SlideShare a Scribd company logo
1 of 16
Download to read offline
1




A Performance Study on the VM
   Startup Time in the Cloud

                 Cloud 2012
            (June 25, Honolulu, Hawaii)




      Ming Mao, Marty Humphrey
      CS Department, University of Virginia
Introduction
2


       One great advantage of the cloud is dynamic scalability,
        BUT…
          Only meaningful when the compute resource is available in time
          The fact is that acquired instances are NOT ready immediately

       Why study the cloud VM startup time
          Important  to time sensitive applications
          Important to resource auto-scaling mechanisms
          Inconsistent definitions across cloud providers
          To set up a reference for the community
          To know the difference from two years ago
Experiment setup
3


    Type                    OS Image                   Location        AWS, Windows Azure,
                           Amazon EC2                                   Rackspace
    m1.small   Linux(Fedora) ami-48aa4921              us-east-1a
    m1.small   Windows (Win Server 2008)               us-east-1a      Oct 15th, 2011 – Feb 15th 2012
               ami-fbf93092
                         Windows Azure                                 A client periodically collects the
    Small      WebRole                                 South            instance startup information
               default WebRole app in Azure SDK        Central US
    Small      WorkerRole                              South           By different factors – time,
               default WorkerRole app in Azure SDK     Central US
    Small      VMRole                                  South            image size, machine type,
               Win Server 2008R2                       Central US
                            Rackspace
                                                                        location, etc.
    Type IV    Linux (Fedora) flavor 71                N/A
    Type IV    Windows (Win Server 2008R2) flavor 28   N/A



                                                             VM              Definition of VM startup time
                                                         Linux VMs                The first successful ssh login
                                                        Windows VMs     The first successful remote desktop connection
                                                         WebRoles               The first successful http request
                                                        WorkerRoles                The first successful logging
Results (1) – by time
4




        Cloud           OS       Average VM startup time
        EC2         Linux               96.9 seconds
        EC2         Windows            810.2 seconds
        Azure       WebRole            374.8 seconds
        Azure       WorkerRole         406.2 seconds
        Azure       VMRole             356.6 seconds
        Rackspace   Linux               44.2 seconds
        Rackspace   Windows            429.2 seconds
Results (2) – by image size
5




           Cloud        Average Data Transfer Rate
                       between VM and Image Store
           EC2                  10.9 MB/s
           Azure                 1.1MB/s
           Rackspace            22.5 MB/s
Results (3) – by machine type
6




       Azure, Rackspace – the VM startup time goes longer as the
        instance type goes larger
       EC2 – no significant differences among different instance
        types, except the micro ones
Results (4) – by location
7




       No significant differences among locations
       EC2 - The newly established data center has a slightly slower
        startup performance and greater variance
Results (5) – multiple instances
8




       EC2 – VM startup time is relatively constant.
       Azure – the last VM instance sometimes took significantly
        longer than the first one.
       RS – does not support instance acquisition in batch
Results (6 - 1) – spot instances, by time
9




       Spot instances have a longer waiting time but similar VM
        startup performance compared to on-demand instances
Results (6 - 2) – spot instances, by vm type
10




        Similar to on-demand instances, the VM startup time shows
         no significant differences across different machine types
Results (6 - 3) – spot instances, by location
11




        The US east region shows longer spot instance acquisition
         time
Results (6 - 4) – spot instances, multiple instances
12




        Similar to on-demand instances, spot instances show
         constant performance across all the 16 machines acquired at
         the same time
Results (6 - 5) – spot instances, by real-time price
13




        No significant relations between the VM startup time and the
         real-time price
Result (7) – other facts
14


                                       Spot instances are not
                                        always cheaper than on-
                                        demand VMs
                                       Acquisition requests are
                                        not always successfully
                                        served (unusable
                                        machines).


