ACTS:AutonomousCost-EfficientTask OrchestrationforServerlessAnalytics

Jananie, Jarachanthan; Li, Chen; Fei, Xu

dc.contributor.author	Jananie, Jarachanthan
dc.contributor.author	Li, Chen
dc.contributor.author	Fei, Xu
dc.date.accessioned	2023-12-19T04:19:43Z
dc.date.available	2023-12-19T04:19:43Z
dc.date.issued	2023
dc.identifier.issn	979-8-3503-9973-8
dc.identifier.uri	http://repo.lib.jfn.ac.lk/ujrr/handle/123456789/9956
dc.description.abstract	—Serverlesscomputinghasbecomeincreasinglypopularforcloudapplications,duetoitscompellingpropertiesof high-levelabstractions,lightweightruntime,highelasticityand pay-per-usebilling.Inthisrevolutionarycomputingparadigm shift,challengesarisewhenadaptingdataanalyticsapplications totheserverlessenvironment,duetothelackofsupport forefficientstatesharing,whichattractever-growingresearch attention.Inthispaper,weaimtoexploittheadvantagesoftasklevelorchestrationandfine-grainedresourceprovisioningfor dataanalyticsonserverlessplatforms,withthehopeoffulfilling thepromiseofserverlessdeploymenttothemaximumextent. Tothisend,wepresentACTS,anautonomouscost-efficient taskorchestrationframeworkforserverlessanalytics.ACTS judiciouslyschedulesandcoordinatesfunctiontaskstomitigate cold-startlatencyandstatesharingoverhead.Inaddition,ACTS explorestheoptimizationspaceoffine-grainedworkloaddistributionandfunctionresourceconfigurationforcostefficiency. WehavedeployedandimplementedACTSonAWSLambda, evaluatedwithvariousdataanalyticsworkloads.Resultsfrom extensiveexperimentsdemonstratethatACTSachievesupto 98%monetarycostreductionwhilemaintainingsuperiorjob completiontimeperformance,incomparisonwiththestate-ofthe-artbaselines.	en_US
dc.language.iso	en	en_US
dc.publisher	IEEE	en_US
dc.subject	Serverless computing,	en_US
dc.subject	Cost-efficiency,	en_US
dc.subject	Data analytics,	en_US
dc.subject	Cloud resource provisioning	en_US
dc.title	ACTS:AutonomousCost-EfficientTask OrchestrationforServerlessAnalytics	en_US
dc.type	Article	en_US