June 28, 2016 Cluster Management for Non-XSEDE Systems Barbara Hallock, Eric Coulter, Sudhakar Pamidighantam.

Slides:



Advertisements
Similar presentations
Condor use in Department of Computing, Imperial College Stephen M c Gough, David McBride London e-Science Centre.
Advertisements

3rd Campus Grid SIG Meeting. Agenda Welcome OMII Requirements document Grid Data Group HTC Workshop Research Computing SIG? AOB Next meeting (AG)
Teaching Courses in Scientific Computing 30 September 2010 Roger Bielefeld Director, Advanced Research Computing.
Information Technology Center Introduction to High Performance Computing at KFUPM.
Copyright 2007, Information Builders. Slide 1 Workload Distribution for the Enterprise Mark Nesson, Vashti Ragoonath June, 2008.
Technology Steering Group January 31, 2007 Academic Affairs Technology Steering Group February 13, 2008.
Two Broad Categories of Software
Software Frameworks for Acquisition and Control European PhD – 2009 Horácio Fernandes.
Building a Cluster Support Service Implementation of the SCS Program UC Computing Services Conference Gary Jung SCS Project Manager
DataGrid Kimmo Soikkeli Ilkka Sormunen. What is DataGrid? DataGrid is a project that aims to enable access to geographically distributed computing power.
Technology Steering Group January 31, 2007 Academic Affairs Technology Steering Group February 13, 2008.
SM3121 Software Technology Mark Green School of Creative Media.
High Performance Computing (HPC) at Center for Information Communication and Technology in UTM.
Cloud Computing How secure is it? Author: Marziyeh Arabnejad Revised/Edited: James Childress April 2014 Tandy School of Computer Science.
Communicating with Users about HTCondor and High Throughput Computing Lauren Michael, Research Computing Facilitator HTCondor Week 2015.
Issue: Unknown / Unrecognized Filesystems Initial Analysis Extract Metadata Identify Restricted Info Identify Duplicates Generate Reports.
Rocks cluster : a cluster oriented linux distribution or how to install a computer cluster in a day.
About the Presentations The presentations cover the objectives found in the opening of each chapter. All chapter objectives are listed in the beginning.
Operating Systems Networking for Home and Small Businesses – Chapter 2 – Introduction To Networking.
Cloud computing is the use of computing resources (hardware and software) that are delivered as a service over the Internet. Cloud is the metaphor for.
Randy Diddel A+ Certified Technician Apple Certified Associate-Mac Integration OS X ITIL Foundations v3 Mac Team Technical Support Analyst II UNM IT Workstation.
DMF Configuration for JCU HPC Dr. Wayne Mallett Systems Manager James Cook University.
Module 13: Maintaining Software by Using Windows Server Update Services.
PCGRID ‘08 Workshop, Miami, FL April 18, 2008 Preston Smith Implementing an Industrial-Strength Academic Cyberinfrastructure at Purdue University.
So, Jung-ki Distributed Computing System LAB School of Computer Science and Engineering Seoul National University Implementation of Package Management.
Weekly Report By: Devin Trejo Week of May 30, > June 5, 2015.
การติดตั้งและทดสอบการทำคลัสเต อร์เสมือนบน Xen, ROCKS, และไท ยกริด Roll Implementation of Virtualization Clusters based on Xen, ROCKS, and ThaiGrid Roll.
Introduction to Interactive Media Interactive Media Tools: Software.
We’re going to a New Version? Bill Marlow School of Business, IT & Management.
