Running Virtual Machines in a Slurm Batch System

Save this PDF as:
 WORD  PNG  TXT  JPG

Size: px
Start display at page:

Download "Running Virtual Machines in a Slurm Batch System"

Transcription

1 Center for Information Services and High Performance Computing (ZIH) Running Virtual Machines in a Slurm Batch System Slurm User Group Meeting September 2015 Ulf Markwardt

2 Acknowledgements These colleagues contribute to our VM / Slurm infrastructure: Matthias Jurenz - Slurm plugins/provisioning Danny Rotscher - Slurm scripts/programming Ralph Müller-Pfefferkorn - integrating the G-Lite infrastructre, Rene Jäkel - Open Nebula Bert Wesarg - programming of the VM scheduler, SUG 2015, Ulf Markwardt (2/26)

3 Motivation We have a new nice HPC system (#66 in Top 500): over 1 PFLOPS performance about 5 PB scratch file system heterogeneity CPU (Westmere, Sandybridge, Haswell), 9 different sitzes of memory (from 2 GB/core to 2 TB/node), different accelerators Slurm is running fine with version (thanks to SchedMD for a few bug fixes...) SUG 2015, Ulf Markwardt (3/26)

4 Motivation We have a new nice HPC system (#66 in Top 500): over 1 PFLOPS performance about 5 PB scratch file system heterogeneity CPU (Westmere, Sandybridge, Haswell), 9 different sitzes of memory (from 2 GB/core to 2 TB/node), different accelerators Slurm is running fine with version (thanks to SchedMD for a few bug fixes...) But some research communities need a special access to use the resources easily... SUG 2015, Ulf Markwardt (4/26)

5 Motivation Source: CERN Requirements of scientific experiments providing data for thousands of physicists (eg. CERN): Grid computing scalable data infrastructure, well-defined software stack to enable reproducible analyses, easy-to-use, standardized access to compute resources SUG 2015, Ulf Markwardt (5/26)

6 Motivation Infrastructure works (too) well... Science gateways for other communities provide easy access to specific software and workflows rising demands for compute resources SUG 2015, Ulf Markwardt (6/26)

7 Motivation Infrastructure works (too) well... Science gateways for other communities provide easy access to specific software and workflows rising demands for compute resources But HPC software is somewhat restricted when it comes to specific versions: specific hardware (IB cards, accelerators), software dependencies (Lustre), vendor support policies. SUG 2015, Ulf Markwardt (7/26)

8 Motivation Solution: use virtual machines in an HPC / throughput system. arbitrary software stack (...), complete system isolation, fast start-up and shut-down stable and usable virtualization environment (KVM) Requirements: hide virtualization from user multiple users can share a VM / image multiple images SUG 2015, Ulf Markwardt (8/26)

9 Design Taurus slurmctld Grid infrastructure Compute nodes Login node(s) Testcase: How can we run jobs from ATLAS Grid on our HPC system? Requirements: Scientific Linux 6.6 no persistent file systems about 40 GB local disk space needed SUG 2015, Ulf Markwardt (9/26)

10 Requirements Slurm needs a unique hostname for each VM. Start a VM with a well-defined MAC address, Dynamically enter MAC/IP/name into local DHCP and DNS servers. Once the VM powers up it then has exactly the hostname Slurm knows, like vm-sl6-003 Jobs in the queue need different images We use --gres=vm_<imagename>; node definition like: NodeName=vm-SL6-[ ] Gres=vm_SL6:24... We have plenty of (mostly unused) hostnames in the configuration, for each image. We have a maximum runtime of 7 days Garbage collector.. SUG 2015, Ulf Markwardt (10/26)

11 Process overview A VM scheduler keeps track on the Slurm queue and on the VMs. VM Scheduler Daemon Slurm Controller VM Compute Node get jobs states from partition vm job states: pending/running VM Life-Cycle Triggered by the VM Scheduler kvmadmin submits a job to get a node start VM on host using KVM, image according to to gres start specified image slurmd on VM is available + res Scheduling jobs for this gres schedule job done SUG 2015, Ulf Markwardt (11/26)

12 Process overview A VM scheduler keeps track on the Slurm queue and on the VMs. VM Scheduler Daemon Slurm Controller VM Compute Node get jobs states from partition vm job states: pending/running VM Life-Cycle Triggered by the VM Scheduler kvmadmin submits a job to get a node start VM on host using KVM, image according to to gres start specified image slurmd on VM is available + res Scheduling jobs for this gres schedule job Remark: We have tried to run the jobs for donethe VMs in the normal Slurm: This became unstable after a couple of minutes; nodes vanished and re-appered. agent/is node resp: node:vm-sl6-009 rpc:1017 : Can t find an address, check slurm.conf SUG 2015, Ulf Markwardt (11/26)

13 Design Taurus slurmctld Taurus-VM squeue VM scheduler sinfo slurmctld Grid infrastructure Login node(s) Compute nodes Solution: Run the virtual machines in their own batch system. VM scheduler looks at the VM queue and the hosts Slurm. SUG 2015, Ulf Markwardt (12/26)

14 Design Grid infrastructure slurmctld Taurus-VM VM scheduler Job-Script: * define VM * start VM slurmctld Compute nodes Taurus Login node(s) According to a set of rules, VM scheduler now schedules an exclusive job in the normal batch system to start a new VM with these attributes: name of the image MAC address hostname The VM scheduler adds the new attributes to the DHCP and DNS servers. SUG 2015, Ulf Markwardt (13/26)

15 Design Taurus slurmctld Taurus-VM VM scheduler slurmctld Grid infrastructure * Do things * startvm vm-sl6-007 Compute nodes Login node(s) Things to do when the job starts: create a raw file for the VM s /tmp create Slurm reservation for the VM in 7 days (maximum job run time) start the VM SUG 2015, Ulf Markwardt (14/26)

16 Design Taurus Grid infrastructure slurmctld Taurus-VM VM scheduler vm-sl6-007 available slurmctld Compute nodes Login node(s) At boot time: get/expand a tarball including start scripts, Slurm binaries get slurm.conf and others scontrol update state=idle... SUG 2015, Ulf Markwardt (15/26)