         Cloud       Success Rate          VM Release Time
      EC2               99.2%                 3-8 seconds
      Azure             99.6%                 8-21 seconds
      Rackspace         92.0%                 3-8 seconds
Summary
15


        Summary
      VM startup time is independent of time of the day.
      EC2, Rackspace - Windows instances take 9 times longer
       than Linux instances. Azure - all three Role instances show
       similar performances.
      Both the size of the OS image and the instance type can
       largely affect the VM startup time.
      spot instances - longer VM startup time and greater variance
      RS had a higher instance acquisition failure rate (8%) > EC2
       (0.8%) & Azure (0.4%).
      Azure – 200-second improvement and smaller variances;
       EC2 – performance does not change.
Questions
16




                 Thank you!

More Related Content

What's hot

BACD July 2012 : The Xen Cloud Platform
BACD July 2012 : The Xen Cloud Platform BACD July 2012 : The Xen Cloud Platform
BACD July 2012 : The Xen Cloud Platform The Linux Foundation
 
Securing Your Cloud With the Xen Hypervisor by Russell Pavlicek
Securing Your Cloud With the Xen Hypervisor by Russell PavlicekSecuring Your Cloud With the Xen Hypervisor by Russell Pavlicek
Securing Your Cloud With the Xen Hypervisor by Russell Pavlicekbuildacloud
 
Techdays SE 2016 - Micros.. err Microcosmos
Techdays SE 2016 - Micros.. err MicrocosmosTechdays SE 2016 - Micros.. err Microcosmos
Techdays SE 2016 - Micros.. err MicrocosmosMike Martin
 
Containers and docker
Containers and dockerContainers and docker
Containers and dockerSUDIP GHOSH
 
Building your own personal cloud with Eucalyptus
Building your own personal cloud with EucalyptusBuilding your own personal cloud with Eucalyptus
Building your own personal cloud with EucalyptusOrlando_Ruby_Users_Group
 
Dockerization of Azure Platform
Dockerization of Azure PlatformDockerization of Azure Platform
Dockerization of Azure Platformnirajrules
 
Automating Dev Environment - Introduction to Docker and Chef
Automating Dev Environment - Introduction to Docker and ChefAutomating Dev Environment - Introduction to Docker and Chef
Automating Dev Environment - Introduction to Docker and Chefkamalikamj
 
From Monolith to Docker Distributed Applications. JavaOne
From Monolith to Docker Distributed Applications. JavaOneFrom Monolith to Docker Distributed Applications. JavaOne
From Monolith to Docker Distributed Applications. JavaOneCarlos Sanchez
 
vSphere APIs for performance monitoring
vSphere APIs for performance monitoringvSphere APIs for performance monitoring
vSphere APIs for performance monitoringAlan Renouf
 
Virtual machines and containers
Virtual machines and containersVirtual machines and containers
Virtual machines and containersPatrick Pierson
 
Deploying private cloud with eucalyptus
Deploying private cloud with eucalyptusDeploying private cloud with eucalyptus
Deploying private cloud with eucalyptusBeny Raja
 
CIF16/Scale14x: The latest from the Xen Project (Lars Kurth, Chairman of Xen ...
CIF16/Scale14x: The latest from the Xen Project (Lars Kurth, Chairman of Xen ...CIF16/Scale14x: The latest from the Xen Project (Lars Kurth, Chairman of Xen ...
CIF16/Scale14x: The latest from the Xen Project (Lars Kurth, Chairman of Xen ...The Linux Foundation
 
Docker: Testing to Production
Docker: Testing to ProductionDocker: Testing to Production
Docker: Testing to ProductionEdwin Fuquen
 
Introduction to Docker and deployment and Azure
Introduction to Docker and deployment and AzureIntroduction to Docker and deployment and Azure
Introduction to Docker and deployment and AzureJérôme Petazzoni
 
Hoverboards, Jetpacks, Clusters and Flux Capacitors
Hoverboards, Jetpacks,  Clusters and Flux CapacitorsHoverboards, Jetpacks,  Clusters and Flux Capacitors
Hoverboards, Jetpacks, Clusters and Flux CapacitorsKubeAcademy
 
