Sun Grid Engine Grid Computing Assignment – Fall 2005 James Ruff Senior Department of Mathematics and Computer Science Western Carolina University.

Slides:



Advertisements
Similar presentations
Installing DOS and Windows98 under MS Virtual PC.
Advertisements

Lesson 4 0x Operating Systems.
GALVESTON COUNTY, TX P-CARD TRAINING GALVESTON COUNTY.
Grid Computing, Barry Wilkinson, 2004A3.1 Assignment 3 Simple Job Submission Using GRAM.
DCMS 5.0 Product Demonstration. START OF EXERCISE #1.
CCPR Workshop Lexis Cluster Introduction October 19, 2007 David Ash.
Southgreen HPC system Concepts Cluster : compute farm i.e. a collection of compute servers that can be shared and accessed through a single “portal”
ALEPH version 21 Task Manager. New Task Manager Interface Admin tab 2 The Task Manager interface has been removed from the ALEPH menu, and is now found.
Software Tools Using PBS. Software tools Portland compilers pgf77 pgf90 pghpf pgcc pgCC Portland debugger GNU compilers g77 gcc Intel ifort icc.
DCC/FCUP Grid Computing 1 Resource Management Systems.
GRID workload management system and CMS fall production Massimo Sgaravatto INFN Padova.
6/2/20071 Grid Computing Sun Grid Engine (SGE) Manoj Katwal.
By Rashid Khan Lesson 7-Preparing for Output: Printer Configuration and Management.
Assignment 3 Using GRAM to Submit a Job to the Grid James Ruff Senior Western Carolina University Department of Mathematics and Computer Science.
Asynchronous Solution Appendix Eleven. Training Manual Asynchronous Solution August 26, 2005 Inventory # A11-2 Chapter Overview In this chapter,
Installing and running COMSOL on a Windows HPCS2008(R2) cluster
1 Chapter Overview Introduction to Windows XP Professional Printing Setting Up Network Printers Connecting to Network Printers Configuring Network Printers.
1 Using Compressed Files and Folders Applications and operating systems read and write to compressed files. NTFS uncompresses the file before making it.
Introduction to UNIX/Linux Exercises Dan Stanzione.
Linux Operations and Administration
MagicInfo Pro Server Software All control, content, and scheduling is performed within the MagicInfo Pro Server software previously installed. Before.
Cobian Backup 7 Tutorial Welcome to Cobian Backup 7. This tutorial will show you how to install the program, configure it and create your first scheduled.
CIM6400 CTNW (04/05) 1 CIM6400 CTNW Lesson 6 – More on Windows 2000.
Copyright © 2007, Oracle. All rights reserved. Managing Concurrent Requests.
Bigben Pittsburgh Supercomputing Center J. Ray Scott
MagicInfo Pro Scheduler Now that a template has been created from content imported into the Library, the user is ready to begin scheduling content to.
03/27/2003CHEP20031 Remote Operation of a Monte Carlo Production Farm Using Globus Dirk Hufnagel, Teela Pulliam, Thomas Allmendinger, Klaus Honscheid (Ohio.
CCPR Workshop Introduction to the Cluster July 13, 2006.
PNPI HEPD seminar 4 th November Andrey Shevel Distributed computing in High Energy Physics with Grid Technologies (Grid tools at PHENIX)
Rochester Institute of Technology Job Submission Andrew Pangborn & Myles Maxfield 10/19/2015Service Oriented Cyberinfrastructure Lab,
How to use Oracle Grid OEM to deploy J2EE applications By: Charles Li Date: 5/2/2005 Updated: 10/12/2005.
1 Short Course on Grid Computing Jornadas Chilenas de Computación 2010 INFONOR-CHILE 2010 November 15th - 19th, 2010 Antofagasta, Chile Dr. Barry Wilkinson.
July 11-15, 2005Lecture3: Grid Job Management1 Grid Compute Resources and Job Management.
Part Five: Globus Job Management A: GRAM B: Globus Job Commands C: Laboratory: globusrun.
1 / 61 Using the Customer Support Web Site © 2006, Universal Tax Systems, Inc. All Rights Reserved. Customer Support Site Objectives –In this chapter you.
Information Security 493. Lab 10.1: Look for Security Events within Windows XP Logging is automatically enabled for many items on a workstation, but the.
1 High-Performance Grid Computing and Research Networking Presented by David Villegas Instructor: S. Masoud Sadjadi
How to for compiling and running MPI Programs. Prepared by Kiriti Venkat.
Software Tools Using PBS. Software tools Portland compilers pgf77 pgf90 pghpf pgcc pgCC Portland debugger GNU compilers g77 gcc Intel ifort icc.
Cluster Computing Applications for Bioinformatics Thurs., Sept. 20, 2007 process management shell scripting Sun Grid Engine running parallel programs.
Job Submission with Globus, Condor, and Condor-G Selim Kalayci Florida International University 07/21/2009 Note: Slides are compiled from various TeraGrid.
Linux Operations and Administration
Submitting Jobs to the Sun Grid Engine at Sheffield and Leeds (Node1)
TUTORIAL #3 INFORMATION SECURITY : LOOK FOR SECURITY EVENTS WITHIN WINDOWS XP Logging is automatically enabled for many items on a workstation,
SQL SERVER 2008 Installation Guide A Step by Step Guide Prepared by Hassan Tariq.
Grid Compute Resources and Job Management. 2 Grid middleware - “glues” all pieces together Offers services that couple users with remote resources through.
Remote & Collaborative Visualization. TACC Remote Visualization Systems Longhorn – Dell XD Visualization Cluster –256 nodes, each with 48 GB (or 144 GB)
Portable Batch System – Definition and 3 Primary Roles Definition: PBS is a distributed workload management system. It handles the management and monitoring.
INFSO-RI Enabling Grids for E-sciencE Using of GANGA interface for Athena applications A. Zalite / PNPI.
Manchester Computing Supercomputing, Visualization & eScience Seamless Access to Multiple Datasets Mike AS Jones ● Demo Run-through.
A GANGA tutorial Professor Roger W.L. Jones Lancaster University.
HTCondor’s Grid Universe Jaime Frey Center for High Throughput Computing Department of Computer Sciences University of Wisconsin-Madison.
Wouter Verkerke, NIKHEF 1 Using ‘stoomboot’ for NIKHEF-ATLAS batch computing What is ‘stoomboot’ – Hardware –16 machines, each 2x quad-core Pentium = 128.
Active-HDL Server Farm Course 11. All materials updated on: September 30, 2004 Outline 1.Introduction 2.Advantages 3.Requirements 4.Installation 5.Architecture.
CACI Proprietary Information | Date 1 PD² v4.2 Increment 2 SR13 and FPDS Engine v3.5 Database Upgrade Name: Semarria Rosemond Title: Systems Analyst, Lead.
IPEmotion License Management PM (V1.2).
CFI 2004 UW A quick overview with lots of time for Q&A and exploration.
Grid Computing: An Overview and Tutorial Kenny Daily BIT Presentation 22/09/2016.
An Brief Introduction Charlie Taylor Associate Director, Research Computing UF Research Computing.
Fundamentals of Windows Mouse n 4 Basic Operations: –Pointing –Clicking –Double Clicking –Dragging.
Interacting with the cluster ssh, sftp, & slurm batch scripts
GRID COMPUTING.
PARADOX Cluster job management
Development Environment
OpenPBS – Distributed Workload Management System
Welcome to Indiana University Clusters
Using Paraguin to Create Parallel Programs
Globus Job Management. Globus Job Management Globus Job Management A: GRAM B: Globus Job Commands C: Laboratory: globusrun.
Working in The IITJ HPC System
Presentation transcript:

