New Scheduling Algorithms: Improving Fairness and Quality of Service

Slides:



Advertisements
Similar presentations
School of Computing FACULTY OF ENGINEERING Grids and QoS Grid Computing has emerged in the last two decades, initially as a model for large-scale, resource-intensive.
Advertisements

Coordination Mechanisms for Unrelated Machine Scheduling Yossi Azar joint work with Kamal Jain Vahab Mirrokni.
Real-Time Competitive Environments: Truthful Mechanisms for Allocating a Single Processor to Sporadic Tasks Anwar Mohammadi, Nathan Fisher, and Daniel.
Energy-efficient Task Scheduling in Heterogeneous Environment 2013/10/25.
Mafijul Islam, PhD Software Systems, Electrical and Embedded Systems Advanced Technology & Research Research Issues in Computing Systems: An Automotive.
Class-constrained Packing Problems with Application to Storage Management in Multimedia Systems Tami Tamir Department of Computer Science The Technion.
Thomas Moscibroda Distributed Systems Research, Redmond Onur Mutlu
1 Presenter: Chien-Chih Chen. 2 Dynamic Scheduler for Multi-core Systems Analysis of The Linux 2.6 Kernel Scheduler Optimal Task Scheduler for Multi-core.
CALTECH CS137 Winter DeHon CS137: Electronic Design Automation Day 14: March 3, 2004 Scheduling Heuristics and Approximation.
S ELFISH M IGRATE : A Scalable Algorithm for Non-clairvoyantly Scheduling Heterogeneous Processors Janardhan Kulkarni, Duke University Sungjin Im (UC Merced.
SCALABLE PARALLEL COMPUTING ON CLOUDS : EFFICIENT AND SCALABLE ARCHITECTURES TO PERFORM PLEASINGLY PARALLEL, MAPREDUCE AND ITERATIVE DATA INTENSIVE COMPUTATIONS.
An Overview of the BSP Model of Parallel Computation Overview Only.
Atomistic Protein Folding Simulations on the Submillisecond Timescale Using Worldwide Distributed Computing Qing Lu CMSC 838 Presentation.
On-line adaptive parallel prefix computation Jean-Louis Roch, Daouda Traoré and Julien Bernard Presented by Andreas Söderström, ITN.
EDA (CS286.5b) Day 11 Scheduling (List, Force, Approximation) N.B. no class Thursday (FPGA) …
1 Scheduling on Heterogeneous Machines: Minimize Total Energy + Flowtime Ravishankar Krishnaswamy Carnegie Mellon University Joint work with Anupam Gupta.
1 Scheduling Jobs with Varying Parallelizability Ravishankar Krishnaswamy Carnegie Mellon University.
Little Demonstration of the Power in Discovery Jason Hill, Steve Ross David E. Culler Computer Science Division U.C. Berkeley.
By- Jaideep Moses, Ravi Iyer , Ramesh Illikkal and
1 Energy Efficient Communication in Wireless Sensor Networks Yingyue Xu 8/14/2015.
Distributed Systems (15-440) Mohammad Hammoud December 4 th, 2013.
Course Outline DayContents Day 1 Introduction Motivation, definitions, properties of embedded systems, outline of the current course How to specify embedded.
Parallel Processing CS453 Lecture 2.  The role of parallelism in accelerating computing speeds has been recognized for several decades.  Its role in.
NSF Critical Infrastructures Workshop Nov , 2006 Kannan Ramchandran University of California at Berkeley Current research interests related to workshop.
CIS4930/CDA5125 Parallel and Distributed Systems Florida State University CIS4930/CDA5125: Parallel and Distributed Systems Instructor: Xin Yuan, 168 Love,
Extreme scale parallel and distributed systems – High performance computing systems Current No. 1 supercomputer Tianhe-2 at petaflops Pushing toward.
CALTECH CS137 Winter DeHon CS137: Electronic Design Automation Day 12: February 13, 2002 Scheduling Heuristics and Approximation.
1 Server Scheduling in the L p norm Nikhil Bansal (CMU) Kirk Pruhs (Univ. of Pittsburgh)
Computational Thinking in K-12 and Scalable Game Design Michael Shuffett.
A Maiden Analysis of Longest Wait First Jeff Edmonds York University Kirk Pruhs University of Pittsburgh.
©Ian Sommerville 2000Software Engineering, 6th edition. Chapter 1 Slide 1 Chapter 1 Introduction.
By Garrett Kelly. 3 types or reasons for distributed applications Data Data used by the application is distributed Computation Computation is distributed.
Summary of Distributed Computing Security Yifeng Zou Georgia State University
Computer Organization and Architecture Tutorial 1 Kenneth Lee.
Fluid Software: Handling Heterogeneous Many-Core for Programmer Productivity Nate Clark.
Service-oriented Resource Broker for QoS-Guaranteed in Grid Computing System Yichao Yang, Jin Wu, Lei Lang, Yanbo Zhou and Zhili Sun Centre for communication.
1 Uwe Schwiegelshohn, 2 Andrei Tchernykh, 1 Ramin Yahyapour 1 Technische Universität Dortmund, Germany
Rassul Ayani 1 Performance of parallel and distributed systems  What is the purpose of measurement?  To evaluate a system (or an architecture)  To compare.
DynamicMR: A Dynamic Slot Allocation Optimization Framework for MapReduce Clusters Nanyang Technological University Shanjiang Tang, Bu-Sung Lee, Bingsheng.
Data Structures and Algorithms in Parallel Computing Lecture 7.
Server HW CSIS 4490 n-Tier Client/Server Dr. Hoganson Server Hardware Mission-critical –High reliability –redundancy Massive storage (disk) –RAID for redundancy.
Hybrid Multi-Core Architecture for Boosting Single-Threaded Performance Presented by: Peyman Nov 2007.
Hierarchies, Clouds, and Specialization Phillip B. Gibbons Intel Labs Pittsburgh June 28, 2012 NSF Workshop on Research Directions in the Principles of.
Multicast Pull Scheduling Kirk Pruhs. The Big Problem Movie Distribution Database Replication via Internet Harry Potter Book Download Software Download.
1 Cache-Oblivious Query Processing Bingsheng He, Qiong Luo {saven, Department of Computer Science & Engineering Hong Kong University of.
Load Rebalancing for Distributed File Systems in Clouds.
Scheduling Parallel DAG Jobs to Minimize the Average Flow Time K. Agrawal, J. Li, K. Lu, B. Moseley.
Spark on Entropy : A Reliable & Efficient Scheduler for Low-latency Parallel Jobs in Heterogeneous Cloud Huankai Chen PhD Student at University of Kent.
Panel: Beyond Exascale Computing
Use of Cloud Computing for Implementation of e-Governance Services
Parallel Computing in the Multicore Era
Data and Applications Security Developments and Directions
Computing Resource Allocation and Scheduling in A Data Center
2008 Team Building Workshop
Advanced Operating Systems – Fall 2009
FROM STRATEGY TO TOTAL REDESIGN
By: Greg Boyarko, Jordan Sutton, and Shaun Parkison
Parallel Computing in the Multicore Era
University of Pittsburgh
2008 Team Building Workshop
University of Pittsburgh
2008 Team Building Workshop
Data and Applications Security Developments and Directions
Data and Applications Security Developments and Directions
Enabling Contribution Awareness in an Overlay Broadcasting System
What I've done in past 6 months
Data and Applications Security Developments and Directions
Distributed Systems (15-440)
Data and Applications Security Developments and Directions
Research: Past, Present and Future
Presentation transcript:

New Scheduling Algorithms: Improving Fairness and Quality of Service [1] Sungjin Im, Benjamin Moseley, Kirk Pruhs: A tutorial on amortized local competitiveness in online scheduling. SIGACT News 42(2): 83-97 (2011) [2] Sungjin Im, Benjamin Moseley: An online scalable algorithm for average flow time in broadcast scheduling. ACM Transactions on Algorithms 8(4): 39 (2012) [3] Kyle Fox, Sungjin Im, Benjamin Moseley: Energy Efficient Scheduling of Parallelizable Jobs. SODA 2013: 948-957 Schedule 1 optimizes the average flow time and schedule 2 optimizes the L2-norm of the flow times. The example shows how the first scheduler can be unfair as compared to the second scheduler. Develop new algorithmic and analysis techniques to discover scheduling algorithms with strong worst-case performance guarantees Optimize quality of service objectives that enforce both fairness to clients and enforce good quality of service Discover fundamental algorithmic ideas that transcend different objectives and environments Introduce analysis techniques that are widely applicable Improve upon the state of the art performance guarantees Introduce models for scheduling environments in new and emerging technologies Heterogenous parallel processor environments Distributed computing Energy aware scheduling for more sustainable computing In the past, most work focus on settings where all machines are identical. In the future, it is thought that heterogeneous environments will be the dominant architecture. This is so that slow low power energy efficient machines can process less critical jobs. Scheduling in heterogenous environments is challenging and requires sophisticated algorithmic techniques