EMEA OpenStack Day, July 13th 2011 in London - Jim Curry intro
EMEA OpenStack Day, July 13th 2011 in London - Jim Curry introEMEA OpenStack Day, July 13th 2011 in London - Jim Curry intro
EMEA OpenStack Day, July 13th 2011 in London - Jim Curry introOpen Stack
 
ASP.NET Core and Docker
ASP.NET Core and DockerASP.NET Core and Docker
ASP.NET Core and DockerChuck Megivern
 

What's hot (20)

BACD July 2012 : The Xen Cloud Platform
BACD July 2012 : The Xen Cloud Platform BACD July 2012 : The Xen Cloud Platform
BACD July 2012 : The Xen Cloud Platform
 
Securing Your Cloud With the Xen Hypervisor by Russell Pavlicek
Securing Your Cloud With the Xen Hypervisor by Russell PavlicekSecuring Your Cloud With the Xen Hypervisor by Russell Pavlicek
Securing Your Cloud With the Xen Hypervisor by Russell Pavlicek
 
Xen time machine
Xen time machineXen time machine
Xen time machine
 
Techdays SE 2016 - Micros.. err Microcosmos
Techdays SE 2016 - Micros.. err MicrocosmosTechdays SE 2016 - Micros.. err Microcosmos
Techdays SE 2016 - Micros.. err Microcosmos
 
Containers and docker
Containers and dockerContainers and docker
Containers and docker
 
Building your own personal cloud with Eucalyptus
Building your own personal cloud with EucalyptusBuilding your own personal cloud with Eucalyptus
Building your own personal cloud with Eucalyptus
 
Dockerization of Azure Platform
Dockerization of Azure PlatformDockerization of Azure Platform
Dockerization of Azure Platform
 
Automating Dev Environment - Introduction to Docker and Chef
Automating Dev Environment - Introduction to Docker and ChefAutomating Dev Environment - Introduction to Docker and Chef
Automating Dev Environment - Introduction to Docker and Chef
 
From Monolith to Docker Distributed Applications. JavaOne
From Monolith to Docker Distributed Applications. JavaOneFrom Monolith to Docker Distributed Applications. JavaOne
From Monolith to Docker Distributed Applications. JavaOne
 
vSphere APIs for performance monitoring
vSphere APIs for performance monitoringvSphere APIs for performance monitoring
vSphere APIs for performance monitoring
 
Virtual machines and containers
Virtual machines and containersVirtual machines and containers
Virtual machines and containers
 
Deploying private cloud with eucalyptus
Deploying private cloud with eucalyptusDeploying private cloud with eucalyptus
Deploying private cloud with eucalyptus
 
CIF16/Scale14x: The latest from the Xen Project (Lars Kurth, Chairman of Xen ...
CIF16/Scale14x: The latest from the Xen Project (Lars Kurth, Chairman of Xen ...CIF16/Scale14x: The latest from the Xen Project (Lars Kurth, Chairman of Xen ...
CIF16/Scale14x: The latest from the Xen Project (Lars Kurth, Chairman of Xen ...
 
Windows Server 2012 R2! Что нового в Hyper-V?
Windows Server 2012 R2! Что нового в Hyper-V?Windows Server 2012 R2! Что нового в Hyper-V?
Windows Server 2012 R2! Что нового в Hyper-V?
 
Docker: Testing to Production
Docker: Testing to ProductionDocker: Testing to Production
Docker: Testing to Production
 
Introduction to Docker and deployment and Azure
Introduction to Docker and deployment and AzureIntroduction to Docker and deployment and Azure
Introduction to Docker and deployment and Azure
 
Xen in Linux (aka PVOPS update)
Xen in Linux (aka PVOPS update)Xen in Linux (aka PVOPS update)
Xen in Linux (aka PVOPS update)
 
Hoverboards, Jetpacks, Clusters and Flux Capacitors
Hoverboards, Jetpacks,  Clusters and Flux CapacitorsHoverboards, Jetpacks,  Clusters and Flux Capacitors
Hoverboards, Jetpacks, Clusters and Flux Capacitors
 
