UPC/SHMEM PAT High-level Design v.1.1 Hung-Hsun Su UPC Group, HCS lab 6/21/2005.

Slides:



Advertisements
Similar presentations
K T A U Kernel Tuning and Analysis Utilities Department of Computer and Information Science Performance Research Laboratory University of Oregon.
Advertisements

ICS103 Programming in C Lecture 1: Overview of Computers & Programming
Systems Software.
Jill Goryca, Richard Hill American Control Conference June 17, 2013.
ITE PC v4.0 Chapter 1 1 Operating Systems Computer Networks– 2.
CS511 - Spring 2006 Final Presentation Project 3 - Team 1 Ching Chang Panagiotis Papapetrou Raymond Sweha.
DISTRIBUTED CONSISTENCY MANAGEMENT IN A SINGLE ADDRESS SPACE DISTRIBUTED OPERATING SYSTEM Sombrero.
Mahapatra-Texas A&M-Fall'001 cosynthesis Introduction to cosynthesis Rabi Mahapatra CPSC498.
The C++ Tracing Tutor: Visualizing Computer Program Behavior for Beginning Programming Courses Rika Yoshii Alastair Milne Computer Science Department California.
CAD/CAM Design Process and the role of CAD. Design Process Engineering and manufacturing together form largest single economic activity of western civilization.
The Data Warehouse and Design. Summary The design of the data warehouse begins with the data model The primary concern of the data warehouse developer.
Chapter 13: Advanced GUI and Graphics
Chapter 6: An Introduction to System Software and Virtual Machines
Cambodia-India Entrepreneurship Development Centre - : :.... :-:-
Tiered architectures 1 to N tiers. 2 An architectural history of computing 1 tier architecture – monolithic Information Systems – Presentation / frontend,
 A data processing system is a combination of machines and people that for a set of inputs produces a defined set of outputs. The inputs and outputs.
Scalability Module 6.
Microsoft ® Official Course Monitoring and Troubleshooting Custom SharePoint Solutions SharePoint Practice Microsoft SharePoint 2013.
Scott Pinkerton Sample GUI/Application Portfolio 1.
SOFTWARE.
Digital Graphics and Computers. Hardware and Software Working with graphic images requires suitable hardware and software to produce the best results.
DIVA. What Is Diva ? Diva is a software infrastructure for visualizing and interacting with dynamic information spaces. Visualizations are built by hooking.
1 Shawlands Academy Higher Computing Software Development Unit.
1 CSE 2102 CSE 2102 CSE 2102: Introduction to Software Engineering Ch9: Software Engineering Tools and Environments.
Chapter 3: Operating-System Structures System Components Operating System Services System Calls System Programs System Structure Virtual Machines System.
Advanced PI Calculation Engine Makes Complex PI Calculations Easy! Use of EDICTvb for Multi-Plant Advanced PI Calculations Dane OverfieldEXELE Information.
CCS APPS CODE COVERAGE. CCS APPS Code Coverage Definition: –The amount of code within a program that is exercised Uses: –Important for discovering code.
Behavior-based Spyware Detection By Engin Kirda and Christopher Kruegel Secure Systems Lab Technical University Vienna Greg Banks, Giovanni Vigna, and.
At A Glance VOLT is a freeware, platform independent tool set that coordinates cross-mission observation planning and scheduling among one or more space.
A Metadata Based Approach For Supporting Subsetting Queries Over Parallel HDF5 Datasets Vignesh Santhanagopalan Graduate Student Department Of CSE.
Magnetic Field Measurement System as Part of a Software Family Jerzy M. Nogiec Joe DiMarco Fermilab.
Scalable Analysis of Distributed Workflow Traces Daniel K. Gunter and Brian Tierney Distributed Systems Department Lawrence Berkeley National Laboratory.
11 July 2005 Tool Evaluation Scoring Criteria Professor Alan D. George, Principal Investigator Mr. Hung-Hsun Su, Sr. Research Assistant Mr. Adam Leko,
Performance Model & Tools Summary Hung-Hsun Su UPC Group, HCS lab 2/5/2004.
Contents 1.Introduction, architecture 2.Live demonstration 3.Extensibility.
Invitation to Computer Science 5 th Edition Chapter 6 An Introduction to System Software and Virtual Machine s.
INVITATION TO COMPUTER SCIENCE, JAVA VERSION, THIRD EDITION Chapter 6: An Introduction to System Software and Virtual Machines.
Silberschatz, Galvin and Gagne  Operating System Concepts Chapter 3: Operating-System Structures System Components Operating System Services.
We have developed a GUI-based user interface for Chandra data processing automation, data quality evaluation, and control of the system. This system, known.
Human-computer interfaces. Operating systems are software (i.e. programs) used to control the hardware directly used to run the applications software.
Pre-OTS Testing in Penticton Sonja Vrcic Socorro, December 11, 2007.
The Functions of Operating Systems Desktop PC Operating Systems.
Issues Autonomic operation (fault tolerance) Minimize interference to applications Hardware support for new operating systems Resource management (global.
CCSB223/SAD/CHAPTER131 Chapter 13 Designing the System Internals.
Foundation year Lec.3: Computer SoftwareLec.3: Computer Software Lecturer: Dalia Mirghani Year: 2014/2015.
Construction Planning and Prerequisite
Operating System Principles And Multitasking
The Software Development Process
Tool Visualizations, Metrics, and Profiled Entities Overview [Brief Version] Adam Leko HCS Research Laboratory University of Florida.
UNIX Unit 1- Architecture of Unix - By Pratima.
CPSC 171 Introduction to Computer Science System Software and Virtual Machines.
Page 1 Printing & Terminal Services Lecture 8 Hassan Shuja 11/16/2004.
Application Software System Software.
21 Sep UPC Performance Analysis Tool: Status and Plans Professor Alan D. George, Principal Investigator Mr. Hung-Hsun Su, Sr. Research Assistant.
HOW A COMPUTER PROCESSES DATA. What is hardware? Hardware is the electric, electronic and mechanical equipment that makes up a computer What is software?
Concept Diagram Hung-Hsun Su UPC Group, HCS lab 1/27/2004.
1 Presentation Methodology Summary B. Golden. 2 Introduction Why use visualizations?  To facilitate user comprehension  To convey complexity and intricacy.
Testing plan outline Adam Leko Hans Sherburne HCS Research Laboratory University of Florida.
Chapter – 8 Software Tools.
EPICS and LabVIEW Tony Vento, National Instruments
Lecture #1: Introduction to Algorithms and Problem Solving Dr. Hmood Al-Dossari King Saud University Department of Computer Science 6 February 2012.
Active-HDL Server Farm Course 11. All materials updated on: September 30, 2004 Outline 1.Introduction 2.Advantages 3.Requirements 4.Installation 5.Architecture.
Profiling/Tracing Method and Tool Evaluation Strategy Summary Slides Hung-Hsun Su UPC Group, HCS lab 1/25/2005.
Parallel Performance Wizard: A Generalized Performance Analysis Tool Hung-Hsun Su, Max Billingsley III, Seth Koehler, John Curreri, Alan D. George PPW.
In-situ Visualization using VisIt
Chapter 16: Distributed System Structures
Networking for Home and Small Businesses – Chapter 2
Networking for Home and Small Businesses – Chapter 2
Human–computer interfaces
Presentation transcript:

