DYNAMIC DERIVATION OF ANALYTICAL PERFORMANCE MODELS IN AUTONOMIC SYSTEMS

Simple item page
dc.contributor.advisorMenasce, Daniel A
dc.contributor.authorAwad, Mahmoud
dc.creatorAwad, Mahmoud
dc.date.accessioned2018-10-22T01:21:16Z
dc.date.available2018-10-22T01:21:16Z
dc.date.issued2017
dc.description.abstractDeriving analytical performance models requires detailed knowledge of the architecture and behavior of the computer system being modeled as well as modeling skills. In autonomic computing environments, this detailed knowledge may not be readily available (or it may be impractical to gather) given the dynamic nature of these environments. This dissertation presents a framework, called iModel, for dynamically deriving and parameterizing analytical performance models in autonomic systems. Analytical performance models consist of a workload model and a system model. iModel uses system logs and configuration files to generate a high level characterization of the system; e.g., open queuing network (QN) model versus closed QN model. By harvesting more information from the system logs and configuration files, iModel generates a workload model by inferring user-system interaction patterns in the form of a Customer Behavior Model Graph (CBMG) and generates a system model by discovering system components and their interaction patterns in the form of a Client-Server Interaction Diagram (CSID). iModel includes a library of well-known QN models stored in an XML-based repository. The generated workload model and system model are compared to the model repository to determine which model in the repository best matches the system's observable behavior and architecture. The best-fit model is then parameterized and solved and its output metrics are compared to the system's measured output metrics to determine the model's accuracy. iModel takes into consideration the balance between producing costly but accurate performance models versus sufficient levels of model accuracy even when detailed system logs are available. To that effect, this dissertation presents a black-box optimization approach that is used to derive analytical model parameters by observing only the input-output relationships of a real system. The important question is whether the dynamically generated and parameterized performance model has predictive power, i.e., can the derived model predict the output values that would be observed in the real system for different values of the input? The results presented in this dissertation demonstrate that the analytical performance models derived by iModel are relatively robust and have predictive power over a wide range of input values.
dc.format.extent162 pages
dc.identifier.urihttps://hdl.handle.net/1920/11305
dc.language.isoen
dc.rightsCopyright 2017 MAHMOUD AWAD
dc.subjectInformation technology
dc.subjectAUTONOMIC SYSTEMS
dc.subjectNON-LINEAR OPTIMIZATION
dc.subjectPARAMETER ESTIMATION
dc.subjectPERFORMANCE MODELING
dc.subjectQUEUING NETWORKS
dc.subjectWORKLOAD MODELING
dc.titleDYNAMIC DERIVATION OF ANALYTICAL PERFORMANCE MODELS IN AUTONOMIC SYSTEMS
dc.typeDissertation
thesis.degree.disciplineInformation Technology
thesis.degree.grantorGeorge Mason University
thesis.degree.levelPh.D.

Files

Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
AWAD_gmu_0883E_11384.pdf
Size:
2.41 MB
Format:
Adobe Portable Document Format
Download

Collections

College of Engineering and Computing