Network-aware Job Scheduling for the HPC Cloud

Save this PDF as:
 WORD  PNG  TXT  JPG

Size: px
Start display at page:

Download "Network-aware Job Scheduling for the HPC Cloud"

Transcription

1 16 th Workshop on Sustained Simulation Performance (WSSP) Network-aware Job Scheduling for the HPC Cloud 2012/12/11 Yoshitomo Murata Cyberscience Center, Tohoku University

2 Agenda Background Vector Computing Cloud Network-aware Job Scheduling Evaluation Conclusions 2012/12/11 16th Workshop on Sustained Simulation Performance (WSSP) 2

3 Background High Performance Computing (HPC) requires more computing resources. HPC is used in many fields such as research and development. Distributed computing systems provide high-computing power to HPC users Multi computing resources coordinate on resolving large-scale problems. Ex) Grid Computing, Volunteer Computing 2012/12/11 16th Workshop on Sustained Simulation Performance (WSSP) 3

4 Distributed HPC Systems TOHOKU Univ. SX-9 OSAKA Univ. SX-9 16TFLOPS 2012/12/11 16th Workshop on Sustained Simulation Performance (WSSP) 4

5 Vector Computing Cloud Virtualized Vector Resource TOHOKU Univ. SX-9 OSAKA Univ. SX-9 16TFLOPS 2012/12/11 16th Workshop on Sustained Simulation Performance (WSSP) 5

6 Advantages of Vector Computing Cloud Load-Balancing The submitted job is automatically assigned to low-load node. High utilization of multiple systems is realized. Realizing Virtual Parallel Computer The virtual parallel computer is not limited hardware configuration of each site. High-Load Low-Load No Low-Load Load Different Site Submit parallel job Submit single job Single node is available Execution time time is is long. short. Improving the flexibility of job execution and resource management 2012/12/11 16th Workshop on Sustained Simulation Performance (WSSP) 6

7 Composition of Wide-Area Vector co-operation Environment Two supercomputer sites Two SX-9 nodes at Tohoku University Two SX-9 nodes at Osaka University Two high-speed wide area network SINET-3 (1Gbps) JGN2Plus (10Gbps) 2012/12/11 16th Workshop on Sustained Simulation Performance (WSSP) 7

8 Gflop/s Potential of Wide-Area Vector co-operation Environment - HPL Benchmark MPI/SX (IXS) GridMPI (IXS) GridMPI (JF) GridMPI (NF) GridMPI (SINET3) GridMPI (JGN2Plus) ,254 2,780 2,913 1,149 2, , Inhouse Performance Wide area co-operation Performance 1,330 1,517 1,683 Efficiency 51.3% 1.6TFlop/s single node Peak performance 1 Problem Size N 2012/12/11 16th Workshop on Sustained Simulation Performance (WSSP) 8

9 Potential of Wide-Area Vector co-operation Environment - Real Application The fluid flow around a cylinder is simulated by a particle method on the wide-area environment. Execution Time [sec.] Prob. size # of node Wide-area local 16 x x 32 1 x 1 3, x 2 3, x 1 20,096 7,675 2 x 2 24,995 4,406 The visualized results shows the vortexes behind the cylinder. The wide-area vector co-operation has the ability for executing MPI programs successfully. 2012/12/11 16th Workshop on Sustained Simulation Performance (WSSP) 9

10 Job Scheduling on Vector Computing Cloud Scheduling Problem There is no standard to select an appropriate site from distributed resources. History-based job scheduling The execution time of parallel application is limited by the performance of wide-area network. To reduce the limitation by the wide-are network, new network selection mechanism is required. 2012/12/11 16th Workshop on Sustained Simulation Performance (WSSP) 10

11 History-based Job Scheduling Scheduling Problem Difference in the site operation policies Osaka University: Reservation-based Guarantee of the job-start time Time restriction of the job-execution Tohoku University: Queuing-based High usability of the computing resources No guarantee of the job-start time No fairness and efficiency in the job scheduling between sites. 2012/12/11 16th Workshop on Sustained Simulation Performance (WSSP) 11

12 History-based Job Scheduling Assigning a job to the site which has least waiting-time Achieving fairness and efficiency resource selection 2012/12/11 16th Workshop on Sustained Simulation Performance (WSSP) 12

13 Job Scheduling on Vector Computing Cloud Scheduling Problem There is no standard to select an appropriate site from distributed resources. History-based job scheduling The execution time of parallel application is limited by the performance of wide-area network. To reduce the limitation by the wide-are network, new network selection mechanism is required. Network-aware job scheduling 2012/12/11 16th Workshop on Sustained Simulation Performance (WSSP) 13

14 NETWORK-AWARE JOB SCHEDULING 2012/12/11 16th Workshop on Sustained Simulation Performance (WSSP) 14

15 Network-aware Job scheduling Objective Realizing the node and network allocation to minimize the influence of network in the wide-area co-operation environment. Approach Modeling the execution time of parallel-job Considering the network performance Designing the node and network allocation mechanism 2012/12/11 16th Workshop on Sustained Simulation Performance (WSSP) 15

16 Modeling the Job Execution Time Presupposition The changes of execution time depends on only network performance Vector Computing Cloud consists of same kind of vector processors Equation T exec = T comp + T comm T comp : The time spent for computation (constant) T comm : The time spent for communication (variable) 2012/12/11 16th Workshop on Sustained Simulation Performance (WSSP) 16

17 Modeling the Network Communication Time Presuppositions The communication time is decided on only the network bandwidth and latency. the larger size of data is, the longer the communication time is. the more count of communication is, the longer the communication time is. Equation T comm = C band F band + C lat F lat C band : The normalized communication time regarding network bandwidth C lat : The normalized communication time regarding network latency F band : Factor of network bandwidth F lat : Factor of network latency ( local network has 1.0, wide area network has more than 1.0) 2012/12/11 16th Workshop on Sustained Simulation Performance (WSSP) 17