UPC/SHMEM PAT High-level Design v.1.1 Hung-Hsun Su UPC Group, HCS lab 6/21/2005

High-level Design

Instrumentation Units Automatic source instrumentation manager – deals with the system- assisted instrumentation of source code [i.e. all source level instrumentation except line-by-line instrumentation done by the user]. This determines what to instrument and how to instrument it from the manual instrumentation configuration [for user defined constructs] and automatic instrumentation configuration [for all upc and shmem constructs]. Binary Instrumentation manager – responsible for adding binary instrumentation to the object code. It also includes the capability of performing dynamic instrumentation at the binary level.

Measurement Units Measurement modules – perform the actual measurement of events either through software or hardware (most likely HW specific). They make calls to the low-level event write module to actually record the basic event data. Low-level event read/write modules – responsible for reading/writing low-level (raw, basic) event data to a storage space (file for tracing, memory for profiling). Event file manager – primarily used to merge all basic event data from all nodes into a single event file (needed to handle the case where a single event file will be too big). The sub-module trace format converter is responsible for converting from one trace format to another.

Analysis Units High-level event module – provides methods to access basic and complex event data. It calls the low-level event read module to access the basic event data. It includes an aggregation module to produce complex event data and a search/indexing module for efficient accessing of both the basic and complex event data. Simulator – provides simulation support using the event data.

Analysis Units Analysis modules – each module is responsible for performing and generating the necessary data for a specific analysis.  Standard analysis modules - basic load balancing analysis (ex: % busy), multiple runs data comparison  Advanced analysis modules - advanced load balancing analysis (ex: best scheduling analysis), scalability analysis, memory system analysis.  Other analysis modules - performance prediction, based-line execution time calculation, bottleneck detection, bottleneck management, bottleneck resolution (manual and automatic).

Analysis Units Analysis/data manager – interface to access all types of data (basic, complex, and analysis data). It uses high-level event module to read basic and complex event data. It also provides the mechanism for invoking the appropriate analysis modules based on the default view configuration and view configuration to produce appropriate analysis data (supplies the invoked analysis module with necessary data). The sub-module search module provides the capability for searching analysis data and basic/complex data (through calls to the high-level event module). Finally, it supplies the visualization manager with all or a subset of the various data generated. The other sub-module, performance data manager, is responsible for more advanced data management

Presentation Units Input interface – text-based (command line) and graphic-based (GUI) interface for performing tasks that can be done before [instrumenting the code, selecting views] and after [selecting additional views, data processing operations] running the program. Design of the GUI depends solely on the Command line interface.

Presentation Units Visualization manager – responsible for producing the various views (standard views, advanced views, plug-in views). Each module interacts only with the analysis/data manger to obtain all or a subset of the data (can cause analysis/data manager to invoke analysis modules). Has the capability of relating views together that affects the amount of data received from the analysis/data manager  Standard view modules - timeline display, call graph tree, speedup chart, communication volume display, and array placement display.  Advanced view modules – TBD.  Plug-in view modules - plug-in views defined by the user [an interface will be defined on how to add new plug-in view modules].

Flow Diagram