17 Design Taurus slurmctld Taurus-VM squeue VM scheduler sinfo slurmctld Grid infrastructure Normal batch system service Login node(s) Compute nodes VM scheduler supervises the VM queue and the VM infrastructure... SUG 2015, Ulf Markwardt (16/26)

18 Design Taurus Grid infrastructure slurmctld Taurus-VM VM scheduler shutdown slurmctld Login node(s) Compute nodes According to a set of rules, it now shuts down the VM: scontrol update state=down... delete the entries in DHCP and DNS completely delete the virtual machine on the host clean up the temp space on the host SUG 2015, Ulf Markwardt (17/26)

19 Provisioning Our Slurm version and our config files change faster than images :-) use the same Slurm binaries as the host provision the current configuration to /etc/slurm at boot time We already have a good provisioning framework... SUG 2015, Ulf Markwardt (18/26)

20 Provisioning Heterogeneous systems need robust provisioning: GPU / non-gpu energy accounting: IPMI-raw(Bull) / IPMI batch / interactive test nodes (hardware/software) SUG 2015, Ulf Markwardt (19/26)

21 Provisioning Heterogeneous systems need robust provisioning: GPU / non-gpu energy accounting: IPMI-raw(Bull) / IPMI batch / interactive test nodes (hardware/software) Different files: slurm.conf (AcctGatherEnergyType) gres.conf prolog, epilogs, plugstack.conf (e.g. CPU/GPU frequency) sanity checks SUG 2015, Ulf Markwardt (19/26)

22 Provisioning by filename At the end of slurm.conf we include slurm.local.conf coming in different flavours (covering AcctGatherEnergyType): slurm.local.conf.0.n_#admin_nodes slurm.local.conf.0.n_#compute_nodes slurm.local.conf.1.n_#gpu_nodes slurm.local.conf.remain Aliases are in nodeset syntax (GPU_NODES=taurusi[ ]) Each host uses the files with a match of hostname/alias. Additional integer makes overloading possible for special cases. SUG 2015, Ulf Markwardt (20/26)

23 Provisioning by filename At the end of slurm.conf we include slurm.local.conf coming in different flavours (covering AcctGatherEnergyType): slurm.local.conf.0.n_#admin_nodes slurm.local.conf.0.n_#compute_nodes slurm.local.conf.1.n_#gpu_nodes slurm.local.conf.remain Aliases are in nodeset syntax (GPU_NODES=taurusi[ ]) Each host uses the files with a match of hostname/alias. Additional integer makes overloading possible for special cases. We have our configuration on a backed-up file system. Two ways of provisioning: Nodes with shared file systems (compute, login) pull their config during /etc/init.d/slurm. We push the configs for all other (admins). SUG 2015, Ulf Markwardt (20/26)

24 Comparison between VM and Host with SPEC CPU 400.perlbench 401.bzip2 100% 483.xalancbmk 403.gcc 473.astar 0% 429.mcf -100% 471.omnetpp 445.gobmk 456.hmmer 458.sjeng 462.libquantum 464.h264ref Overhead SSE4.2 Overhead AVX2 CINT perlbench 401.bzip2 403.gcc 429.mcf 445.gobmk 456.hmmer 458.sjeng 462.libquantum 464.h264ref 471.omnetpp 473.astar 483.xalancbmk Programming Language Compression C Compiler Combinatorial Optimization Artificial Intelligence: Go Search Gene Sequence Artificial Intelligence: chess Physics / Quantum Compu Video Compression Discrete Event Simulation Path-finding Algorithms XML Processing SUG 2015, Ulf Markwardt (21/26)

25 Comparison between VM and Host with SPEC CPU 416.gamess 410.bwaves 100% 482.sphinx3 433.milc 481.wrf 434.zeusmp 0% 470.lbm 435.gromacs 465.tonto -100% CFP cactusADM 410.bwaves 459.GemsFDTD 416.gamess Fluid Dynamics Quantum Chemistry. 433.milc Physics / Quantum Chromodynamics 434.zeusmp Physics / CFD 435.gromacs Biochemistry / Molecular Dynamics 436.cactusADM Physics / General Relativity 437.leslie3d 454.calculix 437.leslie3d 444.namd 447.dealII Fluid Dynamics Biology / Molecular Dynamics Finite Element Analysis 450.soplex Linear Programming, Optimization 453.povray Image Ray-tracing 444.namd 447.dealII 450.soplex 453.povray Overhead SSE4.2 Overhead AVX2 454.calculix 459.GemsFDTD 465.tonto 470.lbm 481.wrf 482.sphinx3 Structural Mechanics Computational Electromagnetics Quantum Chemistry Fluid Dynamics Weather Speech recognition SUG 2015, Ulf Markwardt (22/26)

26 Next steps What about Linux containers / Docker? KVM some performance degradation moderate memory ovehead sw completely independent from host easy standalone configuration migration and long runs possible Container same performance as native very low memory overhead sw stack bases on kernel of the host modular configuration possible (like host) SUG 2015, Ulf Markwardt (23/26)

27 Next steps What about Linux containers / Docker? KVM some performance degradation moderate memory ovehead sw completely independent from host easy standalone configuration migration and long runs possible Container same performance as native very low memory overhead sw stack bases on kernel of the host modular configuration possible (like host) A 3.10 Linux kernel is the minimum requirement for Docker - we have :-( Implementation of Docker containers is planned with RHEL 7. SUG 2015, Ulf Markwardt (24/26)

28 Next steps Installation issues... This week, phase 1 of our installation (130 TFlops) will be integrated into the newer system. New IP ranges are created with plenty of vacancies for virtual machines. afterwards VM/gLite goes productive Learn about the State=CLOUD stuff... and possibly use it. SUG 2015, Ulf Markwardt (25/26)

