2026-04-15T18:50:01Z https://meral.edu.mm/oai

oai:meral.edu.mm:recid/4250 2022-03-24T23:12:18Z 1582963302567:1597824273898 user-ucsy

Working Set Prediction for Live Virtual Machine Migration Zaw, Ei Phyu Thein, Ni Lar Live migration of virtual machines has been apowerful tool to facilitate system maintenance,load balancing, fault tolerance, and powersavingin clusters or data centers. In pre-copyapproach that is mainly used in live migration,VM service downtime is expected to be minimalbut total migration time is prolonged by iterativecopy operations and the significant amount oftransferred data during the whole migrationprocess. In this paper, we present a frameworkthat includes pre-processing phase, push phaseand stop phase, for reducing the amount oftransferred data in pre-copy based livemigration. In pre-processing phase of theframework, we propose Least Recently Used(LRU) based working set prediction algorithm inphysical memory of virtual machine. Theproposed algorithm classify working set list thatcollects the most recent used memory pages andinactive list that collects the least recent usedmemory pages. Applying the proposed workingset prediction algorithm, we can reduce theamount of iteration time of push phase andreduce the total migration time. Fromcalculation result compared with pre copy basedlive migration that is not include pre-processingphase, the proposed framework can reduce thetotal migration time. 2011-05-05 http://hdl.handle.net/20.500.12678/0000004250 https://meral.edu.mm/records/4250