SlideShare a Scribd company logo
Dynamic Optimization of Heterogeneous
Resource Provisioning in Cloud Computing
Masoumeh Tajvidi
Supervisors:
A/Prof. Michael Maher
Dr. Daryl Essam
School of Engineering and
Information Technology (SEIT)
• Introduction
- Cloud computing and its potential benefits
- Challenges!
• Research question
• Related Work
• Research Plan
• Preliminary Work
Outline
1/1
6
What is Cloud Computing?
A broad and deep platform that helps customers build
sophisticated, scalable applications
2/1
6
Why are companies adopting cloud
computing so quickly?
Agility
3/1
6
Main benefits of utility-like computing
Convert CAPEX to OPEX
 Lower Total Cost
No Need to Guess Capacity
Lower Maintenance Cost
4/1
6
Effectively using of cloud computing
 Over-provisioning: purchased resources are not fully utilized
cost more than necessary
 Under-provisioning: purchased resources are not sufficient to
meet the actual demand  hurts application performance,
5/1
6
Efficient resource provisioning:
challenge #1 : Various cloud providers offering
multiple Virtual Machine (VM) types
• Virtualization technologies help providers
pack their hardware resources into
different type of Virtual Machines
• The end-users encounter a complicated
decision making problem for choosing
the right mix of VMs!
6/1
6
Efficient resource provisioning:
challenge #2 : Multiple pricing models
7/1
6
Efficient resource provisioning:
challenge #3 : Deal with COST and DEMAND
uncertainty
 The application’s demand is not
known in advance
• e.g. online video streaming
applications, like YouTube
Channel
 The cost of instances in both on-
demand and spot is varying!
8/1
6
Efficient resource provisioning:
challenge #4 : Multi-objective problem
 Cost is not the only objective for this problem
 The trade-off among cost and QoS must also be
handled
 For example response time is very critical in latency
sensitive applications , online gaming services, as
the users tend to be very impatient
9/1
6
Research Question ?
How to dynamically optimize cloud resource
provisioning plan as a multi-objective problem in
the real-world?
10/1
6
Literature review
11/1
6
stochastic optimization
 The time at which parameters become known, divide the
problem into stages
 Main Question: what to decide in the first stage?
 Main question: how many VM to reserve
 Recourse: how to deal with under or over provisioning
 Objective function is to minimize the expected cost
12/1
6
 Add complexities of the real-world problem into our model.
Take into account the heterogeneity of VM types
Take into account all three pricing schemes into our model
Solve the problem as a multi-objective optimization problem
 Dealing with uncertainty by solving the problem with both
stochastic and approximate programming approaches
 enhance the performance of the optimization problem by
using Machine Learning techniques
Research Plan : Big Picture
Challenge
#1
Challenge
#2
Challenge
#4
Challenge
#3
13/1
6
Preliminary results
 Replicated one of the available multi stage
programming approach
 Modelled the problem in Stochastic MiniZinc
modelling language
 Added spot-instance pricing model into our model
14/1
6
Stochastic MiniZinc Model Results
Before and After Adding Spot Instances
15/1
6
Thank you .

More Related Content

What's hot

Cloud computing (IT-703) UNIT 1 & 2
Cloud computing (IT-703) UNIT 1 & 2Cloud computing (IT-703) UNIT 1 & 2
Cloud computing (IT-703) UNIT 1 & 2
Jitendra s Rathore
 
NIST Cloud Computing Reference Architecture
NIST Cloud Computing Reference ArchitectureNIST Cloud Computing Reference Architecture
NIST Cloud Computing Reference Architecture
Thanakrit Lersmethasakul
 
SLA Agreement, types and Life Cycle
SLA Agreement, types and Life Cycle SLA Agreement, types and Life Cycle
SLA Agreement, types and Life Cycle
Dr Neelesh Jain
 
Distributed Systems Real Life Applications
Distributed Systems Real Life ApplicationsDistributed Systems Real Life Applications
Distributed Systems Real Life Applications
Aman Srivastava
 
Virtualization for Cloud Environment
Virtualization for Cloud EnvironmentVirtualization for Cloud Environment
Virtualization for Cloud Environment
Dr. Sunil Kr. Pandey
 
distributed shared memory
 distributed shared memory distributed shared memory
distributed shared memory
Ashish Kumar
 
Cloud computing
Cloud computingCloud computing
Cloud computing
Pallavi Rai
 
Cloud Computing Security Challenges
Cloud Computing Security ChallengesCloud Computing Security Challenges
Cloud Computing Security Challenges
Yateesh Yadav
 
Virtual machine security
Virtual machine securityVirtual machine security
Virtual machine security
Jacob Zvirikuzhe
 
Introduction to CloudStack
Introduction to CloudStack Introduction to CloudStack
Cloud computing & energy efficiency using cloud to decrease the energy use in...
Cloud computing & energy efficiency using cloud to decrease the energy use in...Cloud computing & energy efficiency using cloud to decrease the energy use in...
Cloud computing & energy efficiency using cloud to decrease the energy use in...
Puru Agrawal
 