29 SchedMD and Slurm Community Thank you for the great and fast support! SUG 2015, Ulf Markwardt (26/26)

Virtualisierung im HPC-Kontext - Werkstattbericht

Virtualisierung im HPC-Kontext - Werkstattbericht Center for Information Services and High Performance Computing (ZIH) Virtualisierung im HPC-Kontext - Werkstattbericht ZKI Tagung AK Supercomputing 19. Oktober 2015 Ulf Markwardt +49 351-463 33640 ulf.markwardt@tu-dresden.de

More information

Achieving QoS in Server Virtualization

Achieving QoS in Server Virtualization Achieving QoS in Server Virtualization Intel Platform Shared Resource Monitoring/Control in Xen Chao Peng (chao.p.peng@intel.com) 1 Increasing QoS demand in Server Virtualization Data center & Cloud infrastructure

More information

Computer Architecture

Computer Architecture Computer Architecture Slide Sets WS 2013/2014 Prof. Dr. Uwe Brinkschulte M.Sc. Benjamin Betting Part 6 Fundamentals in Performance Evaluation Computer Architecture Part 6 page 1 of 22 Prof. Dr. Uwe Brinkschulte,

More information

Comparison of Processor Performance of SPECint2006 Benchmarks of some Intel Xeon Processors

Comparison of Processor Performance of SPECint2006 Benchmarks of some Intel Xeon Processors Comparison of Processor Performance of SPECint2006 Benchmarks of some Intel Xeon Processors Abdul Kareem PARCHUR * and Ram Asaray SINGH Department of Physics and Electronics, Dr. H. S. Gour University,

More information

How Much Power Oversubscription is Safe and Allowed in Data Centers?

How Much Power Oversubscription is Safe and Allowed in Data Centers? How Much Power Oversubscription is Safe and Allowed in Data Centers? Xing Fu 1,2, Xiaorui Wang 1,2, Charles Lefurgy 3 1 EECS @ University of Tennessee, Knoxville 2 ECE @ The Ohio State University 3 IBM

More information

An OS-oriented performance monitoring tool for multicore systems

An OS-oriented performance monitoring tool for multicore systems An OS-oriented performance monitoring tool for multicore systems J.C. Sáez, J. Casas, A. Serrano, R. Rodríguez-Rodríguez, F. Castro, D. Chaver, M. Prieto-Matias Department of Computer Architecture Complutense

More information

Performance Characterization of SPEC CPU2006 Integer Benchmarks on x86-64 64 Architecture

Performance Characterization of SPEC CPU2006 Integer Benchmarks on x86-64 64 Architecture Performance Characterization of SPEC CPU2006 Integer Benchmarks on x86-64 64 Architecture Dong Ye David Kaeli Northeastern University Joydeep Ray Christophe Harle AMD Inc. IISWC 2006 1 Outline Motivation

More information

Compiler-Assisted Binary Parsing

Compiler-Assisted Binary Parsing Compiler-Assisted Binary Parsing Tugrul Ince tugrul@cs.umd.edu PD Week 2012 26 27 March 2012 Parsing Binary Files Binary analysis is common for o Performance modeling o Computer security o Maintenance

More information

CPU Performance Evaluation: Cycles Per Instruction (CPI) Most computers run synchronously utilizing a CPU clock running at a constant clock rate:

CPU Performance Evaluation: Cycles Per Instruction (CPI) Most computers run synchronously utilizing a CPU clock running at a constant clock rate: CPU Performance Evaluation: Cycles Per Instruction (CPI) Most computers run synchronously utilizing a CPU clock running at a constant clock rate: Clock cycle where: Clock rate = 1 / clock cycle f = 1 /C

More information

Full Speed Ahead: Detailed Architectural Simulation at Near-Native Speed

Full Speed Ahead: Detailed Architectural Simulation at Near-Native Speed 1 Full Speed Ahead: Detailed Architectural Simulation at Near-Native Speed Andreas Sandberg Erik Hagersten David Black-Schaffer Abstract Popular microarchitecture simulators are typically several orders

More information

Review and Fundamentals

Review and Fundamentals Review and Fundamentals Nima Honarmand Measuring and Reporting Performance Performance Metrics Latency (execution/response time): time to finish one task Throughput (bandwidth): number of tasks/unit time

More information

Reducing Dynamic Compilation Latency

Reducing Dynamic Compilation Latency LLVM 12 - European Conference, London Reducing Dynamic Compilation Latency Igor Böhm Processor Automated Synthesis by iterative Analysis The University of Edinburgh LLVM 12 - European Conference, London

More information

Schedulability Analysis for Memory Bandwidth Regulated Multicore Real-Time Systems

Schedulability Analysis for Memory Bandwidth Regulated Multicore Real-Time Systems Schedulability for Memory Bandwidth Regulated Multicore Real-Time Systems Gang Yao, Heechul Yun, Zheng Pei Wu, Rodolfo Pellizzoni, Marco Caccamo, Lui Sha University of Illinois at Urbana-Champaign, USA.

More information

Secure Cloud Computing: The Monitoring Perspective

Secure Cloud Computing: The Monitoring Perspective Secure Cloud Computing: The Monitoring Perspective Peng Liu Penn State University 1 Cloud Computing is Less about Computer Design More about Use of Computing (UoC) CPU, OS, VMM, PL, Parallel computing

More information

A-DRM: Architecture-aware Distributed Resource Management of Virtualized Clusters

A-DRM: Architecture-aware Distributed Resource Management of Virtualized Clusters A-DRM: Architecture-aware Distributed Resource Management of Virtualized Clusters Hui Wang, Canturk Isci, Lavanya Subramanian, Jongmoo Choi, Depei Qian, Onur Mutlu Beihang University, IBM Thomas J. Watson

More information

When Prefetching Works, When It Doesn t, and Why

When Prefetching Works, When It Doesn t, and Why When Prefetching Works, When It Doesn t, and Why JAEKYU LEE, HYESOON KIM, and RICHARD VUDUC, Georgia Institute of Technology In emerging and future high-end processor systems, tolerating increasing cache

More information

Leistungsanalyse von Rechnersystemen

Leistungsanalyse von Rechnersystemen Center for Information Services and High Performance Computing (ZIH) Leistungsanalyse von Rechnersystemen 29. Oktober 2008 Nöthnitzer Straße 46 Raum 1026 Tel. +49 351-463 - 35048 Holger Brunst (holger.brunst@tu-dresden.de)

