning programming. We propose to change the image of computer science by exposing students to applications of computing and its impact on their lives through reading and discussing recent news articles in 15 minutes of class. We call this component of our courses the Broader Issues in computer science. In this paper, we present our motivation for the Broader Issues component, how we implemented the component in CS0 and CS1 courses, and our evaluation of the component, including student reactions to and impact of the component.

The Ancient Graffiti Project: Geo-Spatial Visualization and Search Tools for Ancient Handwritten Inscriptions

This paper discusses how the Ancient Graffiti Project publishes the digital content of ancient epigraphic material and makes handwritten inscriptions from the first century AD more accessible through the use of geo-referenced, spatial interfaces, interlinked and expanded reference data, and innovative tools that enhance searching. Ancient graffiti are texts, but they are texts that exist within a physical environment. We have designed the Ancient Graffiti Project (AGP) to provide a richer understanding of these handwritten inscriptions from the first century AD in their archaeological context. Interactive maps allow a user to retrieve and analyze all the graffiti in a particular location. These and other tools, from filters that refine searches to brief descriptions and translations to explain the content of these writings, have been designed to reach multiple audiences, including scholars, students, and interested members of the public. We discuss how we have designed AGP from its inception to be integrated within and interoperable with the Epigraphic Database Roma. The spatial and physical context of graffiti also allows us to link with other spatially oriented digital projects on the ancient world, including the ancient gazetteer Pleiades.