Sun Grid Engine Grid Computing Assignment – Fall 2005 James Ruff Senior Department of Mathematics and Computer Science Western Carolina University

We will: Discover what Sun Grid Engine is and what it accomplishes. Find out the advantages and disadvantages of Sun Grid Engine. Submit an example job submission using the command line. Submit and monitor a job using the Graphical User Interface. Discover a real application of Sun Grid Engine. (Using Scheme)

What is Sun Grid Engine? Sun Grid Engine, basically speaking, is software designed for cluster computing. Sun Grid Engine has two main versions: a professional version that may be purchased and a free open source implementation.

What Does Sun Grid Engine Allow Me to do? Run jobs on a cluster (1+ nodes) that may include parallel jobs. Submit job(s) to a cluster, which will then select which machine the job(s) will run on. Basically, SGE allows a user to run jobs on remote machines.

Pros/Cons of Sun Grid Engine Relatively easy installation. Wide usage, thus large knowledge base. Allows for GUI interface. Requires NFS. Large network strain. (Worst Case) Requires each user’s account to be mounted across the cluster. Pros Cons

An Example Setup of Sun Grid Engine Master Host Execution Host / Administration Host Submission Host Shadow Master Host Execution Host / Administration Host / Submission Host

Demonstration: Finding Hosts Sun Grid Engine allows any user to see how many hosts are available to run jobs. This is done using the ‘ qhost ’ command. % qhost This should result with several different hosts, along with how many processors each node has, the amount of ram, the amount of swap space and the usage of those resources.