More information

SURFsara HPC Cloud Workshop

SURFsara HPC Cloud Workshop SURFsara HPC Cloud Workshop doc.hpccloud.surfsara.nl UvA workshop 2016-01-25 UvA HPC Course Jan 2016 Anatoli Danezi, Markus van Dijk cloud-support@surfsara.nl Agenda Introduction and Overview (current

More information

Analysis of Memory Sensitive SPEC CPU2006 Integer Benchmarks for Big Data Benchmarking

Analysis of Memory Sensitive SPEC CPU2006 Integer Benchmarks for Big Data Benchmarking Analysis of Memory Sensitive SPEC CPU2006 Integer Benchmarks for Big Data Benchmarking Kathlene Hurt and Eugene John Department of Electrical and Computer Engineering University of Texas at San Antonio

More information

How Much Power Oversubscription is Safe and Allowed in Data Centers?

How Much Power Oversubscription is Safe and Allowed in Data Centers? How Much Power Oversubscription is Safe and Allowed in Data Centers? Xing Fu, Xiaorui Wang University of Tennessee, Knoxville, TN 37996 The Ohio State University, Columbus, OH 43210 {xfu1, xwang}@eecs.utk.edu

More information

Computing in High- Energy-Physics: How Virtualization meets the Grid

Computing in High- Energy-Physics: How Virtualization meets the Grid Computing in High- Energy-Physics: How Virtualization meets the Grid Yves Kemp Institut für Experimentelle Kernphysik Universität Karlsruhe Yves Kemp Barcelona, 10/23/2006 Outline: Problems encountered

More information

Dynamic Virtual Machine Scheduling in Clouds for Architectural Shared Resources

Dynamic Virtual Machine Scheduling in Clouds for Architectural Shared Resources Dynamic Virtual Machine Scheduling in Clouds for Architectural Shared Resources JeongseobAhn,Changdae Kim, JaeungHan,Young-ri Choi,and JaehyukHuh KAIST UNIST {jeongseob, cdkim, juhan, and jhuh}@calab.kaist.ac.kr

More information

secubt : Hacking the Hackers with User-Space Virtualization

secubt : Hacking the Hackers with User-Space Virtualization secubt : Hacking the Hackers with User-Space Virtualization Mathias Payer Department of Computer Science ETH Zurich Abstract In the age of coordinated malware distribution and zero-day exploits security

More information

STABILIZER: Statistically Sound Performance Evaluation

STABILIZER: Statistically Sound Performance Evaluation STABILIZER: Statistically Sound Performance Evaluation Charlie Curtsinger Emery D. Berger Department of Computer Science University of Massachusetts Amherst Amherst, MA 01003 {charlie,emery}@cs.umass.edu

More information

Integration of Virtualized Workernodes in Batch Queueing Systems The ViBatch Concept

Integration of Virtualized Workernodes in Batch Queueing Systems The ViBatch Concept Integration of Virtualized Workernodes in Batch Queueing Systems, Dr. Armin Scheurer, Oliver Oberst, Prof. Günter Quast INSTITUT FÜR EXPERIMENTELLE KERNPHYSIK FAKULTÄT FÜR PHYSIK KIT University of the

More information

U-LITE: a proposal for scientific computing at LNGS. S. Parlati, P. Spinnato, S. Stalio LNGS 13 Sep. 2011

U-LITE: a proposal for scientific computing at LNGS. S. Parlati, P. Spinnato, S. Stalio LNGS 13 Sep. 2011 U-LITE: a proposal for scientific computing at LNGS S. Parlati, P. Spinnato, S. Stalio LNGS 13 Sep. 2011 20 years of Scientific Computing at LNGS Early 90s: highly centralized structure based on VMS cluster

More information

HQEMU: A Multi-Threaded and Retargetable Dynamic Binary Translator on Multicores

HQEMU: A Multi-Threaded and Retargetable Dynamic Binary Translator on Multicores H: A Multi-Threaded and Retargetable Dynamic Binary Translator on Multicores Ding-Yong Hong National Tsing Hua University Institute of Information Science Academia Sinica, Taiwan dyhong@sslab.cs.nthu.edu.tw

More information

Characterizing the Unique and Diverse Behaviors in Existing and Emerging General-Purpose and Domain-Specific Benchmark Suites

Characterizing the Unique and Diverse Behaviors in Existing and Emerging General-Purpose and Domain-Specific Benchmark Suites Characterizing the Unique and Diverse Behaviors in Existing and Emerging General-Purpose and Domain-Specific Benchmark Suites Kenneth Hoste Lieven Eeckhout ELIS Department, Ghent University Sint-Pietersnieuwstraat

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

SURFsara HPC Cloud Workshop

SURFsara HPC Cloud Workshop SURFsara HPC Cloud Workshop www.cloud.sara.nl Tutorial 2014-06-11 UvA HPC and Big Data Course June 2014 Anatoli Danezi, Markus van Dijk cloud-support@surfsara.nl Agenda Introduction and Overview (current

More information

Practical Memory Checking with Dr. Memory

Practical Memory Checking with Dr. Memory Practical Memory Checking with Dr. Memory Derek Bruening Google bruening@google.com Qin Zhao Massachusetts Institute of Technology qin zhao@csail.mit.edu Abstract Memory corruption, reading uninitialized

More information

KerMon: Framework for in-kernel performance and energy monitoring

KerMon: Framework for in-kernel performance and energy monitoring 1 KerMon: Framework for in-kernel performance and energy monitoring Diogo Antão Abstract Accurate on-the-fly characterization of application behavior requires assessing a set of execution related parameters

More information

<Insert Picture Here> Introducing Oracle VM: Oracle s Virtualization Product Strategy

<Insert Picture Here> Introducing Oracle VM: Oracle s Virtualization Product Strategy Introducing Oracle VM: Oracle s Virtualization Product Strategy SAFE HARBOR STATEMENT The following is intended to outline our general product direction. It is intended for information

