Patrick J. O'Sullivan

Patterns of memory inefficiency

Large applications often suffer from excessive memory consumption. The nature of these heaps, their scale and complex interconnections, makes it difficult to find the low hanging fruit. Techniques relying on dominance or allocation tracking fail to account for sharing, and overwhelm users with small details. More fundamentally, a programmer still needs to know whether high levels of consumption are too high. We present a solution that discovers a small set of high-impact memory problems, by detecting patterns within a heap. Patterns are expressed over a novel ContainerOrContained relation, which overcomes challenges of reuse, delegation, sharing; it induces equivalence classes of objects, based on how they participate in a hierarchy of data structures. We present results on 34 applications, and case studies for nine of these. We demonstrate that eleven patterns cover most memory problems, and that users need inspect only a small number of pattern occurrences to reap large benefits.

Occurrence and Effects of Leader Delegation in Virtual Software Teams

Virtual teams are an important work structure in software development projects. However, little is known about what constitutes effective virtual software team leadership, in particular, the amount of leader delegation that is appropriate in a virtual software-development environment. This study investigates virtual software team leader delegation and explores the impact of delegation strategies on virtual team performance mediated by team motivation, team flexibility and team satisfaction with the team leader. This research is a report of a pilot study run on student teams carried out to refine and test the research constructs and research model for a larger study run in corporations. The study found that virtual team leaders delegate more to competent virtual teams and that such delegation is positively correlated with team member satisfaction with their leader and with team member motivation. Overall, the work provides important information for software-based organizations interested in developing virtual team leadership skills.

Scalable Run-Time Correlation Engine for Monitoring in a Cloud Computing Environment

Monitoring the status of running applications is a real life requirement and important research area. In particular log analysis is often required to understand how the system is behaving during execution. For example it is common for system administrators to collect and view logs from different hardware and software components to gain an understanding into system behavior, especially during activities such as problem determination. A recent research project in this area, the Run Time Correlation Engine (RTCE), provides a framework for run-time correlation of distributed log files in a scalable manner for enterprise applications. The framework has been designed for enterprise applications consisting of distributed software components and is in use in real industry environments. The purpose of this paper is to explore how the RTCE can scale for cloud computing environments where providers of cloud services will require large architectures (e.g. data centers) to deploy such services.

Delegation in Virtual Team: the Moderating Effects of Team Maturity and Team Distance

Virtual teams are becoming an important work structure in software development projects. However, a number of issues arise due to the complexity and newness of the virtual team context. One such issue relates to when and how team leaders should delegate authority and responsibility to the team. Previous studies have yielded conflicting results. This work aims to answer this question about delegation by investigating the moderating effects of team maturity and team distance on the relationship between leader delegation and team outcomes. A research model and specific propositions are presented. This paper provides useful insights for future virtual team leadership research and for organizations interested in developing virtual team leadership

Cultural Differences in Temporal Perceptions and its Application to Running Efficient Global Software Teams

Global software development has been found to be a difficult undertaking, in particular, when members of a single team are not co-located. Studies have looked at the impact of different cultural backgrounds, communication structures and temporal distance on the teams effectiveness. This research proposes to examine the impact of culturally based perceptions of time. A gap analysis is proposed to carry out this examination. The gap that will be measured is the gap between time-based attitudes and behavior in team unit A and team unit B where units A and B are part of the same team but are not co-located. These time-based attitudes and behavior will be compared to measures of team satisfaction and team effectiveness. A model of the impact of the temporal cultural differences and their effect on team performance is presented and the proposed research for testing this model is described

High volumes of event stream indexing and efficient multi-keyword searching for cloud monitoring

In the Cloud computing environment, enterprise applications can produce high volumes of data, which need to be effectively analyzed for administrators to understand system behaviors. Many run-time data analysis tools can give the most up-to-date knowledge of the system to administrators. However, when troubleshooting a problem in depth, the offline data analysis functionality is necessarily required to get the complete knowledge for system diagnosis. In this paper, we propose a composite tree index structure for the Run-Time Correlation Engine framework to achieve efficient event indexing and searching tasks. The framework has been previously proposed as an automatic tool for easing many run-time data analysis tasks. By integrating the data indexing solution to the framework, we are able to further enhance the tool to perform offline data analysis tasks to provide a more sophisticated monitoring service in the Cloud computing environment.

Information "bridging" in a global organization

This paper describes an interview study investigating the collaborative information-seeking and sharing practices of a global software testing team. A site located in Europe was used as a temporal bridge to help in managing time zome differences between the US, China and India, All sites utilized this bridge for critical, synchronous information seeking. Interviews suggest that bridging can be a taxing job and that the success of the bridging arragement depended upon an intricate balance of temporal, infrastructure and cultural factors.

Event Indexing and Searching for High Volumes of Event Streams in the Cloud

Deployed software applications use log files to keep a record of system events. Log analysis provides support for system administrators to gain the knowledge of system health and behavior. As a result, the ability to efficiently search for patterns in historical events has become a major requirement for timely analysis. Enterprise systems today produce high volumes of log data, regularly in the order of thousands of events per second, which requires to build inverted indexes for quick data retrieval. However, current inverted indexing techniques are rarely designed to handle high volumes of dynamic stream data and often resource consuming. We propose an efficient indexing solution, which reduces the necessary resources by employing bloom filter techniques. The solution builds a generic indexing engine for the Run Time Correlation Engine logging framework to achieve efficient monitoring in the Cloud. In particular, our solution is able to deliver significant performance improvement over existing indexing engines.

IBM's smarter care: challenges and strategies

In this paper, we discuss how IBM is working with Lero -- the Irish Software Research Centre and the Ireland Smart Aging Exchange. The proposed project will observe, analyse and coordinate Smarter Care models as they apply to active ageing and healthy living, building on models taken from Connected Health. We consider these as socio-technical models, extending from the older persons home and family into wider society. They can and will include healthcare models, but will also include further elements such as social interaction, exercise, intellectual requirements and personal assistance.

Symptom matching for event streams

Enterprise systems produce a vast amount of logging data. This critical and valuable information must be processed automatically for timely system analysis and recovery. As a result of industry demands, a standard database containing known issues has been introduced - a symptom database. Each symptom consists of a rule pattern and corresponding solutions. Patterns used for symptom identification are encoded as a XPath expression and matched against a stream of events in a standardised WSGI format common base event. The ability of an efficient matching for symptom patterns has been raised as an important requirement by industries. The authors present a real-time symptom identification in a stream of events. The implementation will allow multiple autonomic computing components such as self-monitoring sensors to effectively match known patterns in large datasets in run time. Unlike current state of the art approaches, the proposed solution allows users to define patterns using all the complex XPath functions in addition to standard numeric and Boolean operators. In particular, it was aimed at efficient simultaneous matching of a large set of XPath-based symptom patterns against a high-volume event stream, which is crucial for symptom identification but was not addressed efficiently by currently available XPath-matching engines.