18 Processes of Job Scheduling Initial Step: Scheduler push the network information F band and F lat of each path to DB. Site A Site B Site C Waiting Time 10 minutes 1 hour 1 hour Start: User submits a job. job Network-aware job scheduler DB Step1: Scheduler retrieves the profiling information T comp, C band and C lat from DB. History-based job scheduler Step2: History-based job scheduler estimates the waiting time of each node. Step3: Scheduler makes combinations of nodes which can start processes at the same time. Step4: Scheduler retrieves the network information F band and F lat corresponding to above node combinations. Step5: Scheduler calculates the make span of each combination, and selects the combination which has the shortest make span. 2012/12/11 16th Workshop on Sustained Simulation Performance (WSSP) 18

19 Job Assignment Mechanism job scheduler Application Profile DB Job submiting Resource Requirement Network Requirement OpenFlow Network Routing 2012/12/11 16th Workshop on Sustained Simulation Performance (WSSP) 19

20 PERFORMANCE EVALUATION 2012/12/11 16th Workshop on Sustained Simulation Performance (WSSP) 20

21 Simulation Parameters (1) Parameter Value T comp ( frequency ) [sec.] 100(50%), 200(25%), 400(12.5%), 800(6.3%), 1600(3.1%), 3200(1.6%), 6400(0.8%), 12800(0.4%), 25600(0.2%), 51200(0.1%) C band T comp x uniform[0.1, 0.5] C lat T comp x uniform[0.1, 0.5] Concurrency of Parallel Job 1(single), 2 Occurrence Rate of Parallel Job [%] 0, 1, 5, 10, 25, 50 Job Generation Number of Sites 3 Number of Nodes per Site 2 Scheduling Method 120 jobs always exist Round-Robin, History-based, Proposal Total Simulation Time[sec.] 1,000, /12/11 16th Workshop on Sustained Simulation Performance (WSSP) 21

22 Simulation Parameters (2) VPU VPU Site A VPU VPU VPU VPU Site B Site C Connection F band F lat Intra-Site A B 2 4 B C 4 2 C A /12/11 16th Workshop on Sustained Simulation Performance (WSSP) 22

23 Number of Jobs The Number of Executed Jobs 12,000 Round-Robin History-based Proposal 10,000 8,000 6,000 4,000 2, Occurrence Rate of Parallel Job [%] Proposal improves the number of executed jobs 2012/10/12 ADVNET

24 Rate of the Idle Time [%] The Idle Time of Node 25 Round-Robin History-based Proposal Occurrence Rate of Parallel Job [%] The idle time of nodes are caused by the startup synchronization of two processes assigned to different nodes. 2012/12/11 16th Workshop on Sustained Simulation Performance (WSSP) 24

25 Waiting Time [sec.] Waiting Time [sec.] The Distribution of the Waiting Time Average waiting time Round-Robin History-based Proposal Maximum waiting time Round-Robin History-based Proposal 30, ,000 25, ,000 20,000 80,000 15,000 60,000 10,000 40,000 5,000 20, Occurrence Rate of Parallel Job [%] Occurrence Rate of Parallel Job [%] The load-balancing by proposal realizes smallest variation in waiting time. 2012/12/11 16th Workshop on Sustained Simulation Performance (WSSP) 25

26 Future Plan The network configuration of Vector Computing Cloud is being updated. Updating SINET3 to SINET4(10Gbps) Updating JGN2Plus to JGN-X (10Gbps, OpenFlow support) There are multi path between Tohoku University and Osaka University. Plan of experiment Evaluating the influence of bandwidth and latency on the communication time of various parallel applications. Improving the accuracy of job execution time model 2012/12/11 16th Workshop on Sustained Simulation Performance (WSSP) 26

27 Conclusions Vector Computing Cloud Vector Computing Cloud improves the flexibility of job execution and resource management The wide-area vector co-operation has the potential for executing parallel job. Network-aware Job Scheduling Modeling the execution time of parallel-job Designing the resource allocation mechanism The simulation results show that the Network-aware job scheduling realizes the appropriate resource allocation in Vector Computing cloud. 2012/12/11 16th Workshop on Sustained Simulation Performance (WSSP) 27

28 Thank you 2012/12/11 16th Workshop on Sustained Simulation Performance (WSSP) 28

High Performance Computing in CST STUDIO SUITE

High Performance Computing in CST STUDIO SUITE High Performance Computing in CST STUDIO SUITE Felix Wolfheimer GPU Computing Performance Speedup 18 16 14 12 10 8 6 4 2 0 Promo offer for EUC participants: 25% discount for K40 cards Speedup of Solver

More information

High Performance Computing Infrastructure in JAPAN

High Performance Computing Infrastructure in JAPAN High Performance Computing Infrastructure in JAPAN Yutaka Ishikawa Director of Information Technology Center Professor of Computer Science Department University of Tokyo Leader of System Software Research

More information

Methodology for predicting the energy consumption of SPMD application on virtualized environments *

Methodology for predicting the energy consumption of SPMD application on virtualized environments * Methodology for predicting the energy consumption of SPMD application on virtualized environments * Javier Balladini, Ronal Muresano +, Remo Suppi +, Dolores Rexachs + and Emilio Luque + * Computer Engineering

More information

Amazon Cloud Performance Compared. David Adams

Amazon Cloud Performance Compared. David Adams Amazon Cloud Performance Compared David Adams Amazon EC2 performance comparison How does EC2 compare to traditional supercomputer for scientific applications? "Performance Analysis of High Performance

More information

SX-ACE Processor: NEC s Brand-New Vector Processor

SX-ACE Processor: NEC s Brand-New Vector Processor HOTCHIPS26 SX-ACE Processor: NEC s Brand-New Vector Processor Shintaro Momose, Ph.D. NEC Corporation IT Platform Division Manager of SX vector supercomputer development August 11th, 2014 Performance SX

More information

Grid Computing Approach for Dynamic Load Balancing

Grid Computing Approach for Dynamic Load Balancing International Journal of Computer Sciences and Engineering Open Access Review Paper Volume-4, Issue-1 E-ISSN: 2347-2693 Grid Computing Approach for Dynamic Load Balancing Kapil B. Morey 1*, Sachin B. Jadhav