EMEA OpenStack Day, July 13th 2011 in London - Jim Curry intro
EMEA OpenStack Day, July 13th 2011 in London - Jim Curry introEMEA OpenStack Day, July 13th 2011 in London - Jim Curry intro
EMEA OpenStack Day, July 13th 2011 in London - Jim Curry intro
 
ASP.NET Core and Docker
ASP.NET Core and DockerASP.NET Core and Docker
ASP.NET Core and Docker
 

Similar to A performance study on the vm startup time in the cloud

20090911 virtualizationandcloud
20090911 virtualizationandcloud20090911 virtualizationandcloud
20090911 virtualizationandcloudSupratik Ghatak
 
20090911 virtualizationandcloud
20090911 virtualizationandcloud20090911 virtualizationandcloud
20090911 virtualizationandcloudMeenal Joshi
 
virtualization and cloud
virtualization and cloudvirtualization and cloud
virtualization and cloudsankarimsc
 
Wars I’ve Seen From Java EE to Spring and more, Azure has you covered
Wars I’ve SeenFrom Java EE to Spring and more, Azure has you coveredWars I’ve SeenFrom Java EE to Spring and more, Azure has you covered
Wars I’ve Seen From Java EE to Spring and more, Azure has you coveredEdward Burns
 
Operating System extension.docx
Operating System extension.docxOperating System extension.docx
Operating System extension.docxkumari36
 
Intro to virtualization
Intro to virtualizationIntro to virtualization
Intro to virtualizationKalpna Saharan
 
20090911 virtualizationandcloud
20090911 virtualizationandcloud20090911 virtualizationandcloud
20090911 virtualizationandcloudDebabrata Debnath
 
2020-02-10 Java on Azure Solution Briefing
2020-02-10 Java on Azure Solution Briefing2020-02-10 Java on Azure Solution Briefing
2020-02-10 Java on Azure Solution BriefingEd Burns
 
ICALEPCS 2011: Testing Environments using Virtualization
ICALEPCS 2011: Testing Environments using VirtualizationICALEPCS 2011: Testing Environments using Virtualization
ICALEPCS 2011: Testing Environments using VirtualizationOmer Khalid
 
HotLink DR Express
HotLink DR ExpressHotLink DR Express
HotLink DR Expressdean1609
 
01 overview-servlets-and-environment-setup
01 overview-servlets-and-environment-setup01 overview-servlets-and-environment-setup
01 overview-servlets-and-environment-setupdhrubo kayal
 
Enterprise Java on Microsoft Azure: From Java EE to Spring, we’ve got you cov...
Enterprise Java on Microsoft Azure: From Java EE to Spring, we’ve got you cov...Enterprise Java on Microsoft Azure: From Java EE to Spring, we’ve got you cov...
Enterprise Java on Microsoft Azure: From Java EE to Spring, we’ve got you cov...Ed Burns
 
Rails in the Cloud - Experiences from running on EC2
Rails in the Cloud - Experiences from running on EC2Rails in the Cloud - Experiences from running on EC2
Rails in the Cloud - Experiences from running on EC2Jonathan Weiss
 
South jersey sql virtualization
South jersey sql virtualizationSouth jersey sql virtualization
South jersey sql virtualizationJoseph D'Antoni
 
Cloud computing virtualization
Cloud computing virtualizationCloud computing virtualization
Cloud computing virtualizationAyaz Shahid
 

Similar to A performance study on the vm startup time in the cloud (20)

20090911 virtualizationandcloud
20090911 virtualizationandcloud20090911 virtualizationandcloud
20090911 virtualizationandcloud
 
20090911 virtualizationandcloud
20090911 virtualizationandcloud20090911 virtualizationandcloud
20090911 virtualizationandcloud
 
virtualization and cloud
virtualization and cloudvirtualization and cloud
virtualization and cloud
 
Wars I’ve Seen From Java EE to Spring and more, Azure has you covered
Wars I’ve SeenFrom Java EE to Spring and more, Azure has you coveredWars I’ve SeenFrom Java EE to Spring and more, Azure has you covered
Wars I’ve Seen From Java EE to Spring and more, Azure has you covered
 