History and Evolution of Cloud computing (Safaricom cloud)
History and Evolution of Cloud computing (Safaricom cloud)History and Evolution of Cloud computing (Safaricom cloud)
History and Evolution of Cloud computing (Safaricom cloud)
Ben Wakhungu
 
HCI 3e - Ch 14: Communication and collaboration models
HCI 3e - Ch 14:  Communication and collaboration modelsHCI 3e - Ch 14:  Communication and collaboration models
HCI 3e - Ch 14: Communication and collaboration models
Alan Dix
 
Communications is distributed systems
Communications is distributed systemsCommunications is distributed systems
Communications is distributed systems
SHATHAN
 
Resource management
Resource managementResource management
Resource management
Dr Sandeep Kumar Poonia
 
Vision of cloud computing
Vision of cloud computingVision of cloud computing
Vision of cloud computing
gaurav jain
 
Implementation levels of virtualization
Implementation levels of virtualizationImplementation levels of virtualization
Implementation levels of virtualization
Gokulnath S
 
Task programming
Task programmingTask programming
Task programming
Yogendra Tamang
 
Cloud computing
Cloud computingCloud computing
Cloud computing
Reetesh Gupta
 
Cloud deployment models
Cloud deployment modelsCloud deployment models
Cloud deployment models
Ashok Kumar
 

What's hot (20)

Cloud computing (IT-703) UNIT 1 & 2
Cloud computing (IT-703) UNIT 1 & 2Cloud computing (IT-703) UNIT 1 & 2
Cloud computing (IT-703) UNIT 1 & 2
 
NIST Cloud Computing Reference Architecture
NIST Cloud Computing Reference ArchitectureNIST Cloud Computing Reference Architecture
NIST Cloud Computing Reference Architecture
 
SLA Agreement, types and Life Cycle
SLA Agreement, types and Life Cycle SLA Agreement, types and Life Cycle
SLA Agreement, types and Life Cycle
 
Distributed Systems Real Life Applications
Distributed Systems Real Life ApplicationsDistributed Systems Real Life Applications
Distributed Systems Real Life Applications
 
Virtualization for Cloud Environment
Virtualization for Cloud EnvironmentVirtualization for Cloud Environment
Virtualization for Cloud Environment
 
distributed shared memory
 distributed shared memory distributed shared memory
distributed shared memory
 
Cloud computing
Cloud computingCloud computing
Cloud computing
 
Cloud Computing Security Challenges
Cloud Computing Security ChallengesCloud Computing Security Challenges
Cloud Computing Security Challenges
 
Virtual machine security
Virtual machine securityVirtual machine security
Virtual machine security
 
Introduction to CloudStack
Introduction to CloudStack Introduction to CloudStack
Introduction to CloudStack
 
Cloud computing & energy efficiency using cloud to decrease the energy use in...
Cloud computing & energy efficiency using cloud to decrease the energy use in...Cloud computing & energy efficiency using cloud to decrease the energy use in...
Cloud computing & energy efficiency using cloud to decrease the energy use in...
 
History and Evolution of Cloud computing (Safaricom cloud)
History and Evolution of Cloud computing (Safaricom cloud)History and Evolution of Cloud computing (Safaricom cloud)
History and Evolution of Cloud computing (Safaricom cloud)
 
HCI 3e - Ch 14: Communication and collaboration models
HCI 3e - Ch 14:  Communication and collaboration modelsHCI 3e - Ch 14:  Communication and collaboration models
HCI 3e - Ch 14: Communication and collaboration models
 
Communications is distributed systems
Communications is distributed systemsCommunications is distributed systems
Communications is distributed systems
 
Resource management
Resource managementResource management
Resource management
 
Vision of cloud computing
Vision of cloud computingVision of cloud computing
Vision of cloud computing
 
Implementation levels of virtualization
Implementation levels of virtualizationImplementation levels of virtualization
Implementation levels of virtualization
 
Task programming
Task programmingTask programming
Task programming
 
Cloud computing
Cloud computingCloud computing
Cloud computing
 
Cloud deployment models
Cloud deployment modelsCloud deployment models
Cloud deployment models
 

Similar to Resource provisioning optimization in cloud computing

Above the Clouds: A Berkeley View of Cloud Computing: Paper Review
Above the Clouds: A Berkeley View of Cloud Computing:  Paper Review Above the Clouds: A Berkeley View of Cloud Computing:  Paper Review
Above the Clouds: A Berkeley View of Cloud Computing: Paper Review
Mala Deep Upadhaya
 
NCMS UberCloud Experiment Webinar .
NCMS UberCloud Experiment Webinar .NCMS UberCloud Experiment Webinar .
NCMS UberCloud Experiment Webinar .
hpcexperiment
 