More information

Cluster Implementation and Management; Scheduling

Cluster Implementation and Management; Scheduling Cluster Implementation and Management; Scheduling CPS343 Parallel and High Performance Computing Spring 2013 CPS343 (Parallel and HPC) Cluster Implementation and Management; Scheduling Spring 2013 1 /

More information

SLURM Workload Manager

SLURM Workload Manager SLURM Workload Manager What is SLURM? SLURM (Simple Linux Utility for Resource Management) is the native scheduler software that runs on ASTI's HPC cluster. Free and open-source job scheduler for the Linux

More information

Next Generation Operating Systems

Next Generation Operating Systems Next Generation Operating Systems Zeljko Susnjar, Cisco CTG June 2015 The end of CPU scaling Future computing challenges Power efficiency Performance == parallelism Cisco Confidential 2 Paradox of the

More information

Intro to Virtualization

Intro to Virtualization Cloud@Ceid Seminars Intro to Virtualization Christos Alexakos Computer Engineer, MSc, PhD C. Sysadmin at Pattern Recognition Lab 1 st Seminar 19/3/2014 Contents What is virtualization How it works Hypervisor

More information

TECHNICAL GUIDELINES FOR APPLICANTS TO PRACE 7 th CALL (Tier-0)

TECHNICAL GUIDELINES FOR APPLICANTS TO PRACE 7 th CALL (Tier-0) TECHNICAL GUIDELINES FOR APPLICANTS TO PRACE 7 th CALL (Tier-0) Contributing sites and the corresponding computer systems for this call are: GCS@Jülich, Germany IBM Blue Gene/Q GENCI@CEA, France Bull Bullx

More information

An introduction to Fyrkat

An introduction to Fyrkat Cluster Computing May 25, 2011 How to get an account https://fyrkat.grid.aau.dk/useraccount How to get help https://fyrkat.grid.aau.dk/wiki What is a Cluster Anyway It is NOT something that does any of

More information

Solution for private cloud computing

Solution for private cloud computing The CC1 system Solution for private cloud computing 1 Outline What is CC1? Features Technical details Use cases By scientist By HEP experiment System requirements and installation How to get it? 2 What

More information

SUSE Cloud 2.0. Pete Chadwick. Douglas Jarvis. Senior Product Manager pchadwick@suse.com. Product Marketing Manager djarvis@suse.

SUSE Cloud 2.0. Pete Chadwick. Douglas Jarvis. Senior Product Manager pchadwick@suse.com. Product Marketing Manager djarvis@suse. SUSE Cloud 2.0 Pete Chadwick Douglas Jarvis Senior Product Manager pchadwick@suse.com Product Marketing Manager djarvis@suse.com SUSE Cloud SUSE Cloud is an open source software solution based on OpenStack

More information

Cloud Computing for Control Systems CERN Openlab Summer Student Program 9/9/2011 ARSALAAN AHMED SHAIKH

Cloud Computing for Control Systems CERN Openlab Summer Student Program 9/9/2011 ARSALAAN AHMED SHAIKH Cloud Computing for Control Systems CERN Openlab Summer Student Program 9/9/2011 ARSALAAN AHMED SHAIKH CONTENTS Introduction... 4 System Components... 4 OpenNebula Cloud Management Toolkit... 4 VMware

More information

Cloud Bursting with SLURM and Bright Cluster Manager. Martijn de Vries CTO

Cloud Bursting with SLURM and Bright Cluster Manager. Martijn de Vries CTO Cloud Bursting with SLURM and Bright Cluster Manager Martijn de Vries CTO Architecture CMDaemon 2 Management Interfaces Graphical User Interface (GUI) Offers administrator full cluster control Standalone

More information

Benchmarking the Amazon Elastic Compute Cloud (EC2)

Benchmarking the Amazon Elastic Compute Cloud (EC2) Benchmarking the Amazon Elastic Compute Cloud (EC2) A Major Qualifying Project Report submitted to the Faculty of the WORCESTER POLYTECHNIC INSTITUTE in partial fulfillment of the requirements for the

More information

Performance of HPC Applications on the Amazon Web Services Cloud

Performance of HPC Applications on the Amazon Web Services Cloud Cloudcom 2010 November 1, 2010 Indianapolis, IN Performance of HPC Applications on the Amazon Web Services Cloud Keith R. Jackson, Lavanya Ramakrishnan, Krishna Muriki, Shane Canon, Shreyas Cholia, Harvey

More information

Efficient Load Balancing using VM Migration by QEMU-KVM

Efficient Load Balancing using VM Migration by QEMU-KVM International Journal of Computer Science and Telecommunications [Volume 5, Issue 8, August 2014] 49 ISSN 2047-3338 Efficient Load Balancing using VM Migration by QEMU-KVM Sharang Telkikar 1, Shreyas Talele

More information

Cache Capacity and Memory Bandwidth Scaling Limits of Highly Threaded Processors

Cache Capacity and Memory Bandwidth Scaling Limits of Highly Threaded Processors Cache Capacity and Memory Bandwidth Scaling Limits of Highly Threaded Processors Jeff Stuecheli 12 Lizy Kurian John 1 1 Department of Electrical and Computer Engineering, University of Texas at Austin

More information

Integration of Virtualized Worker Nodes in Batch Systems

Integration of Virtualized Worker Nodes in Batch Systems Integration of Virtualized Worker Nodes Dr. A. Scheurer, Dr. V. Büge, O. Oberst, P. Krauß Linuxtag 2010, Berlin, Session: Cloud Computing, Talk ID: #16197 KIT University of the State of Baden-Wuerttemberg

More information

Week Overview. Installing Linux Linux on your Desktop Virtualization Basic Linux system administration

Week Overview. Installing Linux Linux on your Desktop Virtualization Basic Linux system administration ULI101 Week 06b Week Overview Installing Linux Linux on your Desktop Virtualization Basic Linux system administration Installing Linux Standalone installation Linux is the only OS on the computer Any existing

More information

Evaluation of the Intel Core i7 Turbo Boost feature

Evaluation of the Intel Core i7 Turbo Boost feature 1 Evaluation of the Intel Core i7 Turbo Boost feature James Charles, Preet Jassi, Ananth Narayan S, Abbas Sadat and Alexandra Fedorova Abstract The Intel Core i7 processor code named Nehalem has a novel