Sql saturday dc vm ware
Sql saturday dc vm wareSql saturday dc vm ware
Sql saturday dc vm ware
 
Operating System extension.docx
Operating System extension.docxOperating System extension.docx
Operating System extension.docx
 
Intro to virtualization
Intro to virtualizationIntro to virtualization
Intro to virtualization
 
Azure IaaS
Azure IaaSAzure IaaS
Azure IaaS
 
20090911 virtualizationandcloud
20090911 virtualizationandcloud20090911 virtualizationandcloud
20090911 virtualizationandcloud
 
2020-02-10 Java on Azure Solution Briefing
2020-02-10 Java on Azure Solution Briefing2020-02-10 Java on Azure Solution Briefing
2020-02-10 Java on Azure Solution Briefing
 
Your First Week with Amazon EC2
Your First Week with Amazon EC2Your First Week with Amazon EC2
Your First Week with Amazon EC2
 
ICALEPCS 2011: Testing Environments using Virtualization
ICALEPCS 2011: Testing Environments using VirtualizationICALEPCS 2011: Testing Environments using Virtualization
ICALEPCS 2011: Testing Environments using Virtualization
 
HotLink DR Express
HotLink DR ExpressHotLink DR Express
HotLink DR Express
 
OSGi Cloud Ecosystems
OSGi Cloud EcosystemsOSGi Cloud Ecosystems
OSGi Cloud Ecosystems
 
Rails in the Cloud
Rails in the CloudRails in the Cloud
Rails in the Cloud
 
01 overview-servlets-and-environment-setup
01 overview-servlets-and-environment-setup01 overview-servlets-and-environment-setup
01 overview-servlets-and-environment-setup
 
Enterprise Java on Microsoft Azure: From Java EE to Spring, we’ve got you cov...
Enterprise Java on Microsoft Azure: From Java EE to Spring, we’ve got you cov...Enterprise Java on Microsoft Azure: From Java EE to Spring, we’ve got you cov...
Enterprise Java on Microsoft Azure: From Java EE to Spring, we’ve got you cov...
 
Rails in the Cloud - Experiences from running on EC2
Rails in the Cloud - Experiences from running on EC2Rails in the Cloud - Experiences from running on EC2
Rails in the Cloud - Experiences from running on EC2
 
South jersey sql virtualization
South jersey sql virtualizationSouth jersey sql virtualization
South jersey sql virtualization
 
Cloud computing virtualization
Cloud computing virtualizationCloud computing virtualization
Cloud computing virtualization
 

Recently uploaded

Advanced Test Driven-Development @ php[tek] 2024
Advanced Test Driven-Development @ php[tek] 2024Advanced Test Driven-Development @ php[tek] 2024
Advanced Test Driven-Development @ php[tek] 2024Scott Keck-Warren
 
Streamlining Python Development: A Guide to a Modern Project Setup
Streamlining Python Development: A Guide to a Modern Project SetupStreamlining Python Development: A Guide to a Modern Project Setup
Streamlining Python Development: A Guide to a Modern Project SetupFlorian Wilhelm
 
Ensuring Technical Readiness For Copilot in Microsoft 365
Ensuring Technical Readiness For Copilot in Microsoft 365Ensuring Technical Readiness For Copilot in Microsoft 365
Ensuring Technical Readiness For Copilot in Microsoft 3652toLead Limited
 
The Future of Software Development - Devin AI Innovative Approach.pdf
The Future of Software Development - Devin AI Innovative Approach.pdfThe Future of Software Development - Devin AI Innovative Approach.pdf
The Future of Software Development - Devin AI Innovative Approach.pdfSeasiaInfotech2
 
Designing IA for AI - Information Architecture Conference 2024
Designing IA for AI - Information Architecture Conference 2024Designing IA for AI - Information Architecture Conference 2024
Designing IA for AI - Information Architecture Conference 2024Enterprise Knowledge
 
WordPress Websites for Engineers: Elevate Your Brand
WordPress Websites for Engineers: Elevate Your BrandWordPress Websites for Engineers: Elevate Your Brand
WordPress Websites for Engineers: Elevate Your Brandgvaughan
 