More information

Building a Top500-class Supercomputing Cluster at LNS-BUAP

Building a Top500-class Supercomputing Cluster at LNS-BUAP Building a Top500-class Supercomputing Cluster at LNS-BUAP Dr. José Luis Ricardo Chávez Dr. Humberto Salazar Ibargüen Dr. Enrique Varela Carlos Laboratorio Nacional de Supercómputo Benemérita Universidad

More information

A Load Balancing Algorithm based on the Variation Trend of Entropy in Homogeneous Cluster

A Load Balancing Algorithm based on the Variation Trend of Entropy in Homogeneous Cluster , pp.11-20 http://dx.doi.org/10.14257/ ijgdc.2014.7.2.02 A Load Balancing Algorithm based on the Variation Trend of Entropy in Homogeneous Cluster Kehe Wu 1, Long Chen 2, Shichao Ye 2 and Yi Li 2 1 Beijing

More information

PARALLEL & CLUSTER COMPUTING CS 6260 PROFESSOR: ELISE DE DONCKER BY: LINA HUSSEIN

PARALLEL & CLUSTER COMPUTING CS 6260 PROFESSOR: ELISE DE DONCKER BY: LINA HUSSEIN 1 PARALLEL & CLUSTER COMPUTING CS 6260 PROFESSOR: ELISE DE DONCKER BY: LINA HUSSEIN Introduction What is cluster computing? Classification of Cluster Computing Technologies: Beowulf cluster Construction

More information

Performance of the JMA NWP models on the PC cluster TSUBAME.

Performance of the JMA NWP models on the PC cluster TSUBAME. Performance of the JMA NWP models on the PC cluster TSUBAME. K.Takenouchi 1), S.Yokoi 1), T.Hara 1) *, T.Aoki 2), C.Muroi 1), K.Aranami 1), K.Iwamura 1), Y.Aikawa 1) 1) Japan Meteorological Agency (JMA)

More information

HIGH PERFORMANCE COMPUTING COMPETENCE CENTER BADEN-WÜRTTEMBERG

HIGH PERFORMANCE COMPUTING COMPETENCE CENTER BADEN-WÜRTTEMBERG HIGH PERFORMANCE COMPUTING COMPETENCE CENTER BADEN-WÜRTTEMBERG Contents High Performance Computing Competence Center Baden-Württemberg (hkz-bw)... 4 Vector Parallel Supercomputer NEC SX-6X... 8 Massively

More information

Software of the Earth Simulator

Software of the Earth Simulator Journal of the Earth Simulator, Volume 3, September 2005, 52 59 Software of the Earth Simulator Atsuya Uno The Earth Simulator Center, Japan Agency for Marine-Earth Science and Technology, 3173 25 Showa-machi,

More information

Pros and Cons of HPC Cloud Computing

Pros and Cons of HPC Cloud Computing CloudStat 211 Pros and Cons of HPC Cloud Computing Nils gentschen Felde Motivation - Idea HPC Cluster HPC Cloud Cluster Management benefits of virtual HPC Dynamical sizing / partitioning Loadbalancing

More information

Overlapping Data Transfer With Application Execution on Clusters

Overlapping Data Transfer With Application Execution on Clusters Overlapping Data Transfer With Application Execution on Clusters Karen L. Reid and Michael Stumm reid@cs.toronto.edu stumm@eecg.toronto.edu Department of Computer Science Department of Electrical and Computer

More information

Datacenter Wide-area Enterprise

Datacenter Wide-area Enterprise Datacenter Wide-area Enterprise Client LOAD-BALANCER Can t choose path : ( Servers Outline and goals A new architecture for distributed load-balancing joint (server, path) selection Demonstrate a nation-wide

More information

Interaction of Access Patterns on the dnfsp File System Rodrigo Virote Kassick Francieli Zanon Boito Philippe O.A. Navaux

Interaction of Access Patterns on the dnfsp File System Rodrigo Virote Kassick Francieli Zanon Boito Philippe O.A. Navaux [ ] Interaction of Access Patterns on the dnfsp File System Rodrigo Virote Kassick Francieli Zanon Boito Philippe O.A. Navaux GPPD Conferencia Latinoamericana de Computación de Alto Rendimiento 2009 --

More information

Mixed Precision Iterative Refinement Methods Energy Efficiency on Hybrid Hardware Platforms

Mixed Precision Iterative Refinement Methods Energy Efficiency on Hybrid Hardware Platforms Mixed Precision Iterative Refinement Methods Energy Efficiency on Hybrid Hardware Platforms Björn Rocker Hamburg, June 17th 2010 Engineering Mathematics and Computing Lab (EMCL) KIT University of the State

More information

Operating System for the K computer

Operating System for the K computer Operating System for the K computer Jun Moroo Masahiko Yamada Takeharu Kato For the K computer to achieve the world s highest performance, Fujitsu has worked on the following three performance improvements

More information

Marcos Dias de Assunção 1,2, Alexandre di Costanzo 1 and Rajkumar Buyya 1

Marcos Dias de Assunção 1,2, Alexandre di Costanzo 1 and Rajkumar Buyya 1 Marcos Dias de Assunção 1,2, Alexandre di Costanzo 1 and Rajkumar Buyya 1 1 Department of Computer Science and Software Engineering 2 National ICT Australia (NICTA) Victoria Research Laboratory The University

More information

Supercomputing applied to Parallel Network Simulation

Supercomputing applied to Parallel Network Simulation Supercomputing applied to Parallel Network Simulation David Cortés-Polo Research, Technological Innovation and Supercomputing Centre of Extremadura, CenitS. Trujillo, Spain david.cortes@cenits.es Summary

More information

Agenda. HPC Software Stack. HPC Post-Processing Visualization. Case Study National Scientific Center. European HPC Benchmark Center Montpellier PSSC