More information

Dynamic Extension of a Virtualized Cluster by using Cloud Resources CHEP 2012

Dynamic Extension of a Virtualized Cluster by using Cloud Resources CHEP 2012 Dynamic Extension of a Virtualized Cluster by using Cloud Resources CHEP 2012 Thomas Hauth,, Günter Quast IEKP KIT University of the State of Baden-Wuerttemberg and National Research Center of the Helmholtz

More information

Virtualization. Michael Tsai 2015/06/08

Virtualization. Michael Tsai 2015/06/08 Virtualization Michael Tsai 2015/06/08 What is virtualization? Let s first look at a video from VMware http://bcove.me/x9zhalcl Problems? Low utilization Different needs DNS DHCP Web mail 5% 5% 15% 8%

More information

Cloud Computing through Virtualization and HPC technologies

Cloud Computing through Virtualization and HPC technologies Cloud Computing through Virtualization and HPC technologies William Lu, Ph.D. 1 Agenda Cloud Computing & HPC A Case of HPC Implementation Application Performance in VM Summary 2 Cloud Computing & HPC HPC

More information

Cloud Server. Parallels. Key Features and Benefits. White Paper. www.parallels.com

Cloud Server. Parallels. Key Features and Benefits. White Paper. www.parallels.com Parallels Cloud Server White Paper Key Features and Benefits www.parallels.com Table of Contents Introduction... 3 Key Features... 3 Distributed Cloud Storage (Containers and Hypervisors)... 3 Rebootless

More information

Quick Start Guide for Parallels Virtuozzo

Quick Start Guide for Parallels Virtuozzo PROPALMS VDI Version 2.1 Quick Start Guide for Parallels Virtuozzo Rev. 1.1 Published: JULY-2011 1999-2011 Propalms Ltd. All rights reserved. The information contained in this document represents the current

More information

PowerVC 1.2 Q4 2013 Power Systems Virtualization Center

PowerVC 1.2 Q4 2013 Power Systems Virtualization Center PowerVC 1.2 Q4 2013 Power Systems Virtualization Center At last a simple tool to spin-off Power Virtual Machines with very little effort Nigel Griffiths IBM Power Systems Corporation Advanced Technology

More information

DynaMOS: Dynamic Schedule Migration for Heterogeneous Cores

DynaMOS: Dynamic Schedule Migration for Heterogeneous Cores DynaMOS: Dynamic Schedule Migration for Heterogeneous Cores Shruti Padmanabha, Andrew Lukefahr, Reetuparna Das, and Scott Mahlke Advanced Computer Architecture Laboratory University of Michigan, Ann Arbor,

More information

Example of Standard API

Example of Standard API 16 Example of Standard API System Call Implementation Typically, a number associated with each system call System call interface maintains a table indexed according to these numbers The system call interface

More information

Building Clusters for Gromacs and other HPC applications

Building Clusters for Gromacs and other HPC applications Building Clusters for Gromacs and other HPC applications Erik Lindahl lindahl@cbr.su.se CBR Outline: Clusters Clusters vs. small networks of machines Why do YOU need a cluster? Computer hardware Network

More information

OGF25/EGEE User Forum Catania, Italy 2 March 2009

OGF25/EGEE User Forum Catania, Italy 2 March 2009 OGF25/EGEE User Forum Catania, Italy 2 March 2009 Constantino Vázquez Blanco Javier Fontán Muiños Raúl Sampedro Distributed Systems Architecture Research Group Universidad Complutense de Madrid 1/31 Outline

More information

DELL VS. SUN SERVERS: R910 PERFORMANCE COMPARISON SPECint_rate_base2006

DELL VS. SUN SERVERS: R910 PERFORMANCE COMPARISON SPECint_rate_base2006 DELL VS. SUN SERVERS: R910 PERFORMANCE COMPARISON OUR FINDINGS The latest, most powerful Dell PowerEdge servers deliver better performance than Sun SPARC Enterprise servers. In Principled Technologies

More information

Microsoft Intelligent Cloud

Microsoft Intelligent Cloud Microsoft Intelligent Cloud Ulrich (Uli) Homann Distinguished Architect Microsoft Cloud and Enterprise Engineering ulrichh@microsoft.com Why Cloud for transformation? #MSCloudRoadshow Azure IoT Suite Azure

More information

HPC performance applications on Virtual Clusters

HPC performance applications on Virtual Clusters Panagiotis Kritikakos EPCC, School of Physics & Astronomy, University of Edinburgh, Scotland - UK pkritika@epcc.ed.ac.uk 4 th IC-SCCE, Athens 7 th July 2010 This work investigates the performance of (Java)

More information

Fine-Grained User-Space Security Through Virtualization. Mathias Payer and Thomas R. Gross ETH Zurich

Fine-Grained User-Space Security Through Virtualization. Mathias Payer and Thomas R. Gross ETH Zurich Fine-Grained User-Space Security Through Virtualization Mathias Payer and Thomas R. Gross ETH Zurich Motivation Applications often vulnerable to security exploits Solution: restrict application access

More information

Building a More Efficient Data Center from Servers to Software. Aman Kansal

Building a More Efficient Data Center from Servers to Software. Aman Kansal Building a More Efficient Data Center from Servers to Software Aman Kansal Data centers growing in number, Microsoft has more than 10 and less than 100 DCs worldwide Quincy Chicago Dublin Amsterdam Japan

More information

Bright Cluster Manager

Bright Cluster Manager Bright Cluster Manager For HPC, Hadoop and OpenStack Craig Hunneyman Director of Business Development Bright Computing Craig.Hunneyman@BrightComputing.com Agenda Who is Bright Computing? What is Bright

More information

Introduction to parallel computing and UPPMAX

Introduction to parallel computing and UPPMAX Introduction to parallel computing and UPPMAX Intro part of course in Parallel Image Analysis Elias Rudberg elias.rudberg@it.uu.se March 22, 2011 Parallel computing Parallel computing is becoming increasingly

