Files in this item



application/pdfYAN-THESIS-2016.pdf (426kB)
(no description provided)PDF


Title:Performance evaluation of VM-level record-and-replay techniques and applications
Author(s):Yan, Mengjia
Advisor(s):Torrellas, Josep
Department / Program:Computer Science
Discipline:Computer Science
Degree Granting Institution:University of Illinois at Urbana-Champaign
Subject(s):Record and Replay
VMI Analysis
Abstract:Virtual machine level record and replay can be used for complex system debugging and analysis, fault-tolerance replication and forensic analysis. Previous work on performance evaluation of RnR frameworks are not complete enough due to their narrow focuses. RnR related projects either focus on performance evaluation of plain record and replay mechanisms or specifically target the effectiveness of the functionality RnR supports. In order to identify the performance bottlenecks in the complicated RnR system and its various applications, this thesis conducts a thorough evaluation and analysis on 3 different modes of RnR, that is, record, replay with checkpointing and replay with VMI analysis. Both RnR system developer and users can benefit from our work. With our evaluation results, system developer can propose more efficient design accordingly, and RnR users can configure the system properly to achieve expected performance.
Issue Date:2016-04-28
Rights Information:Copyright 2016 Mengjia Yan
Date Available in IDEALS:2016-07-07
Date Deposited:2016-05

This item appears in the following Collection(s)

Item Statistics