Agenda. HPC Software Stack. HPC Post-Processing Visualization. Case Study National Scientific Center. European HPC Benchmark Center Montpellier PSSC HPC Architecture End to End Alexandre Chauvin Agenda HPC Software Stack Visualization National Scientific Center 2 Agenda HPC Software Stack Alexandre Chauvin Typical HPC Software Stack Externes LAN Typical

More information

System Models for Distributed and Cloud Computing

System Models for Distributed and Cloud Computing System Models for Distributed and Cloud Computing Dr. Sanjay P. Ahuja, Ph.D. 2010-14 FIS Distinguished Professor of Computer Science School of Computing, UNF Classification of Distributed Computing Systems

More information

Cost-Benefit Analysis of Cloud Computing versus Desktop Grids

Cost-Benefit Analysis of Cloud Computing versus Desktop Grids Cost-Benefit Analysis of Cloud Computing versus Desktop Grids Derrick Kondo, Bahman Javadi, Paul Malécot, Franck Cappello INRIA, France David P. Anderson UC Berkeley, USA Cloud Background Vision Hide complexity

More information

Workshop on Parallel and Distributed Scientific and Engineering Computing, Shanghai, 25 May 2012

Workshop on Parallel and Distributed Scientific and Engineering Computing, Shanghai, 25 May 2012 Scientific Application Performance on HPC, Private and Public Cloud Resources: A Case Study Using Climate, Cardiac Model Codes and the NPB Benchmark Suite Peter Strazdins (Research School of Computer Science),

More information

QUALITY OF SERVICE METRICS FOR DATA TRANSMISSION IN MESH TOPOLOGIES

QUALITY OF SERVICE METRICS FOR DATA TRANSMISSION IN MESH TOPOLOGIES QUALITY OF SERVICE METRICS FOR DATA TRANSMISSION IN MESH TOPOLOGIES SWATHI NANDURI * ZAHOOR-UL-HUQ * Master of Technology, Associate Professor, G. Pulla Reddy Engineering College, G. Pulla Reddy Engineering

More information

Clusters: Mainstream Technology for CAE

Clusters: Mainstream Technology for CAE Clusters: Mainstream Technology for CAE Alanna Dwyer HPC Division, HP Linux and Clusters Sparked a Revolution in High Performance Computing! Supercomputing performance now affordable and accessible Linux

More information

Volunteer Computing, Grid Computing and Cloud Computing: Opportunities for Synergy. Derrick Kondo INRIA, France

Volunteer Computing, Grid Computing and Cloud Computing: Opportunities for Synergy. Derrick Kondo INRIA, France Volunteer Computing, Grid Computing and Cloud Computing: Opportunities for Synergy Derrick Kondo INRIA, France Outline Cloud Grid Volunteer Computing Cloud Background Vision Hide complexity of hardware

More information

An Architecture Model of Sensor Information System Based on Cloud Computing

An Architecture Model of Sensor Information System Based on Cloud Computing An Architecture Model of Sensor Information System Based on Cloud Computing Pengfei You, Yuxing Peng National Key Laboratory for Parallel and Distributed Processing, School of Computer Science, National

More information

Operations Management Software for the K computer

Operations Management Software for the K computer Operations Management Software for the K computer Kouichi Hirai Yuji Iguchi Atsuya Uno Motoyoshi Kurokawa Supercomputer systems have been increasing steadily in scale (number of CPU cores and number of

More information

Fair Scheduling Algorithm with Dynamic Load Balancing Using In Grid Computing

Fair Scheduling Algorithm with Dynamic Load Balancing Using In Grid Computing Research Inventy: International Journal Of Engineering And Science Vol.2, Issue 10 (April 2013), Pp 53-57 Issn(e): 2278-4721, Issn(p):2319-6483, Www.Researchinventy.Com Fair Scheduling Algorithm with Dynamic

More information

Unleashing the Performance Potential of GPUs for Atmospheric Dynamic Solvers

Unleashing the Performance Potential of GPUs for Atmospheric Dynamic Solvers Unleashing the Performance Potential of GPUs for Atmospheric Dynamic Solvers Haohuan Fu haohuan@tsinghua.edu.cn High Performance Geo-Computing (HPGC) Group Center for Earth System Science Tsinghua University

More information

Collaborative Project in Cloud Computing

Collaborative Project in Cloud Computing Collaborative Project in Cloud Computing Project Title: Virtual Cloud Laboratory (VCL) Services for Education Transfer of Results: from North Carolina State University (NCSU) to IBM Technical Description:

More information

Symmetric Multiprocessing

Symmetric Multiprocessing Multicore Computing A multi-core processor is a processing system composed of two or more independent cores. One can describe it as an integrated circuit to which two or more individual processors (called

More information

Scalable and High Performance Computing for Big Data Analytics in Understanding the Human Dynamics in the Mobile Age

Scalable and High Performance Computing for Big Data Analytics in Understanding the Human Dynamics in the Mobile Age Scalable and High Performance Computing for Big Data Analytics in Understanding the Human Dynamics in the Mobile Age Xuan Shi GRA: Bowei Xue University of Arkansas Spatiotemporal Modeling of Human Dynamics

More information

Performance Evaluation of Amazon EC2 for NASA HPC Applications!

Performance Evaluation of Amazon EC2 for NASA HPC Applications! National Aeronautics and Space Administration Performance Evaluation of Amazon EC2 for NASA HPC Applications! Piyush Mehrotra!! J. Djomehri, S. Heistand, R. Hood, H. Jin, A. Lazanoff,! S. Saini, R. Biswas!

