Full Record

New Search | Similar Records

Author
Title Enabling Richer Insight Into Runtime Executions Of Systems
URL
Publication Date
Degree PhD
Discipline/Department Computer Science
Degree Level doctoral
University/Publisher Purdue University
Abstract Systems software of very large scales are being heavily used today in various important scenarios such as online retail, banking, content services, web search and social networks. As the scale of functionality and complexity grows in these software, managing the implementations becomes a considerable challenge for developers, designers and maintainers. Software needs to be constantly monitored and tuned for optimal efficiency and user satisfaction. With large scale, these systems incorporate significant degrees of asynchrony, parallelism and distributed executions, reducing the manageability of software including performance management. Adding to the complexity, developers are under pressure between developing new functionality for customers and maintaining existing programs. This dissertation argues that the manual effort currently required to manage performance of these systems is very high, and can be automated to both reduce the likelihood of problems and quickly fix them once identified. The execution logs from these systems are easily available and provide rich information about the internals at runtime for diagnosis purposes, but the volume of logs is simply too large for today's techniques. Developers hence spend many human hours observing and investigating executions of their systems during development and diagnosis of software, for performance management. This dissertation proposes the application of machine learning techniques to automatically analyze logs from executions, to challenging tasks in different phases of the software lifecycle. It is shown that the careful application of statistical techniques to features extracted from instrumentation, can distill the rich log data into easily comprehensible forms for the developers.
Subjects/Keywords applied sciences; distributed systems; machine learning; system performance; runtime executions; Computer Sciences
Contributors Charles E. Killian; Jennifer L. Neville; Charles E. Killian; Jennifer L. Neville; Dongyan Xu; Ramana R. Kompella; Patrick T. Eugster
Country of Publication us
Record ID oai:docs.lib.purdue.edu:open_access_dissertations-1118
Repository purdue-diss
Date Retrieved
Date Indexed 2019-10-07
Created Date 2013-10-01 07:00:00

Sample Search Hits | Sample Images

…analyzing instrumentation logs using machine learning. These applications range from analysis of raw system logs [4, 14], diagnosing faults by tracking component dependencies [24,25], summarizing execution behavior for profiling and…

…summarization Reduced effort analysis Analysis models Software code Feature creation Machine Learning Developer Figure 1.2.: Framework for automated analysis of runtime instrumentation to aid in system management 1.5.2 PerfDetect: Tracking Performance…

…occurrences and variable values that (i) most diverge across the sets of logs, and (ii) most affect overall system performance. Distalyzer uses machine learning techniques to automatically infer the strongest associations between system

…separation of classes are as follows: • Different versions of the same system • Different requests in the same system • Different implementations of the same protocol • Different nodes in the same run Distalyzer uses machine learning methods to automatically…

…instrumentation to aid in system management . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 2.1 Four-step log comparison process in Distalyzer leading up to a visual interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15…

…satisfaction. With large scale, these systems incorporate significant degrees of asynchrony, parallelism and distributed executions, reducing the manageability of software including performance management. Adding to the complexity, developers are under pressure…

…techniques. Developers hence spend many human hours observing and investigating executions of their systems during development and diagnosis of software, for performance management. This dissertation proposes the application of machine learning techniques to…

…easily comprehensible forms for the developers. 1 1 INTRODUCTION Systems implementations are ubiquitous in today’s computing platforms, from operating systems running on end-user machines to large distributed systems running over thousands of machines…

.