Install Stable Diffusion in windows machine
Install Stable Diffusion in windows machineInstall Stable Diffusion in windows machine
Install Stable Diffusion in windows machinePadma Pradeep
 
Connect Wave/ connectwave Pitch Deck Presentation
Connect Wave/ connectwave Pitch Deck PresentationConnect Wave/ connectwave Pitch Deck Presentation
Connect Wave/ connectwave Pitch Deck PresentationSlibray Presentation
 
Powerpoint exploring the locations used in television show Time Clash
Powerpoint exploring the locations used in television show Time ClashPowerpoint exploring the locations used in television show Time Clash
Powerpoint exploring the locations used in television show Time Clashcharlottematthew16
 
Kotlin Multiplatform & Compose Multiplatform - Starter kit for pragmatics
Kotlin Multiplatform & Compose Multiplatform - Starter kit for pragmaticsKotlin Multiplatform & Compose Multiplatform - Starter kit for pragmatics
Kotlin Multiplatform & Compose Multiplatform - Starter kit for pragmaticscarlostorres15106
 
Story boards and shot lists for my a level piece
Story boards and shot lists for my a level pieceStory boards and shot lists for my a level piece
Story boards and shot lists for my a level piececharlottematthew16
 
DevEX - reference for building teams, processes, and platforms
DevEX - reference for building teams, processes, and platformsDevEX - reference for building teams, processes, and platforms
DevEX - reference for building teams, processes, and platformsSergiu Bodiu
 
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
"Federated learning: out of reach no matter how close",Oleksandr LapshynFwdays
 
Unraveling Multimodality with Large Language Models.pdf
Unraveling Multimodality with Large Language Models.pdfUnraveling Multimodality with Large Language Models.pdf
Unraveling Multimodality with Large Language Models.pdfAlex Barbosa Coqueiro
 
Gen AI in Business - Global Trends Report 2024.pdf
Gen AI in Business - Global Trends Report 2024.pdfGen AI in Business - Global Trends Report 2024.pdf
Gen AI in Business - Global Trends Report 2024.pdfAddepto
 
Vertex AI Gemini Prompt Engineering Tips
Vertex AI Gemini Prompt Engineering TipsVertex AI Gemini Prompt Engineering Tips
Vertex AI Gemini Prompt Engineering TipsMiki Katsuragi
 
Search Engine Optimization SEO PDF for 2024.pdf
Search Engine Optimization SEO PDF for 2024.pdfSearch Engine Optimization SEO PDF for 2024.pdf
Search Engine Optimization SEO PDF for 2024.pdfRankYa
 
What's New in Teams Calling, Meetings and Devices March 2024
What's New in Teams Calling, Meetings and Devices March 2024What's New in Teams Calling, Meetings and Devices March 2024
What's New in Teams Calling, Meetings and Devices March 2024Stephanie Beckett
 
Beyond Boundaries: Leveraging No-Code Solutions for Industry Innovation
Beyond Boundaries: Leveraging No-Code Solutions for Industry InnovationBeyond Boundaries: Leveraging No-Code Solutions for Industry Innovation
Beyond Boundaries: Leveraging No-Code Solutions for Industry InnovationSafe Software
 

Recently uploaded (20)

Advanced Test Driven-Development @ php[tek] 2024
Advanced Test Driven-Development @ php[tek] 2024Advanced Test Driven-Development @ php[tek] 2024
Advanced Test Driven-Development @ php[tek] 2024
 
Streamlining Python Development: A Guide to a Modern Project Setup
Streamlining Python Development: A Guide to a Modern Project SetupStreamlining Python Development: A Guide to a Modern Project Setup
Streamlining Python Development: A Guide to a Modern Project Setup
 
Ensuring Technical Readiness For Copilot in Microsoft 365
Ensuring Technical Readiness For Copilot in Microsoft 365Ensuring Technical Readiness For Copilot in Microsoft 365
Ensuring Technical Readiness For Copilot in Microsoft 365
 