More information

A Proposed Service Broker Strategy in CloudAnalyst for Cost-Effective Data Center Selection

A Proposed Service Broker Strategy in CloudAnalyst for Cost-Effective Data Center Selection A Proposed Service Broker Strategy in CloudAnalyst for Cost-Effective Selection Dhaval Limbani*, Bhavesh Oza** *(Department of Information Technology, S. S. Engineering College, Bhavnagar) ** (Department

More information

2. Research and Development on the Autonomic Operation. Control Infrastructure Technologies in the Cloud Computing Environment

2. Research and Development on the Autonomic Operation. Control Infrastructure Technologies in the Cloud Computing Environment R&D supporting future cloud computing infrastructure technologies Research and Development on Autonomic Operation Control Infrastructure Technologies in the Cloud Computing Environment DEMPO Hiroshi, KAMI

More information

Microsoft Technical Computing The Advancement of Parallelism. Tom Quinn, Technical Computing Partner Manager

Microsoft Technical Computing The Advancement of Parallelism. Tom Quinn, Technical Computing Partner Manager Presented at the COMSOL Conference 2010 Boston Microsoft Technical Computing The Advancement of Parallelism Tom Quinn, Technical Computing Partner Manager 21 1.2 x 10 New Bytes of Information in 2010 Source:

More information

Cloud-WIEN2k. A Scientific Cloud Computing Platform for Condensed Matter Physics

Cloud-WIEN2k. A Scientific Cloud Computing Platform for Condensed Matter Physics Penn State, August 2013 Cloud-WIEN2k A Scientific Cloud Computing Platform for Condensed Matter Physics K. Jorissen University of Washington, Seattle, U.S.A. Supported by NSF grant OCI-1048052 www.feffproject.org

More information

Volunteer Computing and Cloud Computing: Opportunities for Synergy

Volunteer Computing and Cloud Computing: Opportunities for Synergy Volunteer Computing and Cloud Computing: Opportunities for Synergy Derrick Kondo INRIA, France Performance vs. Reliability vs. Costs high Cost Reliability high low low low Performance high Performance

More information

A Hybrid Load Balancing Policy underlying Cloud Computing Environment

A Hybrid Load Balancing Policy underlying Cloud Computing Environment A Hybrid Load Balancing Policy underlying Cloud Computing Environment S.C. WANG, S.C. TSENG, S.S. WANG*, K.Q. YAN* Chaoyang University of Technology 168, Jifeng E. Rd., Wufeng District, Taichung 41349

More information

COLO: COarse-grain LOck-stepping Virtual Machine for Non-stop Service

COLO: COarse-grain LOck-stepping Virtual Machine for Non-stop Service COLO: COarse-grain LOck-stepping Virtual Machine for Non-stop Service Eddie Dong, Yunhong Jiang 1 Legal Disclaimer INFORMATION IN THIS DOCUMENT IS PROVIDED IN CONNECTION WITH INTEL PRODUCTS. NO LICENSE,

More information

LOAD BALANCING MECHANISMS IN DATA CENTER NETWORKS

LOAD BALANCING MECHANISMS IN DATA CENTER NETWORKS LOAD BALANCING Load Balancing Mechanisms in Data Center Networks Load balancing vs. distributed rate limiting: an unifying framework for cloud control Load Balancing for Internet Distributed Services using

More information

The Assessment of Benchmarks Executed on Bare-Metal and Using Para-Virtualisation

The Assessment of Benchmarks Executed on Bare-Metal and Using Para-Virtualisation The Assessment of Benchmarks Executed on Bare-Metal and Using Para-Virtualisation Mark Baker, Garry Smith and Ahmad Hasaan SSE, University of Reading Paravirtualization A full assessment of paravirtualization

More information

A PERFORMANCE COMPARISON USING HPC BENCHMARKS: WINDOWS HPC SERVER 2008 AND RED HAT ENTERPRISE LINUX 5

A PERFORMANCE COMPARISON USING HPC BENCHMARKS: WINDOWS HPC SERVER 2008 AND RED HAT ENTERPRISE LINUX 5 A PERFORMANCE COMPARISON USING HPC BENCHMARKS: WINDOWS HPC SERVER 2008 AND RED HAT ENTERPRISE LINUX 5 R. Henschel, S. Teige, H. Li, J. Doleschal, M. S. Mueller October 2010 Contents HPC at Indiana University

More information

TRUFFLE Broadband Bonding Network Appliance BBNA6401. A Frequently Asked Question on. Link Bonding vs. Load Balancing

TRUFFLE Broadband Bonding Network Appliance BBNA6401. A Frequently Asked Question on. Link Bonding vs. Load Balancing TRUFFLE Broadband Bonding Network Appliance BBNA6401 A Frequently Asked Question on Link Bonding vs. Load Balancing LBRvsBBNAFeb15_08b 1 Question: What's the difference between a Truffle Broadband Bonding

More information

Performance In the Cloud. White paper

Performance In the Cloud. White paper Hardware vs. Amazon EC2 Cloud Performance In the Cloud White paper September, 2009 Ron Warshawsky Eugene Skobel Copyright 2009 Enteros, Inc. All rights reserved. No part of this publication may be reproduced,

More information

Highly Productive HPC on Modern Vector Supercomputers: Current and Future

Highly Productive HPC on Modern Vector Supercomputers: Current and Future Highly Productive HPC on Modern Vector Supercomputers: Current and Future Hiroaki Kobayashi Director and Professor Cyberscience Center Tohoku University koba@cc.tohoku.ac.jp Moscow, Russia About Tohoku

More information

Gfarm: Present Status and Future Evolution

Gfarm: Present Status and Future Evolution OpenSFS APAC Lustre User Group 2013 Tokyo October 17, 2013 Gfarm: Present Status and Future Evolution Osamu Tatebe University of Tsukuba Gfarm file system Award-winning file system since 2000 Distributed

More information

Keywords: PDAs, VM. 2015, IJARCSSE All Rights Reserved Page 365

Keywords: PDAs, VM. 2015, IJARCSSE All Rights Reserved Page 365 Volume 5, Issue 7, July 2015 ISSN: 2277 128X International Journal of Advanced Research in Computer Science and Software Engineering Research Paper Available online at: www.ijarcsse.com Energy Adaptive

More information

Cluster Sanity Checks

Cluster Sanity Checks Cluster Sanity Checks Christian Terboven terboven@rz.rwth aachen.de Center for Computing and Communication RWTH Aachen University Windows HPC Deployment September 19, RWTH Aachen Agenda o Motivation o