More information

Memory Bandwidth Management for Efficient Performance Isolation in Multi-core Platforms

Memory Bandwidth Management for Efficient Performance Isolation in Multi-core Platforms .9/TC.5.5889, IEEE Transactions on Computers Memory Bandwidth Management for Efficient Performance Isolation in Multi-core Platforms Heechul Yun, Gang Yao, Rodolfo Pellizzoni, Marco Caccamo, Lui Sha University

More information

Cloud Simulator for Scalability Testing

Cloud Simulator for Scalability Testing Cloud Simulator for Scalability Testing Nitin Singhvi (nitin.singhvi@calsoftinc.com) 1 Introduction Nitin Singhvi 11+ Years of experience in technology, especially in Networking QA. Currently playing roles

More information

Scientific and Technical Applications as a Service in the Cloud

Scientific and Technical Applications as a Service in the Cloud Scientific and Technical Applications as a Service in the Cloud University of Bern, 28.11.2011 adapted version Wibke Sudholt CloudBroker GmbH Technoparkstrasse 1, CH-8005 Zurich, Switzerland Phone: +41

More information

Overview of HPC systems and software available within

Overview of HPC systems and software available within Overview of HPC systems and software available within Overview Available HPC Systems Ba Cy-Tera Available Visualization Facilities Software Environments HPC System at Bibliotheca Alexandrina SUN cluster

More information

Virtualization of a Cluster Batch System

Virtualization of a Cluster Batch System Virtualization of a Cluster Batch System Christian Baun, Volker Büge, Benjamin Klein, Jens Mielke, Oliver Oberst and Armin Scheurer Die Kooperation von Cluster Batch System Batch system accepts computational

More information

II. Installing Debian Linux:

II. Installing Debian Linux: Debian Linux Installation Lab Spring 2013 In this lab you will be installing Debian Linux in a KVM (Kernel Virtual Machine). You will be guided through a series of steps to setup the network (IP addresses,

More information

Providing Self-Service, Life-cycle Management for Databases with VMware vfabric Data Director

Providing Self-Service, Life-cycle Management for Databases with VMware vfabric Data Director Providing Self-Service, Life-cycle Management for Databases with VMware vfabric Data Director Graeme Gordon Senior Systems Engineer, VMware 2013 VMware Inc. All rights reserved Traditional IT Application

More information

HTCondor at the RAL Tier-1

HTCondor at the RAL Tier-1 HTCondor at the RAL Tier-1 Andrew Lahiff, Alastair Dewhurst, John Kelly, Ian Collier, James Adams STFC Rutherford Appleton Laboratory HTCondor Week 2014 Outline Overview of HTCondor at RAL Monitoring Multi-core

More information

ovirt and Gluster hyper-converged! HA solution for maximum resource utilization

ovirt and Gluster hyper-converged! HA solution for maximum resource utilization ovirt and Gluster hyper-converged! HA solution for maximum resource utilization 21 st of Aug 2015 Martin Sivák Senior Software Engineer Red Hat Czech KVM Forum Seattle, Aug 2015 1 Agenda (Storage) architecture

More information

Safe Software Updates via Multi-version Execution

Safe Software Updates via Multi-version Execution Safe Software Updates via Multi-version Execution Petr Hosek Cristian Cadar Department of Computing Imperial College London {p.hosek, c.cadar}@imperial.ac.uk Abstract Software systems are constantly evolving,

More information

Working with HPC and HTC Apps. Abhinav Thota Research Technologies Indiana University

Working with HPC and HTC Apps. Abhinav Thota Research Technologies Indiana University Working with HPC and HTC Apps Abhinav Thota Research Technologies Indiana University Outline What are HPC apps? Working with typical HPC apps Compilers - Optimizations and libraries Installation Modules

More information

Cloud Computing with Red Hat Solutions. Sivaram Shunmugam Red Hat Asia Pacific Pte Ltd. sivaram@redhat.com

Cloud Computing with Red Hat Solutions. Sivaram Shunmugam Red Hat Asia Pacific Pte Ltd. sivaram@redhat.com Cloud Computing with Red Hat Solutions Sivaram Shunmugam Red Hat Asia Pacific Pte Ltd sivaram@redhat.com Linux Automation Details Red Hat's Linux Automation strategy for next-generation IT infrastructure

More information

Selective Hardware/Software Memory Virtualization

Selective Hardware/Software Memory Virtualization Selective Hardware/Software Memory Virtualization Xiaolin Wang Dept. of Computer Science and Technology, Peking University, Beijing, China, 100871 wxl@pku.edu.cn Jiarui Zang Dept. of Computer Science and

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

How to Setup Bare Metal Adaxa Suite Demonstration Box

How to Setup Bare Metal Adaxa Suite Demonstration Box How to Setup Bare Metal Adaxa Suite Demonstration Box Level 1 616 St Kilda Road Melbourne Victoria 3004 T:1-300-990-120 Email: info@adaxa.com Web: www.adaxa.com Table of Contents Pre-Requisites 1.1 Hardware

More information

Installing NICE Windows under Parallels for Mac

Installing NICE Windows under Parallels for Mac Installing NICE Windows under Parallels for Mac 1. Prerequisites a) A Mac with OS X connected to a portable socket on the CERN general public network b) Parallels Desktop software installed from: smb://cerndata44.cern.ch/d

More information

JBoss Seam Performance and Scalability on Dell PowerEdge 1855 Blade Servers

JBoss Seam Performance and Scalability on Dell PowerEdge 1855 Blade Servers JBoss Seam Performance and Scalability on Dell PowerEdge 1855 Blade Servers Dave Jaffe, PhD, Dell Inc. Michael Yuan, PhD, JBoss / RedHat June 14th, 2006 JBoss Inc. 2006 About us Dave Jaffe Works for Dell

More information

Self service for software development tools

Self service for software development tools Self service for software development tools Michal Husejko, behalf of colleagues in CERN IT/PES CERN IT Department CH-1211 Genève 23 Switzerland www.cern.ch/it Self service for software development tools