The Future of Software Development - Devin AI Innovative Approach.pdf
The Future of Software Development - Devin AI Innovative Approach.pdfThe Future of Software Development - Devin AI Innovative Approach.pdf
The Future of Software Development - Devin AI Innovative Approach.pdf
 
E-Vehicle_Hacking_by_Parul Sharma_null_owasp.pptx
E-Vehicle_Hacking_by_Parul Sharma_null_owasp.pptxE-Vehicle_Hacking_by_Parul Sharma_null_owasp.pptx
E-Vehicle_Hacking_by_Parul Sharma_null_owasp.pptx
 
Designing IA for AI - Information Architecture Conference 2024
Designing IA for AI - Information Architecture Conference 2024Designing IA for AI - Information Architecture Conference 2024
Designing IA for AI - Information Architecture Conference 2024
 
WordPress Websites for Engineers: Elevate Your Brand
WordPress Websites for Engineers: Elevate Your BrandWordPress Websites for Engineers: Elevate Your Brand
WordPress Websites for Engineers: Elevate Your Brand
 
Install Stable Diffusion in windows machine
Install Stable Diffusion in windows machineInstall Stable Diffusion in windows machine
Install Stable Diffusion in windows machine
 
Connect Wave/ connectwave Pitch Deck Presentation
Connect Wave/ connectwave Pitch Deck PresentationConnect Wave/ connectwave Pitch Deck Presentation
Connect Wave/ connectwave Pitch Deck Presentation
 
Powerpoint exploring the locations used in television show Time Clash
Powerpoint exploring the locations used in television show Time ClashPowerpoint exploring the locations used in television show Time Clash
Powerpoint exploring the locations used in television show Time Clash
 
Kotlin Multiplatform & Compose Multiplatform - Starter kit for pragmatics
Kotlin Multiplatform & Compose Multiplatform - Starter kit for pragmaticsKotlin Multiplatform & Compose Multiplatform - Starter kit for pragmatics
Kotlin Multiplatform & Compose Multiplatform - Starter kit for pragmatics
 
Story boards and shot lists for my a level piece
Story boards and shot lists for my a level pieceStory boards and shot lists for my a level piece
Story boards and shot lists for my a level piece
 
DevEX - reference for building teams, processes, and platforms
DevEX - reference for building teams, processes, and platformsDevEX - reference for building teams, processes, and platforms
DevEX - reference for building teams, processes, and platforms
 
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
 
Unraveling Multimodality with Large Language Models.pdf
Unraveling Multimodality with Large Language Models.pdfUnraveling Multimodality with Large Language Models.pdf
Unraveling Multimodality with Large Language Models.pdf
 
Gen AI in Business - Global Trends Report 2024.pdf
Gen AI in Business - Global Trends Report 2024.pdfGen AI in Business - Global Trends Report 2024.pdf
Gen AI in Business - Global Trends Report 2024.pdf
 
Vertex AI Gemini Prompt Engineering Tips
Vertex AI Gemini Prompt Engineering TipsVertex AI Gemini Prompt Engineering Tips
Vertex AI Gemini Prompt Engineering Tips
 
Search Engine Optimization SEO PDF for 2024.pdf
Search Engine Optimization SEO PDF for 2024.pdfSearch Engine Optimization SEO PDF for 2024.pdf
Search Engine Optimization SEO PDF for 2024.pdf
 
What's New in Teams Calling, Meetings and Devices March 2024
What's New in Teams Calling, Meetings and Devices March 2024What's New in Teams Calling, Meetings and Devices March 2024
What's New in Teams Calling, Meetings and Devices March 2024
 
Beyond Boundaries: Leveraging No-Code Solutions for Industry Innovation
Beyond Boundaries: Leveraging No-Code Solutions for Industry InnovationBeyond Boundaries: Leveraging No-Code Solutions for Industry Innovation
Beyond Boundaries: Leveraging No-Code Solutions for Industry Innovation
 