The Cluster Computing Project Robert L. Tureman Paul D. Camp Community College.
J. J. Rehr & R.C. Albers Rev. Mod. Phys. 72, 621 (2000) A “cluster to cloud” story: Naturally parallel Each CPU calculates a few points in the energy grid.
NUOL Campus Intranet (Team 5) Final presentation NUOL Intranet Project Final Presentation Team 5
Russ Hobby Program Manager Internet2 Cyberinfrastructure Architect UC Davis.
Tool Integration with Data and Computation Grid GWE - “Grid Wizard Enterprise”
Issues Autonomic operation (fault tolerance) Minimize interference to applications Hardware support for new operating systems Resource management (global.
Enabling Grids for E-sciencE EGEE and gLite are registered trademarks Usage of virtualization in gLite certification Andreas Unterkircher.
Bio-Linux 3.0 An integrated bioinformatics solution for the EG community ClustalX showing DNA polymerase alignment GeneSpring showing yeast transcriptome.
EVGM081 Multi-Site Virtual Cluster: A User-Oriented, Distributed Deployment and Management Mechanism for Grid Computing Environments Takahiro Hirofuchi,
LegendCorp What is System Center Virtual Machine Manager (SCVMM)? SCVMM at a glance Features and Benefits Components / Topology /
 2001 Prentice Hall, Inc. All rights reserved. Chapter 1 – Introduction to Computers, the Internet and the World Wide Web Outline 1.1Introduction 1.2What.
Virtualization Technology and Microsoft Virtual PC 2007 YOU ARE WELCOME By : Osama Tamimi.
Hands-On Virtual Computing
Tool Integration with Data and Computation Grid “Grid Wizard 2”
CERN IT Department CH-1211 Genève 23 Switzerland t Migration from ELFMs to Agile Infrastructure CERN, IT Department.
Computing Strategies. A computing strategy should identify – the hardware, – the software, – Internet services, and – the network connectivity needed.
SMOOTHWALL FIREWALL By Nitheish Kumarr. INTRODUCTION  Smooth wall Express is a Linux based firewall produced by the Smooth wall Open Source Project Team.
1 Programming and problem solving in C, Maxima, and Excel.
Configuring SQL Server for a successful SharePoint Server Deployment Haaron Gonzalez Solution Architect & Consultant Microsoft MVP SharePoint Server
EGEE-III INFSO-RI Enabling Grids for E-sciencE EGEE and gLite are registered trademarksEGEE-III INFSO-RI MPI on the grid:
Introduction to Data Analysis with R on HPC Texas Advanced Computing Center Feb
© 2007 UC Regents1 Rocks – Present and Future The State of Things Open Source Grids and Clusters Conference Philip Papadopoulos, Greg Bruno Mason Katz,
Scientific Linux Inventory Project (SLIP) Troy Dawson Connie Sieh.
CFI 2004 UW A quick overview with lots of time for Q&A and exploration.
Accessing the VI-SEEM infrastructure
By Abdul Wahhab Khalid UEL ID
Virtualization and Hosting
Software Defined Storage
Heterogeneous Computation Team HybriLIT
GWE Core Grid Wizard Enterprise (
Fedora in Education BalCCon2k17 17th September 2017 Nemanja Milošević
Integration of Singularity With Makeflow
XSEDE’s Campus Bridging Project
Windows Server 2016 Software Defined Storage
Chapter 6 System and Application Software
OPS235: Lab 2 Virtual Machines – Part I
Chapter 6 System and Application Software
Chapter 6 System and Application Software
Chapter 6 System and Application Software
OU BATTLECARD: Oracle Systems Learning Subscription
IT Next – Transformation Program
Presentation transcript:

June 28, 2016 Cluster Management for Non-XSEDE Systems Barbara Hallock, Eric Coulter, Sudhakar Pamidighantam

Agenda Welcome and Introductions What is a cluster? Campus Bridging User Stories (and solutions) SEAGrid Questions / Discussion 2

Welcome and Introductions 3

Quick Poll Please respond to the poll, indicating whether you are on the IT side or the academic side of things at your institution (or other) 4

Quick Poll 2 Please respond to the poll, indicating whether or not you have current or planned access to a cluster at your institution 5

Quick Poll 3 Please indicate your level of familiarity with cluster technology. 6

What is a cluster? Many computers able to act as one on a single calculation Instead of many cores on the same machine, made of many machines networked together Ideally optimized for inter-machine communication 7

What is a cluster? 8 Image source: started.html

What is a cluster? We just covered hardware, what about software? Cluster administration – Rocks, XCAT, Warewulf Resource Manager/Job Scheduler – SGE, TORQUE+MAUI, SLURM Applications! – mostly HPC oriented 9

What’s the difference between HPC and HTC? HPC means High Performance Computing – “Painfully parallel” – lots of communication between processes – Large-scale simulations, “Scientific Computing”, MPI HTC means High Throughput Computing – “Pleasingly parallel” – little to no communication between processes – CERN, “Big Data”, Hadoop 10

What is a scheduler/resource manager? The pieces that make clusters work Provides awareness about what and where all users are running Prevents resource contention (Get off my lawn!) Shares available resource to users according to administrator policies 11

User Stories The following slides contain “user stories” that are generalized versions of common scenarios we work with in Campus Bridging efforts We will highlight user needs and show how different Campus Bridging resources can meet them in each case 12

User Story I We are a small institution of about 2,000 students, faculty, and staff. We recently received a grant that will allow us to put a cluster in our data center for the campus for everyone to share, but we don’t have room in the budget for costly support contracts or software licensing. It’s also important to us that our students be able to use it as a learning tool.

User Story I – CB Solutions The XSEDE National Integration Toolkit – XNIT – is a YUM repository that contains Open- Source packages Thanks to efforts by staff at Cornell University and funding from NSF, these packages are available at no cost, and can be used on any RPM-compatible cluster 14

XNIT Assembled to simplify the process of converting a "bare-bones" Linux cluster into a high-performance, parallel computing system that can be used to support scientific discovery Updates are packaged as soon as is feasible, which means less work to update a cluster utilizing XNIT 15

XNIT The base set of packages were selected because they are in use on XSEDE Service Providers (SPs) and have been found to be useful in the scientific computing context This means any skill sets picked up while using an XNIT package will translate to larger regional resources later, if needed 16

XNIT XNIT staff have also implemented packages by request from the community, so if there’s something you’d like to see that isn’t listed, let us know! Lots of the pre-installed software is in environment modules. Future XNIT packages will be available as modules, too. 17

Selected Application Categories Scientific Packages Compilers / Languages /Interpreters Linear Algebra /Libraries / MPI Implementation Visualization Utilities 18

Quick install demo repo.iu.xsede.org/xsederepo/readme.xsedere po repo.iu.xsede.org/xsederepo/readme.xsedere po 19

User Story II We are a growing computational department at the intersection of business analytics, math, and economics. We are trying to build a new program aimed at improving the computational education of our students and providing resources to the local community. We need to provide resources that are both easy to maintain and powerful enough to make a real difference for our users. We also aim to educate our researchers and students about resources available outside of our campus. 20

User Story II – CB Solutions The XSEDE Compatible Basic Cluster – Consists of the Rocks Cluster Manager plus a “roll” of open-source software available on XSEDE resources Simplifies the processes of building and maintaining a local cluster for admins with too much to do already Like XNIT, enables users to gain familiarity with the XSEDE “ecosystem” via XSEDE and bio rolls 21

XCBC Requirements Networking: – Static IP & hostname – At least a “dumb” switch for internal network – Two NIC’s on frontend, one on each compute node Hardware: – 1GB RAM – 30GB hard drive space – on frontend and compute nodes 22

XCBC – Easy to build! 23

XCBC – Easy to build! 24

XCBC – Easy to configure and manage! Schedulers: – SGE and TORQUE+MAUI available as rolls – SGE must be installed FROM THE BEGINNING – SGE will detect compute nodes; they must be added to TORQUE via /var/spool/torque/server_priv/nodes Users: – useradd + “rocks sync users” 25

XCBC – easy to manage 26 Rocks commands: – rocks list [host/interface/host attr/boot…] – rocks add/remove [roll/host attr/interface…] – rocks sync [users/config/…] – rocks run host [hostlist] “$command” – rocks [command] help Ganglia – Web-based monitoring system

XCBC: Caveat! Rocks currently supports only Centos 6.6 – Centos 7 Support is not predicted in the near future! – Security updates should continue be available, but this may not be ideal for production systems. Still an excellent tool for educational purposes Good community support with an active mailing list 27

Support and Consulting XSEDE Campus Bridging staff are also funded to travel in order to provide support and consulting with regards to campus bridging products at no cost to the host institution 28

SEAGrid Now, we’ll introduce Sudhakar Pamidighantam to talk about SEAGrid 29

Contact us Campus Bridging: (8a-5p Eastern time, M-F, except holidays) SEAGrid: 30