More information

How to Deploy OpenStack on TH-2 Supercomputer Yusong Tan, Bao Li National Supercomputing Center in Guangzhou April 10, 2014

How to Deploy OpenStack on TH-2 Supercomputer Yusong Tan, Bao Li National Supercomputing Center in Guangzhou April 10, 2014 How to Deploy OpenStack on TH-2 Supercomputer Yusong Tan, Bao Li National Supercomputing Center in Guangzhou April 10, 2014 2014 年 云 计 算 效 率 与 能 耗 暨 第 一 届 国 际 云 计 算 咨 询 委 员 会 中 国 高 峰 论 坛 Contents Background

More information

Distribution transparency. Degree of transparency. Openness of distributed systems

Distribution transparency. Degree of transparency. Openness of distributed systems Distributed Systems Principles and Paradigms Maarten van Steen VU Amsterdam, Dept. Computer Science steen@cs.vu.nl Chapter 01: Version: August 27, 2012 1 / 28 Distributed System: Definition A distributed

More information

Cloud computing. Intelligent Services for Energy-Efficient Design and Life Cycle Simulation. as used by the ISES project

Cloud computing. Intelligent Services for Energy-Efficient Design and Life Cycle Simulation. as used by the ISES project Intelligent Services for Energy-Efficient Design and Life Cycle Simulation Project number: 288819 Call identifier: FP7-ICT-2011-7 Project coordinator: Technische Universität Dresden, Germany Website: ises.eu-project.info

More information

Using In-Memory Data Grids for Global Data Integration

Using In-Memory Data Grids for Global Data Integration SCALEOUT SOFTWARE Using In-Memory Data Grids for Global Data Integration by Dr. William Bain, ScaleOut Software, Inc. 2012 ScaleOut Software, Inc. 12/27/2012 B y enabling extremely fast and scalable data

More information

Internet, adat, biztonság, sebesség

Internet, adat, biztonság, sebesség Internet, adat, biztonság, sebesség Gacsal József Business Development Manager, Intel Hungary Ltd. 2013. április 9. HOUG Siófok Legal Information Today s presentations contain forward-looking statements.

More information

Current Status of FEFS for the K computer

Current Status of FEFS for the K computer Current Status of FEFS for the K computer Shinji Sumimoto Fujitsu Limited Apr.24 2012 LUG2012@Austin Outline RIKEN and Fujitsu are jointly developing the K computer * Development continues with system

More information

PRIMERGY server-based High Performance Computing solutions

PRIMERGY server-based High Performance Computing solutions PRIMERGY server-based High Performance Computing solutions PreSales - May 2010 - HPC Revenue OS & Processor Type Increasing standardization with shift in HPC to x86 with 70% in 2008.. HPC revenue by operating

More information

Interconnect. Jesús Labarta. Index

Interconnect. Jesús Labarta. Index Interconnect Jesús Labarta Index 1 Interconnection networks Need to send messages (commands/responses, message passing) Processors Memory Node Node Interconnection networks Components Links Switches Network

More information

High Performance Computing Infrastructure in Japan

High Performance Computing Infrastructure in Japan High Performance Computing Infrastructure in Japan Kento Aida National Institute of Informatics 2 Overview of HPCI Introduction n High Performance Computing Infrastructure (HPCI) Ø national project promoted

More information

TRUFFLE Broadband Bonding Network Appliance. A Frequently Asked Question on. Link Bonding vs. Load Balancing

TRUFFLE Broadband Bonding Network Appliance. A Frequently Asked Question on. Link Bonding vs. Load Balancing TRUFFLE Broadband Bonding Network Appliance A Frequently Asked Question on Link Bonding vs. Load Balancing 5703 Oberlin Dr Suite 208 San Diego, CA 92121 P:888.842.1231 F: 858.452.1035 info@mushroomnetworks.com

More information

Performance Characteristics of Large SMP Machines

Performance Characteristics of Large SMP Machines Performance Characteristics of Large SMP Machines Dirk Schmidl, Dieter an Mey, Matthias S. Müller schmidl@rz.rwth-aachen.de Rechen- und Kommunikationszentrum (RZ) Agenda Investigated Hardware Kernel Benchmark

More information

GPU System Architecture. Alan Gray EPCC The University of Edinburgh

GPU System Architecture. Alan Gray EPCC The University of Edinburgh GPU System Architecture EPCC The University of Edinburgh Outline Why do we want/need accelerators such as GPUs? GPU-CPU comparison Architectural reasons for GPU performance advantages GPU accelerated systems

More information

ALPS Supercomputing System A Scalable Supercomputer with Flexible Services

ALPS Supercomputing System A Scalable Supercomputer with Flexible Services ALPS Supercomputing System A Scalable Supercomputer with Flexible Services 1 Abstract Supercomputing is moving from the realm of abstract to mainstream with more and more applications and research being

More information

Agenda. Enterprise Application Performance Factors. Current form of Enterprise Applications. Factors to Application Performance.

Agenda. Enterprise Application Performance Factors. Current form of Enterprise Applications. Factors to Application Performance. Agenda Enterprise Performance Factors Overall Enterprise Performance Factors Best Practice for generic Enterprise Best Practice for 3-tiers Enterprise Hardware Load Balancer Basic Unix Tuning Performance

More information

for my computation? Stefano Cozzini Which infrastructure Which infrastructure Democrito and SISSA/eLAB - Trieste

for my computation? Stefano Cozzini Which infrastructure Which infrastructure Democrito and SISSA/eLAB - Trieste Which infrastructure Which infrastructure for my computation? Stefano Cozzini Democrito and SISSA/eLAB - Trieste Agenda Introduction:! E-infrastructure and computing infrastructures! What is available

More information

1 Bull, 2011 Bull Extreme Computing

1 Bull, 2011 Bull Extreme Computing 1 Bull, 2011 Bull Extreme Computing Table of Contents HPC Overview. Cluster Overview. FLOPS. 2 Bull, 2011 Bull Extreme Computing HPC Overview Ares, Gerardo, HPC Team HPC concepts HPC: High Performance