Job scheduling in hybrid cloud using deep reinforcement learning for cost opt...
Job scheduling in hybrid cloud using deep reinforcement learning for cost opt...Job scheduling in hybrid cloud using deep reinforcement learning for cost opt...
Job scheduling in hybrid cloud using deep reinforcement learning for cost opt...
ArchanaKalapgar
 
Key Challenges In Today’S Dynamic Data Center
Key Challenges In Today’S Dynamic Data CenterKey Challenges In Today’S Dynamic Data Center
Key Challenges In Today’S Dynamic Data Center
Birendra Gosai
 
Best Practices for Building Successful Cloud Projects
Best Practices for Building Successful Cloud ProjectsBest Practices for Building Successful Cloud Projects
Best Practices for Building Successful Cloud Projects
Nati Shalom
 
CPET- Project Report
CPET- Project ReportCPET- Project Report
CPET- Project Report
Akshat Kumar Vaish
 
Lecture 15.ppt
Lecture 15.pptLecture 15.ppt
Lecture 15.ppt
YesuRaju8
 
Migrating enterprise applications to cloud
Migrating enterprise applications to cloudMigrating enterprise applications to cloud
Migrating enterprise applications to cloud
Sougata Mitra
 
CloudFIT_CSF_in_cloud_aug16
CloudFIT_CSF_in_cloud_aug16CloudFIT_CSF_in_cloud_aug16
CloudFIT_CSF_in_cloud_aug16
Dennis. Lee
 
VIRTUAL MACHINE SCHEDULING IN CLOUD COMPUTING ENVIRONMENT
VIRTUAL MACHINE SCHEDULING IN CLOUD COMPUTING ENVIRONMENTVIRTUAL MACHINE SCHEDULING IN CLOUD COMPUTING ENVIRONMENT
VIRTUAL MACHINE SCHEDULING IN CLOUD COMPUTING ENVIRONMENT
ijmpict
 
SERVICE LEVEL AGREEMENT BASED FAULT TOLERANT WORKLOAD SCHEDULING IN CLOUD COM...
SERVICE LEVEL AGREEMENT BASED FAULT TOLERANT WORKLOAD SCHEDULING IN CLOUD COM...SERVICE LEVEL AGREEMENT BASED FAULT TOLERANT WORKLOAD SCHEDULING IN CLOUD COM...
SERVICE LEVEL AGREEMENT BASED FAULT TOLERANT WORKLOAD SCHEDULING IN CLOUD COM...
ijgca
 
SERVICE LEVEL AGREEMENT BASED FAULT TOLERANT WORKLOAD SCHEDULING IN CLOUD COM...
SERVICE LEVEL AGREEMENT BASED FAULT TOLERANT WORKLOAD SCHEDULING IN CLOUD COM...SERVICE LEVEL AGREEMENT BASED FAULT TOLERANT WORKLOAD SCHEDULING IN CLOUD COM...
SERVICE LEVEL AGREEMENT BASED FAULT TOLERANT WORKLOAD SCHEDULING IN CLOUD COM...
ijgca
 
SERVICE LEVEL AGREEMENT BASED FAULT TOLERANT WORKLOAD SCHEDULING IN CLOUD COM...
SERVICE LEVEL AGREEMENT BASED FAULT TOLERANT WORKLOAD SCHEDULING IN CLOUD COM...SERVICE LEVEL AGREEMENT BASED FAULT TOLERANT WORKLOAD SCHEDULING IN CLOUD COM...
SERVICE LEVEL AGREEMENT BASED FAULT TOLERANT WORKLOAD SCHEDULING IN CLOUD COM...
ijgca
 
Software Association of Oregon Cloud Computing Presentation
Software Association of Oregon Cloud Computing PresentationSoftware Association of Oregon Cloud Computing Presentation
Software Association of Oregon Cloud Computing Presentation
ddcarr
 
Achieving Cost and Resource Efficiency through Docker, OpenShift and Kubernetes
Achieving Cost and Resource Efficiency through Docker, OpenShift and KubernetesAchieving Cost and Resource Efficiency through Docker, OpenShift and Kubernetes
Achieving Cost and Resource Efficiency through Docker, OpenShift and Kubernetes
Dean Delamont
 
Best practices for application migration to public clouds interop presentation
Best practices for application migration to public clouds interop presentationBest practices for application migration to public clouds interop presentation
Best practices for application migration to public clouds interop presentation
esebeus
 
Cloud computing managing
Cloud computing managingCloud computing managing
Cloud computing managing
Universiti Putra Malaysia
 
Conference Paper: Simulating High Availability Scenarios in Cloud Data Center...
Conference Paper: Simulating High Availability Scenarios in Cloud Data Center...Conference Paper: Simulating High Availability Scenarios in Cloud Data Center...
Conference Paper: Simulating High Availability Scenarios in Cloud Data Center...
Ericsson
 