Execution Host Selection The particular node your job runs on depends on the scheduling policies for your cluster. In general, no single node job should execute on the same node twice in a row. (Given there are at least two execution nodes.)

Demonstration: qrsh (Sun Grid Engine Remote Shell Execution) The binary ‘ /bin/hostname ’ displays the hostname of the current machine. Example: % qrsh /bin/hostname –Try running this command several times. Notice that the execution changes to different nodes.

Batch Jobs in Sun Grid Engine Batch jobs are submitted in the form of a shell script. (ex: bash) The script should be viewable by the entire cluster. (Unless other arrangements have been made to transfer between nodes.)

Demonstration: A Sleepy Job Execution Script: #$ -S /bin/sh # print date and time date sleep 60 # print date and time again date sleep 60 # See which host we are running on hostname

Demonstration: A Sleepy Job (Con’t) Checking the status of our job: % qstat The ‘ qstat ’ command will return the queue that the job is executing on. (The execution host name.) The status will also be listed in the output.

Other Useful Commands for Sun Grid Engine Job Submission qdel – Deletes a specified job. qhold – Holds a specified job. qrls - Releases a specified job from previous hold state. qresub – Submit a copy of a currently executing job.

Demonstration: A Sleepy Job – Gathering Output The filenames for the output (using a default submission) are the following: –(scriptfilename).sh.o(jobID) – The standard output that was produced by your job. –(scriptfilename).sh.e(jobID) – The standard error output that was produced by your job. To view your files, use the following: % cat filename.sh.?(jobID)

Beyond the Command Line – Sun Grid Engine’s Graphical User Interface Sun Grid Engine provides a Graphical User Interface that can complete everything that we did during the previous examples.

Demonstration: Starting the Graphical User Interface Please enter the following command to start the graphical user interface: % qmon >/dev/null 2>&1 &

Exploring the Graphical User Interface The Graphical User Interface has a wide variety of options. We will only concern ourselves with the following icons: View the Status of a Job Exit GUI

Demonstration: Running A Job through a GUI in Sun Grid Engine To bring up the submission window, please click on the submission icon. (Shown to the left)

Demonstration: Checking the status of the job Please click ‘Done’ and select the ‘Job Status’ button. Click on the tab called ‘Running Jobs’.

Demonstration: Closing qmon Please click on the ‘Done’ button. To close qmon, click on the ‘Exit’ button.

Demonstration: Creating our own submission script An existing binary, cal, is used to print out the current month. We can create a submission script to submit the binary to Sun Grid Engine (Filename: CalSubmission.sh ): /usr/bin/cal $1

Demonstration (Con’t) We can submit this script with command line arguments. %qsub CalSubmission.sh 2006 %qsub CalSubmission.sh 2000 The output will appear shortly in files as previously shown.

Submission to Sun Grid Engine Through GRAM Gridwise technologies, with other groups, have created an adapter where users can use the Sun Grid Engine as a scheduler for GRAM job submissions. Thus, the master node of Sun Grid Engine can now become the master node for GRAM submissions to SGE.

Demonstration: Submitting a Job to SGE through GRAM First, we must have a proxy. (As described in the GRAM presentation.) % grid-proxy-init (Enter YOUR password here.) We already have a container running on port 8443 (Secure).

Demonstration (Con’t) The GRAM job submission for SGE is almost identical to Fork. % globusrun-ws –submit \ -streaming –Ft SGE \ -c /usr/bin/uptime The submission may take up to a couple of minutes.

SGE Job Submissions are not limited to pre-existing binaries We have created two scripts thus far: simple.sh and CalSubmission.sh These scripts may be submitted to GRAM: %globusrun-ws –submit \ -streaming –Ft SGE \ -c /home/yourusername/simple.sh

How This Works GRAM is responsible for authorization. After authorization, SGE is used to choose the node to execute on. The streaming capability requires the use of RFT (Reliable File Transfer).

Conclusion We have: –Explored the Pros/Cons of using Sun Grid Engine. –Used the command line and graphical user interface to submit a wide variety of jobs to the Sun Grid Engine. –Used GRAM to submit jobs to the cluster using Sun Grid Engine.

Questions? Please ask!