More information

Directions for VMware Ready Testing for Application Software

Directions for VMware Ready Testing for Application Software Directions for VMware Ready Testing for Application Software Introduction To be awarded the VMware ready logo for your product requires a modest amount of engineering work, assuming that the pre-requisites

More information

The Benefits of Virtualizing

The Benefits of Virtualizing T E C H N I C A L B R I E F The Benefits of Virtualizing Aciduisismodo Microsoft SQL Dolore Server Eolore in Dionseq Hitachi Storage Uatummy Environments Odolorem Vel Leveraging Microsoft Hyper-V By Heidi

More information

Linear-time Modeling of Program Working Set in Shared Cache

Linear-time Modeling of Program Working Set in Shared Cache Linear-time Modeling of Program Working Set in Shared Cache Xiaoya Xiang, Bin Bao, Chen Ding, Yaoqing Gao Computer Science Department, University of Rochester IBM Toronto Software Lab {xiang,bao,cding}@cs.rochester.edu,

More information

Comparison of Windows IaaS Environments

Comparison of Windows IaaS Environments Comparison of Windows IaaS Environments Comparison of Amazon Web Services, Expedient, Microsoft, and Rackspace Public Clouds January 5, 215 TABLE OF CONTENTS Executive Summary 2 vcpu Performance Summary

More information

Deploying Business Virtual Appliances on Open Source Cloud Computing

Deploying Business Virtual Appliances on Open Source Cloud Computing International Journal of Computer Science and Telecommunications [Volume 3, Issue 4, April 2012] 26 ISSN 2047-3338 Deploying Business Virtual Appliances on Open Source Cloud Computing Tran Van Lang 1 and

More information

An introduction to compute resources in Biostatistics. Chris Scheller schelcj@umich.edu

An introduction to compute resources in Biostatistics. Chris Scheller schelcj@umich.edu An introduction to compute resources in Biostatistics Chris Scheller schelcj@umich.edu 1. Resources 1. Hardware 2. Account Allocation 3. Storage 4. Software 2. Usage 1. Environment Modules 2. Tools 3.

More information

Themis Athanassiadou HPC Project Manager. ClusterVision. ClusterVision. Engineer Innovate Integrate

Themis Athanassiadou HPC Project Manager. ClusterVision. ClusterVision. Engineer Innovate Integrate Themis Athanassiadou HPC Project Manager About 12 years Europe's Dedicated Specialist for High-Performance Computing End-to-end hardware/software/services solution provider HPC engineering and innovation

More information

E-mail: guido.negri@cern.ch, shank@bu.edu, dario.barberis@cern.ch, kors.bos@cern.ch, alexei.klimentov@cern.ch, massimo.lamanna@cern.

E-mail: guido.negri@cern.ch, shank@bu.edu, dario.barberis@cern.ch, kors.bos@cern.ch, alexei.klimentov@cern.ch, massimo.lamanna@cern. *a, J. Shank b, D. Barberis c, K. Bos d, A. Klimentov e and M. Lamanna a a CERN Switzerland b Boston University c Università & INFN Genova d NIKHEF Amsterdam e BNL Brookhaven National Laboratories E-mail:

More information

Jonathan Worthington Scarborough Linux User Group

Jonathan Worthington Scarborough Linux User Group Jonathan Worthington Scarborough Linux User Group Introduction What does a Virtual Machine do? Hides away the details of the hardware platform and operating system. Defines a common set of instructions.

More information

Solving I/O Bottlenecks to Enable Superior Cloud Efficiency

Solving I/O Bottlenecks to Enable Superior Cloud Efficiency WHITE PAPER Solving I/O Bottlenecks to Enable Superior Cloud Efficiency Overview...1 Mellanox I/O Virtualization Features and Benefits...2 Summary...6 Overview We already have 8 or even 16 cores on one

More information

VALAR: A BENCHMARK SUITE TO STUDY THE DYNAMIC BEHAVIOR OF HETEROGENEOUS SYSTEMS

VALAR: A BENCHMARK SUITE TO STUDY THE DYNAMIC BEHAVIOR OF HETEROGENEOUS SYSTEMS VALAR: A BENCHMARK SUITE TO STUDY THE DYNAMIC BEHAVIOR OF HETEROGENEOUS SYSTEMS Perhaad Mistry, Yash Ukidave, Dana Schaa, David Kaeli Department of Electrical and Computer Engineering Northeastern University,

More information

EXPRESSCLUSTER X for Windows Quick Start Guide for Microsoft SQL Server 2014. Version 1

EXPRESSCLUSTER X for Windows Quick Start Guide for Microsoft SQL Server 2014. Version 1 EXPRESSCLUSTER X for Windows Quick Start Guide for Microsoft SQL Server 2014 Version 1 NEC EXPRESSCLUSTER X 3.x for Windows SQL Server 2014 Quick Start Guide Document Number ECX-MSSQL2014-QSG, Version

More information

Aspen Cloud Server Management Console

Aspen Cloud Server Management Console Aspen Cloud Server Management Console Management of Cloud Server Resources Power All Networks Ltd. User Guide June 2011, version 1.1.1 Refer to ICP V1.1 PAGE 1 Table of Content 1. Introduction... 4 2.

More information

High-Performance Nested Virtualization With Hitachi Logical Partitioning Feature

High-Performance Nested Virtualization With Hitachi Logical Partitioning Feature High-Performance Nested Virtualization With Hitachi Logical Partitioning Feature olutions Enabled by New Intel Virtualization Technology Extension in the Intel Xeon Processor E5 v3 Family By Hitachi Data

More information

Required Virtual Interface Maps to... mgmt0. bridge network interface = mgmt0 wan0. bridge network interface = wan0 mgmt1

Required Virtual Interface Maps to... mgmt0. bridge network interface = mgmt0 wan0. bridge network interface = wan0 mgmt1 VXOA VIRTUAL APPLIANCE KVM Hypervisor In-Line Deployment (Bridge Mode) 2012 Silver Peak Systems, Inc. Support Limitations In Bridge mode, the virtual appliance only uses mgmt0, wan0, and lan0. This Quick

More information