Energy-Efficient Task Scheduling in Cloud Environment
Energy-Efficient Task Scheduling in Cloud EnvironmentEnergy-Efficient Task Scheduling in Cloud Environment
Energy-Efficient Task Scheduling in Cloud Environment
IRJET Journal
 
Gomez Blazing Fast Cloud Best Practices
Gomez Blazing Fast Cloud Best Practices Gomez Blazing Fast Cloud Best Practices
Gomez Blazing Fast Cloud Best Practices
Compuware APM
 

Similar to Resource provisioning optimization in cloud computing (20)

Above the Clouds: A Berkeley View of Cloud Computing: Paper Review
Above the Clouds: A Berkeley View of Cloud Computing:  Paper Review Above the Clouds: A Berkeley View of Cloud Computing:  Paper Review
Above the Clouds: A Berkeley View of Cloud Computing: Paper Review
 
NCMS UberCloud Experiment Webinar .
NCMS UberCloud Experiment Webinar .NCMS UberCloud Experiment Webinar .
NCMS UberCloud Experiment Webinar .
 
Job scheduling in hybrid cloud using deep reinforcement learning for cost opt...
Job scheduling in hybrid cloud using deep reinforcement learning for cost opt...Job scheduling in hybrid cloud using deep reinforcement learning for cost opt...
Job scheduling in hybrid cloud using deep reinforcement learning for cost opt...
 
Key Challenges In Today’S Dynamic Data Center
Key Challenges In Today’S Dynamic Data CenterKey Challenges In Today’S Dynamic Data Center
Key Challenges In Today’S Dynamic Data Center
 
Best Practices for Building Successful Cloud Projects
Best Practices for Building Successful Cloud ProjectsBest Practices for Building Successful Cloud Projects
Best Practices for Building Successful Cloud Projects
 
CPET- Project Report
CPET- Project ReportCPET- Project Report
CPET- Project Report
 
Lecture 15.ppt
Lecture 15.pptLecture 15.ppt
Lecture 15.ppt
 
Migrating enterprise applications to cloud
Migrating enterprise applications to cloudMigrating enterprise applications to cloud
Migrating enterprise applications to cloud
 
CloudFIT_CSF_in_cloud_aug16
CloudFIT_CSF_in_cloud_aug16CloudFIT_CSF_in_cloud_aug16
CloudFIT_CSF_in_cloud_aug16
 
VIRTUAL MACHINE SCHEDULING IN CLOUD COMPUTING ENVIRONMENT
VIRTUAL MACHINE SCHEDULING IN CLOUD COMPUTING ENVIRONMENTVIRTUAL MACHINE SCHEDULING IN CLOUD COMPUTING ENVIRONMENT
VIRTUAL MACHINE SCHEDULING IN CLOUD COMPUTING ENVIRONMENT
 
SERVICE LEVEL AGREEMENT BASED FAULT TOLERANT WORKLOAD SCHEDULING IN CLOUD COM...
SERVICE LEVEL AGREEMENT BASED FAULT TOLERANT WORKLOAD SCHEDULING IN CLOUD COM...SERVICE LEVEL AGREEMENT BASED FAULT TOLERANT WORKLOAD SCHEDULING IN CLOUD COM...
SERVICE LEVEL AGREEMENT BASED FAULT TOLERANT WORKLOAD SCHEDULING IN CLOUD COM...
 
SERVICE LEVEL AGREEMENT BASED FAULT TOLERANT WORKLOAD SCHEDULING IN CLOUD COM...
SERVICE LEVEL AGREEMENT BASED FAULT TOLERANT WORKLOAD SCHEDULING IN CLOUD COM...SERVICE LEVEL AGREEMENT BASED FAULT TOLERANT WORKLOAD SCHEDULING IN CLOUD COM...
SERVICE LEVEL AGREEMENT BASED FAULT TOLERANT WORKLOAD SCHEDULING IN CLOUD COM...
 
SERVICE LEVEL AGREEMENT BASED FAULT TOLERANT WORKLOAD SCHEDULING IN CLOUD COM...
SERVICE LEVEL AGREEMENT BASED FAULT TOLERANT WORKLOAD SCHEDULING IN CLOUD COM...SERVICE LEVEL AGREEMENT BASED FAULT TOLERANT WORKLOAD SCHEDULING IN CLOUD COM...
SERVICE LEVEL AGREEMENT BASED FAULT TOLERANT WORKLOAD SCHEDULING IN CLOUD COM...
 