More information

Scheduling and Load Balancing in the Parallel ROOT Facility (PROOF)

Scheduling and Load Balancing in the Parallel ROOT Facility (PROOF) Scheduling and Load Balancing in the Parallel ROOT Facility (PROOF) Gerardo Ganis CERN E-mail: Gerardo.Ganis@cern.ch CERN Institute of Informatics, University of Warsaw E-mail: Jan.Iwaszkiewicz@cern.ch

More information

Scalable Source Routing

Scalable Source Routing Scalable Source Routing January 2010 Thomas Fuhrmann Department of Informatics, Self-Organizing Systems Group, Technical University Munich, Germany Routing in Networks You re there. I m here. Scalable

More information

Performance Test Process

Performance Test Process A white Success The performance testing helped the client identify and resolve performance bottlenecks which otherwise crippled the business. The ability to support 500 concurrent users was a performance

More information

Achieving Performance Isolation with Lightweight Co-Kernels

Achieving Performance Isolation with Lightweight Co-Kernels Achieving Performance Isolation with Lightweight Co-Kernels Jiannan Ouyang, Brian Kocoloski, John Lange The Prognostic Lab @ University of Pittsburgh Kevin Pedretti Sandia National Laboratories HPDC 2015

More information

Dynamic Network Resources Allocation in Grids through a Grid Network Resource Broker

Dynamic Network Resources Allocation in Grids through a Grid Network Resource Broker INGRID 2007 Instrumenting the GRID Second International Workshop on Distributed Cooperative Laboratories Session 2: Networking for the GRID Dynamic Network Resources Allocation in Grids through a Grid

More information

A Proposed Service Broker Policy for Data Center Selection in Cloud Environment with Implementation

A Proposed Service Broker Policy for Data Center Selection in Cloud Environment with Implementation A Service Broker Policy for Data Center Selection in Cloud Environment with Implementation Dhaval Limbani*, Bhavesh Oza** *(Department of Information Technology, S. S. Engineering College, Bhavnagar) **

More information

High Performance Applications over the Cloud: Gains and Losses

High Performance Applications over the Cloud: Gains and Losses High Performance Applications over the Cloud: Gains and Losses Dr. Leila Ismail Faculty of Information Technology United Arab Emirates University leila@uaeu.ac.ae http://citweb.uaeu.ac.ae/citweb/profile/leila

More information

Performance Tuning of a CFD Code on the Earth Simulator

Performance Tuning of a CFD Code on the Earth Simulator Applications on HPC Special Issue on High Performance Computing Performance Tuning of a CFD Code on the Earth Simulator By Ken ichi ITAKURA,* Atsuya UNO,* Mitsuo YOKOKAWA, Minoru SAITO, Takashi ISHIHARA

More information

Available online at www.sciencedirect.com Available online at www.sciencedirect.com

Available online at www.sciencedirect.com Available online at www.sciencedirect.com Available online at www.sciencedirect.com Available online at www.sciencedirect.com Physics Physics Procedia Procedia 00 (2011) 24 (2012) 000 000 2293 2297 Physics Procedia www.elsevier.com/locate/procedia

More information

Dynamically Reconfigurable Network Nodes in Cloud Computing Systems

Dynamically Reconfigurable Network Nodes in Cloud Computing Systems Dynamically Reconfigurable Network Nodes in Cloud Computing Systems HAYASHI Takeo, UENO Hiroshi Abstract As communication uses broader bandwidths and communication needs become diversified, a wide variety

More information

SR-IOV In High Performance Computing

SR-IOV In High Performance Computing SR-IOV In High Performance Computing Hoot Thompson & Dan Duffy NASA Goddard Space Flight Center Greenbelt, MD 20771 hoot@ptpnow.com daniel.q.duffy@nasa.gov www.nccs.nasa.gov Focus on the research side

More information

MOSIX: High performance Linux farm

MOSIX: High performance Linux farm MOSIX: High performance Linux farm Paolo Mastroserio [mastroserio@na.infn.it] Francesco Maria Taurino [taurino@na.infn.it] Gennaro Tortone [tortone@na.infn.it] Napoli Index overview on Linux farm farm

More information

Analysis and Research of Cloud Computing System to Comparison of Several Cloud Computing Platforms

Analysis and Research of Cloud Computing System to Comparison of Several Cloud Computing Platforms Volume 1, Issue 1 ISSN: 2320-5288 International Journal of Engineering Technology & Management Research Journal homepage: www.ijetmr.org Analysis and Research of Cloud Computing System to Comparison of

More information

Cluster Scalability of ANSYS FLUENT 12 for a Large Aerodynamics Case on the Darwin Supercomputer

Cluster Scalability of ANSYS FLUENT 12 for a Large Aerodynamics Case on the Darwin Supercomputer Cluster Scalability of ANSYS FLUENT 12 for a Large Aerodynamics Case on the Darwin Supercomputer Stan Posey, MSc and Bill Loewe, PhD Panasas Inc., Fremont, CA, USA Paul Calleja, PhD University of Cambridge,

More information

benchmarking Amazon EC2 for high-performance scientific computing

benchmarking Amazon EC2 for high-performance scientific computing Edward Walker benchmarking Amazon EC2 for high-performance scientific computing Edward Walker is a Research Scientist with the Texas Advanced Computing Center at the University of Texas at Austin. He received

More information

Efficient Service Broker Policy For Large-Scale Cloud Environments

Efficient Service Broker Policy For Large-Scale Cloud Environments www.ijcsi.org 85 Efficient Service Broker Policy For Large-Scale Cloud Environments Mohammed Radi Computer Science Department, Faculty of Applied Science Alaqsa University, Gaza Palestine Abstract Algorithms,

More information

Energy efficiency in HPC :

Energy efficiency in HPC : Energy efficiency in HPC : A new trend? A software approach to save power but still increase the number or the size of scientific studies! 19 Novembre 2012 The EDF Group in brief A GLOBAL LEADER IN ELECTRICITY

More information

Accelerating CFD using OpenFOAM with GPUs

Accelerating CFD using OpenFOAM with GPUs Accelerating CFD using OpenFOAM with GPUs Authors: Saeed Iqbal and Kevin Tubbs The OpenFOAM CFD Toolbox is a free, open source CFD software package produced by OpenCFD Ltd. Its user base represents a wide

More information

Figure 1. The cloud scales: Amazon EC2 growth [2].

