Astrea: Auto-Serverless Analytics Towards Cost-Efficiency and QoS-Awareness

Simple item page
dc.contributor.authorJananie, J.
dc.contributor.authorLi Chen
dc.contributor.authorFei Xu
dc.contributor.authorBoLi
dc.date.accessioned2023-12-29T07:19:33Z
dc.date.available2023-12-29T07:19:33Z
dc.date.issued2022
dc.description.abstractWiththeabilitytosimplifythecodedeploymentwithone-clickuploadandlightweightexecution,serverlesscomputinghas emergedasapromisingparadigmwithincreasingpopularity.However,thereremainopenchallengeswhenadaptingdata-intensive analyticsapplicationstotheserverlesscontext,inwhichusersofserverlessanalyticsencounterthedifficultyincoordinatingcomputation acrossdifferentstagesandprovisioningresourcesinalargeconfigurationspace.Thispaperpresentsourdesignandimplementationof Astrea,whichconfiguresandorchestratesserverlessanalyticsjobsinanautonomousmanner,whiletakingintoaccountflexibly-specified userrequirements.Astreareliesonthemodelingofperformanceandcostwhichcharacterizestheintricateinterplayamongmultidimensionalfactors(e.g.,functionmemorysize,degreeofparallelismateachstage).Weformulateanoptimizationproblembasedon user-specificrequirementstowardsperformanceenhancementorcostreduction,anddevelopasetofalgorithmsbasedongraph theorytoobtaintheoptimaljobexecution.WedeployAstreaintheAWSLambdaplatformandconductreal-worldexperimentsover representativebenchmarks,includingBigDataanalyticsandmachinelearningworkloads,atdifferentscales.Extensiveresults demonstratethatAstreacanachievetheoptimalexecutiondecisionforserverlessdataanalytics,incomparisonwithvariousprovisioning anddeploymentbaselines.Forexample,whencomparedwiththreeprovisioningbaselines,Astreamanagestoreducethejobcompletion timeby21%to69%underagivenbudgetconstraint,whilesavingcostby20%to84%withoutviolatingperformancerequirements.en_US
dc.identifier.urihttp://repo.lib.jfn.ac.lk/ujrr/handle/123456789/10021
dc.language.isoenen_US
dc.publisherIEEEen_US
dc.subjectCloud computingen_US
dc.subjectServerless computingen_US
dc.subjectResource provisioningen_US
dc.subjectModelingen_US
dc.subjectOptimizationen_US
dc.titleAstrea: Auto-Serverless Analytics Towards Cost-Efficiency and QoS-Awarenessen_US
dc.typeArticleen_US

Files

Original bundle

Now showing 1 - 1 of 1
Thumbnail Image
Name:
AstreaAuto-ServerlessAnalyticsTowards Cost-EfficiencyandQoS-Awareness.pdf
Size:
2.66 MB
Format:
Adobe Portable Document Format
Description:
Download

License bundle

Now showing 1 - 1 of 1
Thumbnail Image
Name:
license.txt
Size:
1.71 KB
Format:
Item-specific license agreed upon to submission
Description:
Download

Collections

Computer Engineering