Software Association of Oregon Cloud Computing Presentation
Software Association of Oregon Cloud Computing PresentationSoftware Association of Oregon Cloud Computing Presentation
Software Association of Oregon Cloud Computing Presentation
 
Achieving Cost and Resource Efficiency through Docker, OpenShift and Kubernetes
Achieving Cost and Resource Efficiency through Docker, OpenShift and KubernetesAchieving Cost and Resource Efficiency through Docker, OpenShift and Kubernetes
Achieving Cost and Resource Efficiency through Docker, OpenShift and Kubernetes
 
Best practices for application migration to public clouds interop presentation
Best practices for application migration to public clouds interop presentationBest practices for application migration to public clouds interop presentation
Best practices for application migration to public clouds interop presentation
 
Cloud computing managing
Cloud computing managingCloud computing managing
Cloud computing managing
 
Conference Paper: Simulating High Availability Scenarios in Cloud Data Center...
Conference Paper: Simulating High Availability Scenarios in Cloud Data Center...Conference Paper: Simulating High Availability Scenarios in Cloud Data Center...
Conference Paper: Simulating High Availability Scenarios in Cloud Data Center...
 
Energy-Efficient Task Scheduling in Cloud Environment
Energy-Efficient Task Scheduling in Cloud EnvironmentEnergy-Efficient Task Scheduling in Cloud Environment
Energy-Efficient Task Scheduling in Cloud Environment
 
Gomez Blazing Fast Cloud Best Practices
Gomez Blazing Fast Cloud Best Practices Gomez Blazing Fast Cloud Best Practices
Gomez Blazing Fast Cloud Best Practices
 

Recently uploaded

Driving Business Innovation: Latest Generative AI Advancements & Success Story
Driving Business Innovation: Latest Generative AI Advancements & Success StoryDriving Business Innovation: Latest Generative AI Advancements & Success Story
Driving Business Innovation: Latest Generative AI Advancements & Success Story
Safe Software
 
みなさんこんにちはこれ何文字まで入るの?40文字以下不可とか本当に意味わからないけどこれ限界文字数書いてないからマジでやばい文字数いけるんじゃないの?えこ...
みなさんこんにちはこれ何文字まで入るの?40文字以下不可とか本当に意味わからないけどこれ限界文字数書いてないからマジでやばい文字数いけるんじゃないの?えこ...みなさんこんにちはこれ何文字まで入るの?40文字以下不可とか本当に意味わからないけどこれ限界文字数書いてないからマジでやばい文字数いけるんじゃないの?えこ...
みなさんこんにちはこれ何文字まで入るの?40文字以下不可とか本当に意味わからないけどこれ限界文字数書いてないからマジでやばい文字数いけるんじゃないの?えこ...
名前 です男
 
Climate Impact of Software Testing at Nordic Testing Days
Climate Impact of Software Testing at Nordic Testing DaysClimate Impact of Software Testing at Nordic Testing Days
Climate Impact of Software Testing at Nordic Testing Days
Kari Kakkonen
 
Choosing The Best AWS Service For Your Website + API.pptx
Choosing The Best AWS Service For Your Website + API.pptxChoosing The Best AWS Service For Your Website + API.pptx
Choosing The Best AWS Service For Your Website + API.pptx
Brandon Minnick, MBA
 
Programming Foundation Models with DSPy - Meetup Slides
Programming Foundation Models with DSPy - Meetup SlidesProgramming Foundation Models with DSPy - Meetup Slides
Programming Foundation Models with DSPy - Meetup Slides
Zilliz
 
20240607 QFM018 Elixir Reading List May 2024
20240607 QFM018 Elixir Reading List May 202420240607 QFM018 Elixir Reading List May 2024
20240607 QFM018 Elixir Reading List May 2024
Matthew Sinclair
 
Removing Uninteresting Bytes in Software Fuzzing
Removing Uninteresting Bytes in Software FuzzingRemoving Uninteresting Bytes in Software Fuzzing
Removing Uninteresting Bytes in Software Fuzzing
Aftab Hussain
 
Monitoring and Managing Anomaly Detection on OpenShift.pdf
Monitoring and Managing Anomaly Detection on OpenShift.pdfMonitoring and Managing Anomaly Detection on OpenShift.pdf
Monitoring and Managing Anomaly Detection on OpenShift.pdf
Tosin Akinosho
 
Infrastructure Challenges in Scaling RAG with Custom AI models
Infrastructure Challenges in Scaling RAG with Custom AI modelsInfrastructure Challenges in Scaling RAG with Custom AI models
Infrastructure Challenges in Scaling RAG with Custom AI models
Zilliz
 
Uni Systems Copilot event_05062024_C.Vlachos.pdf
Uni Systems Copilot event_05062024_C.Vlachos.pdfUni Systems Copilot event_05062024_C.Vlachos.pdf
Uni Systems Copilot event_05062024_C.Vlachos.pdf
Uni Systems S.M.S.A.
 