Figure 1. The cloud scales: Amazon EC2 growth [2]. - Chung-Cheng Li and Kuochen Wang Department of Computer Science National Chiao Tung University Hsinchu, Taiwan 300 shinji10343@hotmail.com, kwang@cs.nctu.edu.tw Abstract One of the most important issues

More information

Recommendations for Performance Benchmarking

Recommendations for Performance Benchmarking Recommendations for Performance Benchmarking Shikhar Puri Abstract Performance benchmarking of applications is increasingly becoming essential before deployment. This paper covers recommendations and best

More information

MSC - Scientific Computing Facility (SCF) Supercomputer Status

MSC - Scientific Computing Facility (SCF) Supercomputer Status MSC - Scientific Computing Facility (SCF) Supercomputer Status RFP/Contract Status The formal SCF RFP was released on March 22, 2002 and closed June 10, 2002. The bid was competitive (more than one bid

More information

High Performance Computing for Numerical Weather Prediction

High Performance Computing for Numerical Weather Prediction High Performance Computing for Numerical Weather Prediction Isabella Weger European Centre for Medium-Range Weather Forecasts Slide 1 ECMWF ECMWF A European organisation with headquarters in the UK Established

More information

Silviu Panica, Marian Neagul, Daniela Zaharie and Dana Petcu (Romania)

Silviu Panica, Marian Neagul, Daniela Zaharie and Dana Petcu (Romania) Silviu Panica, Marian Neagul, Daniela Zaharie and Dana Petcu (Romania) Outline Introduction EO challenges; EO and classical/cloud computing; EO Services The computing platform Cluster -> Grid -> Cloud

More information

Introduction to High Performance Cluster Computing. Cluster Training for UCL Part 1

Introduction to High Performance Cluster Computing. Cluster Training for UCL Part 1 Introduction to High Performance Cluster Computing Cluster Training for UCL Part 1 What is HPC HPC = High Performance Computing Includes Supercomputing HPCC = High Performance Cluster Computing Note: these

More information

Performance and Scalability Best Practices in ArcGIS

Performance and Scalability Best Practices in ArcGIS 2013 Europe, Middle East, and Africa User Conference October 23-25, 2013 Munich, Germany Performance and Scalability Best Practices in ArcGIS Andrew Sakowicz asakowicz@esri.com Introductions Target audience

More information

P. Vicat-Blanc Primet

P. Vicat-Blanc Primet NEGST project Tools & software for flow processing in Grid networks Pascale.primet@inria.fr 1 Summary : our NEGST Y2007 activity Context GRIDNET-FJ : INRIA associated team NEGST : JST- CNRS (Naregi-Grid5000)

More information

10- High Performance Compu5ng

10- High Performance Compu5ng 10- High Performance Compu5ng (Herramientas Computacionales Avanzadas para la Inves6gación Aplicada) Rafael Palacios, Fernando de Cuadra MRE Contents Implemen8ng computa8onal tools 1. High Performance

More information

Performance Analysis of a Numerical Weather Prediction Application in Microsoft Azure

Performance Analysis of a Numerical Weather Prediction Application in Microsoft Azure Performance Analysis of a Numerical Weather Prediction Application in Microsoft Azure Emmanuell D Carreño, Eduardo Roloff, Jimmy V. Sanchez, and Philippe O. A. Navaux WSPPD 2015 - XIII Workshop de Processamento

More information

Key words: cloud computing, cluster computing, virtualization, hypervisor, performance evaluation

Key words: cloud computing, cluster computing, virtualization, hypervisor, performance evaluation Hypervisors Performance Evaluation with Help of HPC Challenge Benchmarks Reza Bakhshayeshi; bakhshayeshi.reza@gmail.com Mohammad Kazem Akbari; akbarif@aut.ac.ir Morteza Sargolzaei Javan; msjavan@aut.ac.ir

More information

Mitglied der Helmholtz-Gemeinschaft. System monitoring with LLview and the Parallel Tools Platform

Mitglied der Helmholtz-Gemeinschaft. System monitoring with LLview and the Parallel Tools Platform Mitglied der Helmholtz-Gemeinschaft System monitoring with LLview and the Parallel Tools Platform November 25, 2014 Carsten Karbach Content 1 LLview 2 Parallel Tools Platform (PTP) 3 Latest features 4

More information

User Reports. Time on System. Session Count. Detailed Reports. Summary Reports. Individual Gantt Charts

User Reports. Time on System. Session Count. Detailed Reports. Summary Reports. Individual Gantt Charts DETAILED REPORT LIST Track which users, when and for how long they used an application on Remote Desktop Services (formerly Terminal Services) and Citrix XenApp (known as Citrix Presentation Server). These

More information

JuRoPA. Jülich Research on Petaflop Architecture. One Year on. Hugo R. Falter, COO Lee J Porter, Engineering

JuRoPA. Jülich Research on Petaflop Architecture. One Year on. Hugo R. Falter, COO Lee J Porter, Engineering JuRoPA Jülich Research on Petaflop Architecture One Year on Hugo R. Falter, COO Lee J Porter, Engineering HPC Advisoy Counsil, Workshop 2010, Lugano 1 Outline The work of ParTec on JuRoPA (HF) Overview

More information

Architecture of Hitachi SR-8000

Architecture of Hitachi SR-8000 Architecture of Hitachi SR-8000 University of Stuttgart High-Performance Computing-Center Stuttgart (HLRS) www.hlrs.de Slide 1 Most of the slides from Hitachi Slide 2 the problem modern computer are data

More information