Self-Adaptive QoS Guarantees and Optimization in Clouds Jim (Zhanwen) Li (Carleton University) Murray Woodside (Carleton University) John Chinneck (Carleton.

Slides:



Advertisements
Similar presentations
1/17/20141 Leveraging Cloudbursting To Drive Down IT Costs Eric Burgener Senior Vice President, Product Marketing March 9, 2010.
Advertisements

Performance Testing - Kanwalpreet Singh.
Hadi Goudarzi and Massoud Pedram
SLA-Oriented Resource Provisioning for Cloud Computing
Amazon. Cloud computing also known as on-demand computing or utility computing. Similar to other utility providers like electric, water, and natural gas,
1 Vladimir Knežević Microsoft Software d.o.o.. 80% Održavanje 80% Održavanje 20% New Cost Reduction Keep Business Up & Running End User Productivity End.
CLOUD COMPUTING AN OVERVIEW & QUALITY OF SERVICE Hamzeh Khazaei University of Manitoba Department of Computer Science Jan 28, 2010.
Green Cloud Computing Hadi Salimi Distributed Systems Lab, School of Computer Engineering, Iran University of Science and Technology,
VMware Virtualization Last Update Copyright Kenneth M. Chipps Ph.D.
Efficient Autoscaling in the Cloud using Predictive Models for Workload Forecasting Roy, N., A. Dubey, and A. Gokhale 4th IEEE International Conference.
Copyright 2009 FUJITSU TECHNOLOGY SOLUTIONS PRIMERGY Servers and Windows Server® 2008 R2 Benefit from an efficient, high performance and flexible platform.
Microsoft Virtual Server 2005 Product Overview Mikael Nyström – TrueSec AB MVP Windows Server – Setup/Deployment Mikael Nyström – TrueSec AB MVP Windows.
LDU Parametrized Discrete-Time Multivariable MRAC and Application to A Web Cache System Ying Lu, Gang Tao and Tarek Abdelzaher University of Virginia.
© 2009 IBM Corporation ® IBM Software Group Introduction to Cloud Computing Vivek C Agarwal IBM India Software Labs.
Cloud Computing (101).
© 2008 Hewlett-Packard Development Company, L.P. The information contained herein is subject to change without notice Automated Workload Management in.
Presented by Sujit Tilak. Evolution of Client/Server Architecture Clients & Server on different computer systems Local Area Network for Server and Client.
CloudCmp: Shopping for a Cloud Made Easy Ang Li Xiaowei Yang Duke University Srikanth Kandula Ming Zhang Microsoft Research 6/22/2010HotCloud 2010, Boston1.
New Challenges in Cloud Datacenter Monitoring and Management
Module 2: Information Technology Infrastructure
SOFTWARE AS A SERVICE PLATFORM AS A SERVICE INFRASTRUCTURE AS A SERVICE.
Plan Introduction What is Cloud Computing?
Cloud Attributes Business Challenges Influence Your IT Solutions Business to IT Conversation Microsoft is Changing too Supporting System Center In House.
Security Difficulties of E-Learning in Cloud Computing
CLOUD COMPUTING. A general term for anything that involves delivering hosted services over the Internet. And Cloud is referred to the hardware and software.
VAP What is a Virtual Application ? A virtual application is an application that has been optimized to run on virtual infrastructure. The application software.
CLOUD COMPUTING & COST MANAGEMENT S. Gurubalasubramaniyan, MSc IT, MTech Presented by.
Introduction to Cloud Computing
Osama Shahid ( ) Vishal ( ) BSCS-5B
Cloud computing is the use of computing resources (hardware and software) that are delivered as a service over the Internet. Cloud is the metaphor for.
Dynamic and Decentralized Approaches for Optimal Allocation of Multiple Resources in Virtualized Data Centers Wei Chen, Samuel Hargrove, Heh Miao, Liang.
Cloud Computing. What is Cloud Computing? Cloud computing is a model for enabling convenient, on-demand network access to a shared pool of configurable.
+ CS 325: CS Hardware and Software Organization and Architecture Cloud Architectures.
Cloud Computing 1. Outline  Introduction  Evolution  Cloud architecture  Map reduce operation  Platform 2.
Internet Information Services 7.0 Infrastructure Planning and Design Series.
1 ©2009 Desktone, Inc. All rights reserved. Desktops in the Cloud: It’s not Virtual Desktop Infrastructure (VDI) Danny Allan, Chief Solution Architect.
November , 2009SERVICE COMPUTATION 2009 Analysis of Energy Efficiency in Clouds H. AbdelSalamK. Maly R. MukkamalaM. Zubair Department.
Introduction to Cloud Computing
Adaptive software in cloud computing Marin Litoiu York University Canada.
An Autonomic Framework in Cloud Environment Jiedan Zhu Advisor: Prof. Gagan Agrawal.
LOGO Service and network administration Storage Virtualization.
Plan  Introduction  What is Cloud Computing?  Why is it called ‘’Cloud Computing’’?  Characteristics of Cloud Computing  Advantages of Cloud Computing.
Using Virtual Servers for the CERN Windows infrastructure Emmanuel Ormancey, Alberto Pace CERN, Information Technology Department.
Server Virtualization
VMware vSphere Configuration and Management v6
RESERVOIR RESERVOIR Resources and Services Virtualization without Barriers Philippe Massonet (CETIC)
Enterprise Cloud Computing
June 30 - July 2, 2009AIMS 2009 Towards Energy Efficient Change Management in A Cloud Computing Environment: A Pro-Active Approach H. AbdelSalamK. Maly.
Cloud Computing, Overview and Challenges Marin Litoiu, York University.
Chapter 8 – Cloud Computing
03/03/051 Performance Engineering of Software and Distributed Systems Research Activities at IIT Bombay Varsha Apte March 3 rd, 2005.
AFACT Cloud Computing WG Zon-yin Shae Institute for Information Industry Bangkok, Thailand, Nov. 26, 2014.
CISC 849 : Applications in Fintech Namami Shukla Dept of Computer & Information Sciences University of Delaware A Cloud Computing Methodology Study of.
Web Technologies Lecture 13 Introduction to cloud computing.
Information Systems in Organizations 5.2 Cloud Computing.
RANDY MODOWSKI COSC Cloud Computing. Road Map What is Cloud Computing? History of “The Cloud” Cloud Milestones How Cloud Computing is being used.
Cloud Computing ENG. YOUSSEF ABDELHAKIM. Agenda :  The definitions of Cloud Computing.  Examples of Cloud Computing.  Which companies are using Cloud.
© 2012 Eucalyptus Systems, Inc. Cloud Computing Introduction Eucalyptus Education Services 2.
Cofax Scalability Document Version Scaling Cofax in General The scalability of Cofax is directly related to the system software, hardware and network.
Spark on Entropy : A Reliable & Efficient Scheduler for Low-latency Parallel Jobs in Heterogeneous Cloud Huankai Chen PhD Student at University of Kent.
Prof. Jong-Moon Chung’s Lecture Notes at Yonsei University
Unit 3 Virtualization.
Chapter 6: Securing the Cloud
Understanding The Cloud
”The Ball” Radical Cloud Resource Consolidation
AWS. Introduction AWS launched in 2006 from the internal infrastructure that Amazon.com built to handle its online retail operations. AWS was one of the.
CNIT131 Internet Basics & Beginning HTML
Cloud Computing: Concepts
Cloud Computing Erasmus+ Project
How Dell, SAP and SUSE Deliver Value Quickly
Presentation transcript:

Self-Adaptive QoS Guarantees and Optimization in Clouds Jim (Zhanwen) Li (Carleton University) Murray Woodside (Carleton University) John Chinneck (Carleton University) Marin Litoiu (York University)

Outline (CERAS) Cloud Overview Optimization for Clouds: Definition Optimization Method Case Studies 2

Centre of Excellence for Research in Adaptive Systems Participants IBM Ontario Government Ontario Cancer Institute UofWaterloo, UofToronto, Queen’s, Carleton, York University Three complementary research thrusts to enable cloud computing Service and resource virtualization: what do we offer in a cloud? Programming models for web services: how do we add value? Adaptive computing: how do we manage the cloud? Deliverables A cloud infrastructure (CERAS Cloud) Algorithms and methods to manage the cloud infrastructure Services Tools in cloud Desktops Web services and applications Demonstrate how emerging application can be run more effectively in a cloud infrastructure

Cloud = SaaS+Virtualization … Delivering software functionality online, similar to the one installed on your machine. Flavours Infrastructure as a Service Software as a Service Platform as a Service Desktop as a Service 4 – Pricing models pay per usage ( Amazon) pay a subscription (Microsoft and Salesforce) pay per transaction (Expedia) use it for free (Google) DesktopOffice Databases OS Network Database Storage CPU Servers Web servers ERP CRM Software dev tools

Management in CERAS Cloud 5

Two Deployment Scenarios 6

Outline (CERAS) Cloud Overview Optimization for Clouds: Definition Optimization Method Case Studies 7

Optimization for Clouds Necessities Economic: cost and profit; time: installation, maintenance Quality of maintenance and configurations Challenges Scalability and complexity: thousands to millions of various decisions. Service selection, Service deployment, Workloads balance Optimization must be efficient enough for real-time management. Guaranteed QoS: SLA to workloads and components, software and hardware delays Constraints from system and business capacity and availability of resources and budgets Interaction of configurations Dynamic in virtualization 8

Feedback Control for QoS and Optimization 9 1.Karman Filter for estimation and prediction 2.Quantitative model : Performance Model, 3.Qualitative model: an Optimization Model that can be solved effectively and efficiently 4.Execute Optimal decisions

Model-based Optimization Architecture 10

Service System Metamodel 11

Outline (CERAS) Cloud Overview Optimization for Clouds: Definition Optimization Method Case Studies 12

Scalable Network Flow Model 13 NFM presents the states of the system as an Optimization Model Parameters of the NFM are updated at runtime Scalable to millions of configurations and decisions Applications: Decisions among replication, migration, on or off etc Workloads: Dynamic workload management Resources: License/memory/CPU requirements and availability Costs (or profits): penalization and rewards QoS management

Optimization Loop 14 1.Network optimization allocates the flows to optimize costs and meet average delay constraints, without knowledge of contention delay 2.LQN performance model calculates contention delay 3.Contention delay is inserted into the network model and allocation is iterated 4.Result is an allocation that minimizes costs and meets delay constraint, including contention.

Outline (CERAS) Cloud Overview Optimization for Clouds: Definition Optimization Method Case Studies 15

Simplified Cost Model Assumption: no request cycles Objective Function Constraints 16

Case Study I: Min RT, Min Cost 17

Case Study II: Scalability Consider a cloud with many services all structured like this one 18 A fragment of the network flow model

Case Study II cont 19 Full optimization can save around 20% hosts in useFull optimization can significantly save costs however, full optimization may increase the cost of contentions. Utilizations are increased to the desired upper bound in Full Optimization

Conclusions The combination of NFM and nonlinear performance model Effectively optimizing many interacted configurations subject to quite a few QoS and economic constraints New optimization algorithms Scalability, Efficiency, Flexibility, Autonomic Tuning Full optimization is best but it is less practical Risks and overhead In practice, cloud administrators will settle with incremental optimization and launch full optimization when the COST becomes high 20

21