HCL Notes and Domino License Cost Reduction in the World of DLAU
HCL Notes and Domino License Cost Reduction in the World of DLAUHCL Notes and Domino License Cost Reduction in the World of DLAU
HCL Notes and Domino License Cost Reduction in the World of DLAU
panagenda
 
Serial Arm Control in Real Time Presentation
Serial Arm Control in Real Time PresentationSerial Arm Control in Real Time Presentation
Serial Arm Control in Real Time Presentation
tolgahangng
 
Best 20 SEO Techniques To Improve Website Visibility In SERP
Best 20 SEO Techniques To Improve Website Visibility In SERPBest 20 SEO Techniques To Improve Website Visibility In SERP
Best 20 SEO Techniques To Improve Website Visibility In SERP
Pixlogix Infotech
 
20240605 QFM017 Machine Intelligence Reading List May 2024
20240605 QFM017 Machine Intelligence Reading List May 202420240605 QFM017 Machine Intelligence Reading List May 2024
20240605 QFM017 Machine Intelligence Reading List May 2024
Matthew Sinclair
 
Taking AI to the Next Level in Manufacturing.pdf
Taking AI to the Next Level in Manufacturing.pdfTaking AI to the Next Level in Manufacturing.pdf
Taking AI to the Next Level in Manufacturing.pdf
ssuserfac0301
 
UI5 Controls simplified - UI5con2024 presentation
UI5 Controls simplified - UI5con2024 presentationUI5 Controls simplified - UI5con2024 presentation
UI5 Controls simplified - UI5con2024 presentation
Wouter Lemaire
 
AI-Powered Food Delivery Transforming App Development in Saudi Arabia.pdf
AI-Powered Food Delivery Transforming App Development in Saudi Arabia.pdfAI-Powered Food Delivery Transforming App Development in Saudi Arabia.pdf
AI-Powered Food Delivery Transforming App Development in Saudi Arabia.pdf
Techgropse Pvt.Ltd.
 
Video Streaming: Then, Now, and in the Future
Video Streaming: Then, Now, and in the FutureVideo Streaming: Then, Now, and in the Future
Video Streaming: Then, Now, and in the Future
Alpen-Adria-Universität
 
CAKE: Sharing Slices of Confidential Data on Blockchain
CAKE: Sharing Slices of Confidential Data on BlockchainCAKE: Sharing Slices of Confidential Data on Blockchain
CAKE: Sharing Slices of Confidential Data on Blockchain
Claudio Di Ciccio
 
“I’m still / I’m still / Chaining from the Block”
“I’m still / I’m still / Chaining from the Block”“I’m still / I’m still / Chaining from the Block”
“I’m still / I’m still / Chaining from the Block”
Claudio Di Ciccio
 

Recently uploaded (20)

Driving Business Innovation: Latest Generative AI Advancements & Success Story
Driving Business Innovation: Latest Generative AI Advancements & Success StoryDriving Business Innovation: Latest Generative AI Advancements & Success Story
Driving Business Innovation: Latest Generative AI Advancements & Success Story
 
みなさんこんにちはこれ何文字まで入るの?40文字以下不可とか本当に意味わからないけどこれ限界文字数書いてないからマジでやばい文字数いけるんじゃないの?えこ...
みなさんこんにちはこれ何文字まで入るの?40文字以下不可とか本当に意味わからないけどこれ限界文字数書いてないからマジでやばい文字数いけるんじゃないの?えこ...みなさんこんにちはこれ何文字まで入るの?40文字以下不可とか本当に意味わからないけどこれ限界文字数書いてないからマジでやばい文字数いけるんじゃないの?えこ...
みなさんこんにちはこれ何文字まで入るの?40文字以下不可とか本当に意味わからないけどこれ限界文字数書いてないからマジでやばい文字数いけるんじゃないの?えこ...
 
Climate Impact of Software Testing at Nordic Testing Days
Climate Impact of Software Testing at Nordic Testing DaysClimate Impact of Software Testing at Nordic Testing Days
Climate Impact of Software Testing at Nordic Testing Days
 
Choosing The Best AWS Service For Your Website + API.pptx
Choosing The Best AWS Service For Your Website + API.pptxChoosing The Best AWS Service For Your Website + API.pptx
Choosing The Best AWS Service For Your Website + API.pptx
 
Programming Foundation Models with DSPy - Meetup Slides
Programming Foundation Models with DSPy - Meetup SlidesProgramming Foundation Models with DSPy - Meetup Slides
Programming Foundation Models with DSPy - Meetup Slides
 
20240607 QFM018 Elixir Reading List May 2024
20240607 QFM018 Elixir Reading List May 202420240607 QFM018 Elixir Reading List May 2024
20240607 QFM018 Elixir Reading List May 2024
 
Removing Uninteresting Bytes in Software Fuzzing
Removing Uninteresting Bytes in Software FuzzingRemoving Uninteresting Bytes in Software Fuzzing
Removing Uninteresting Bytes in Software Fuzzing
 
Monitoring and Managing Anomaly Detection on OpenShift.pdf
Monitoring and Managing Anomaly Detection on OpenShift.pdfMonitoring and Managing Anomaly Detection on OpenShift.pdf
Monitoring and Managing Anomaly Detection on OpenShift.pdf
 