A performance study on the vm startup time in the cloud

  • 1. 1 A Performance Study on the VM Startup Time in the Cloud Cloud 2012 (June 25, Honolulu, Hawaii) Ming Mao, Marty Humphrey CS Department, University of Virginia
  • 2. Introduction 2  One great advantage of the cloud is dynamic scalability, BUT…  Only meaningful when the compute resource is available in time  The fact is that acquired instances are NOT ready immediately  Why study the cloud VM startup time  Important to time sensitive applications  Important to resource auto-scaling mechanisms  Inconsistent definitions across cloud providers  To set up a reference for the community  To know the difference from two years ago
  • 3. Experiment setup 3 Type OS Image Location  AWS, Windows Azure, Amazon EC2 Rackspace m1.small Linux(Fedora) ami-48aa4921 us-east-1a m1.small Windows (Win Server 2008) us-east-1a  Oct 15th, 2011 – Feb 15th 2012 ami-fbf93092 Windows Azure  A client periodically collects the Small WebRole South instance startup information default WebRole app in Azure SDK Central US Small WorkerRole South  By different factors – time, default WorkerRole app in Azure SDK Central US Small VMRole South image size, machine type, Win Server 2008R2 Central US Rackspace location, etc. Type IV Linux (Fedora) flavor 71 N/A Type IV Windows (Win Server 2008R2) flavor 28 N/A VM Definition of VM startup time Linux VMs The first successful ssh login Windows VMs The first successful remote desktop connection WebRoles The first successful http request WorkerRoles The first successful logging
  • 4. Results (1) – by time 4 Cloud OS Average VM startup time EC2 Linux 96.9 seconds EC2 Windows 810.2 seconds Azure WebRole 374.8 seconds Azure WorkerRole 406.2 seconds Azure VMRole 356.6 seconds Rackspace Linux 44.2 seconds Rackspace Windows 429.2 seconds
  • 5. Results (2) – by image size 5 Cloud Average Data Transfer Rate between VM and Image Store EC2 10.9 MB/s Azure 1.1MB/s Rackspace 22.5 MB/s
  • 6. Results (3) – by machine type 6  Azure, Rackspace – the VM startup time goes longer as the instance type goes larger  EC2 – no significant differences among different instance types, except the micro ones
  • 7. Results (4) – by location 7  No significant differences among locations  EC2 - The newly established data center has a slightly slower startup performance and greater variance
  • 8. Results (5) – multiple instances 8  EC2 – VM startup time is relatively constant.  Azure – the last VM instance sometimes took significantly longer than the first one.  RS – does not support instance acquisition in batch
  • 9. Results (6 - 1) – spot instances, by time 9  Spot instances have a longer waiting time but similar VM startup performance compared to on-demand instances
  • 10. Results (6 - 2) – spot instances, by vm type 10  Similar to on-demand instances, the VM startup time shows no significant differences across different machine types
  • 11. Results (6 - 3) – spot instances, by location 11  The US east region shows longer spot instance acquisition time
  • 12. Results (6 - 4) – spot instances, multiple instances 12  Similar to on-demand instances, spot instances show constant performance across all the 16 machines acquired at the same time
  • 13. Results (6 - 5) – spot instances, by real-time price 13  No significant relations between the VM startup time and the real-time price
  • 14. Result (7) – other facts 14  Spot instances are not always cheaper than on- demand VMs  Acquisition requests are not always successfully served (unusable machines). Cloud Success Rate VM Release Time EC2 99.2% 3-8 seconds Azure 99.6% 8-21 seconds Rackspace 92.0% 3-8 seconds
  • 15. Summary 15  Summary  VM startup time is independent of time of the day.  EC2, Rackspace - Windows instances take 9 times longer than Linux instances. Azure - all three Role instances show similar performances.  Both the size of the OS image and the instance type can largely affect the VM startup time.  spot instances - longer VM startup time and greater variance  RS had a higher instance acquisition failure rate (8%) > EC2 (0.8%) & Azure (0.4%).  Azure – 200-second improvement and smaller variances; EC2 – performance does not change.
  • 16. Questions 16 Thank you!