Infrastructure Challenges in Scaling RAG with Custom AI models
Infrastructure Challenges in Scaling RAG with Custom AI modelsInfrastructure Challenges in Scaling RAG with Custom AI models
Infrastructure Challenges in Scaling RAG with Custom AI models
 
Uni Systems Copilot event_05062024_C.Vlachos.pdf
Uni Systems Copilot event_05062024_C.Vlachos.pdfUni Systems Copilot event_05062024_C.Vlachos.pdf
Uni Systems Copilot event_05062024_C.Vlachos.pdf
 
HCL Notes and Domino License Cost Reduction in the World of DLAU
HCL Notes and Domino License Cost Reduction in the World of DLAUHCL Notes and Domino License Cost Reduction in the World of DLAU
HCL Notes and Domino License Cost Reduction in the World of DLAU
 
Serial Arm Control in Real Time Presentation
Serial Arm Control in Real Time PresentationSerial Arm Control in Real Time Presentation
Serial Arm Control in Real Time Presentation
 
Best 20 SEO Techniques To Improve Website Visibility In SERP
Best 20 SEO Techniques To Improve Website Visibility In SERPBest 20 SEO Techniques To Improve Website Visibility In SERP
Best 20 SEO Techniques To Improve Website Visibility In SERP
 
20240605 QFM017 Machine Intelligence Reading List May 2024
20240605 QFM017 Machine Intelligence Reading List May 202420240605 QFM017 Machine Intelligence Reading List May 2024
20240605 QFM017 Machine Intelligence Reading List May 2024
 
Taking AI to the Next Level in Manufacturing.pdf
Taking AI to the Next Level in Manufacturing.pdfTaking AI to the Next Level in Manufacturing.pdf
Taking AI to the Next Level in Manufacturing.pdf
 
UI5 Controls simplified - UI5con2024 presentation
UI5 Controls simplified - UI5con2024 presentationUI5 Controls simplified - UI5con2024 presentation
UI5 Controls simplified - UI5con2024 presentation
 
AI-Powered Food Delivery Transforming App Development in Saudi Arabia.pdf
AI-Powered Food Delivery Transforming App Development in Saudi Arabia.pdfAI-Powered Food Delivery Transforming App Development in Saudi Arabia.pdf
AI-Powered Food Delivery Transforming App Development in Saudi Arabia.pdf
 
Video Streaming: Then, Now, and in the Future
Video Streaming: Then, Now, and in the FutureVideo Streaming: Then, Now, and in the Future
Video Streaming: Then, Now, and in the Future
 
CAKE: Sharing Slices of Confidential Data on Blockchain
CAKE: Sharing Slices of Confidential Data on BlockchainCAKE: Sharing Slices of Confidential Data on Blockchain
CAKE: Sharing Slices of Confidential Data on Blockchain
 
“I’m still / I’m still / Chaining from the Block”
“I’m still / I’m still / Chaining from the Block”“I’m still / I’m still / Chaining from the Block”
“I’m still / I’m still / Chaining from the Block”
 

Resource provisioning optimization in cloud computing

  • 1. Dynamic Optimization of Heterogeneous Resource Provisioning in Cloud Computing Masoumeh Tajvidi Supervisors: A/Prof. Michael Maher Dr. Daryl Essam School of Engineering and Information Technology (SEIT)
  • 2. • Introduction - Cloud computing and its potential benefits - Challenges! • Research question • Related Work • Research Plan • Preliminary Work Outline 1/1 6
  • 3. What is Cloud Computing? A broad and deep platform that helps customers build sophisticated, scalable applications 2/1 6
  • 4. Why are companies adopting cloud computing so quickly? Agility 3/1 6
  • 5. Main benefits of utility-like computing Convert CAPEX to OPEX  Lower Total Cost No Need to Guess Capacity Lower Maintenance Cost 4/1 6
  • 6. Effectively using of cloud computing  Over-provisioning: purchased resources are not fully utilized cost more than necessary  Under-provisioning: purchased resources are not sufficient to meet the actual demand  hurts application performance, 5/1 6
  • 7. Efficient resource provisioning: challenge #1 : Various cloud providers offering multiple Virtual Machine (VM) types • Virtualization technologies help providers pack their hardware resources into different type of Virtual Machines • The end-users encounter a complicated decision making problem for choosing the right mix of VMs! 6/1 6
  • 8. Efficient resource provisioning: challenge #2 : Multiple pricing models 7/1 6
  • 9. Efficient resource provisioning: challenge #3 : Deal with COST and DEMAND uncertainty  The application’s demand is not known in advance • e.g. online video streaming applications, like YouTube Channel  The cost of instances in both on- demand and spot is varying! 8/1 6
  • 10. Efficient resource provisioning: challenge #4 : Multi-objective problem  Cost is not the only objective for this problem  The trade-off among cost and QoS must also be handled  For example response time is very critical in latency sensitive applications , online gaming services, as the users tend to be very impatient 9/1 6
  • 11. Research Question ? How to dynamically optimize cloud resource provisioning plan as a multi-objective problem in the real-world? 10/1 6
  • 13. stochastic optimization  The time at which parameters become known, divide the problem into stages  Main Question: what to decide in the first stage?  Main question: how many VM to reserve  Recourse: how to deal with under or over provisioning  Objective function is to minimize the expected cost 12/1 6
  • 14.  Add complexities of the real-world problem into our model. Take into account the heterogeneity of VM types Take into account all three pricing schemes into our model Solve the problem as a multi-objective optimization problem  Dealing with uncertainty by solving the problem with both stochastic and approximate programming approaches  enhance the performance of the optimization problem by using Machine Learning techniques Research Plan : Big Picture Challenge #1 Challenge #2 Challenge #4 Challenge #3 13/1 6
  • 15. Preliminary results  Replicated one of the available multi stage programming approach  Modelled the problem in Stochastic MiniZinc modelling language  Added spot-instance pricing model into our model 14/1 6
  • 16. Stochastic MiniZinc Model Results Before and After Adding Spot Instances 15/1 6

Editor's Notes

  1. Today, What I am going to talk about is as follows:
  2. In the simplest term: cloud computing is the provision of IT resources on-demand using a pay as you go model over the internet
  3. So why are companies adopting cloud computing so quickly? There is a very simple answer to this, which is that by using the cloud you can inject some much-needed agility into your organization. So resource will be available to you when you need them and you can control them through a simple interface and when you don’t need those resources you no longer have to pay for them. 3. You can simply Plug in to the cloud and pay as you use, which reminds us of traditional utilities such as water and electricity!!
  4. 1. Breaking that down a little more, the four areas the customers get most benefit from when they are using cloud computing are as follows: 2. firstly they can convert CAPEX to OPEX, so you no longer have this upfront investment required to buy and build your IT infrastructure rather than that you pay as you go for the services you consume 3. And also you would normally find that by using the cloud your total cost of operation over ownership will be lower. 4. You no longer need to guess the amount of capacity but you need to run your application since the environment is elastic in nature and you can access as much or as little resources you need in order to support your application , you can scale up and down 5. The cloud provider is responsible for maintenance of the infrastructure and security, that can remove the need for you to a lot of staff and help you run a more lean organization which is more focused meeting the needs of your customers (business) 6. There are all very fascinating for the small growing companies or start ups: because they arent heavily invensted on IT infrustruvture , if you want to try something new, go ahead and try it, if it works fantastic , scale it up and improve it! It doesn’t work you can stop your experiment Capital expenditures (CAPEX) and operating expenses (OPEX) represent two basic categories of business expenses
  5. Lets have a look at the traditional infrastructure model and its problems: Explin over provisioning and under provisioning 3. But With cloud compuitng we can match resources to demand as this figure 4. But for getting to such a idealistic resource provisioingn model, the cloud user needs to smartly plan the resource provisioning problem. Which is not that much easy due to available challenges that I am going to talk about them in next slides
  6. Choosing the right mix
  7. Reserved instances the cheapest On-demand instances the highest price Spot Instances allow you to specify the maximum hourly price that you are willing to pay to run a particular instance type, usually lower than the On-Demand rate. 4. (the spot instances are the unused on-demand instances) The Spot Price fluctuates based on supply and demand for instances, but customers will never pay more than the maximum price they have specified. If the Spot Price moves higher than a customer’s maximum price, the customer’s instance will be shut down by the cloud provider
  8. Therefore the customer is not aware of future price of resources. This uncertainty should be considered, since it can affect the final decision on resource provisioning problem.
  9. Last but not the least challenge is that we can not minimize the cost and neglect all other QoS!
  10. These are challenges in order to answer my research question
  11. Little research has been conducted from the cloud end-user view. Most of them focused on deterministic formulation over fixed parametrs where the problem has a perfect foresight. Those that have considered uncertainty in their problem typically just considered one objective function or only one or two pricing models. The techniques : for cloud resource provisionig and explain the difference between stochastic and approximate programming and why we have chosen stochastic? THE DIFFERENCES?
  12. Stochastic optimization consider the problem as some parameters are unknown The problem is divided by some stages, based on the time at which parameters become known 2-stage problem is something like this: we have first stage decision and second stage decision which also called recourse .. The main question is…. We want to know what to do before an stochastic event occurs The objective function is minimizing the expected cost considering all uncertain parameters as different scenarios
  13. Two of the main contenders in the methodology of optimization under uncertainty are “stochastic programming” and “Approximate dynamic programming.” Both take a probabilistic approach to uncertainty and contemplate not just one decision and one observation, but a possible interplay between decisions and observations in stages. Therefore we will solve the problem with these two approaches and compare their results to find out which one is more appropriate for our problem. In order to enhance the performance of our optimization problem, a machine learning technique integrates into this complex optimization problem, such as reinforcement learning technique, to help it learn while optimizing..
  14. What results we get so far?