TPB Models Development Status Report Presentation to the Travel Forecasting Subcommittee Ron Milone National Capital Region Transportation Planning Board.

Slides:



Advertisements
Similar presentations
An efficient solution to toll choices involving multiple toll booths and / or multiple tolling strategies Current approaches to toll road demand forecasting:
Advertisements

Network II.5 simulator ..
Feedback Loops Guy Rousseau Atlanta Regional Commission.
IBM Software Group ® Integrated Server and Virtual Storage Management an IT Optimization Infrastructure Solution from IBM Small and Medium Business Software.
Agenda Overview Why TransCAD Challenges/tips Initiatives Applications.
Lesson 11-Virtual Private Networks. Overview Define Virtual Private Networks (VPNs). Deploy User VPNs. Deploy Site VPNs. Understand standard VPN techniques.
GCSE Computing - The CPU
1 I/O Management in Representative Operating Systems.
1 DOS with Windows 3.1 and 3.11 Operating Environments n Designed to allow applications to have a graphical interface DOS runs in the background as the.
High Performance Computing (HPC) at Center for Information Communication and Technology in UTM.
Module 2: Planning to Install SQL Server. Overview Hardware Installation Considerations SQL Server 2000 Editions Software Installation Considerations.
Challenge 2: Spatial Aggregation Level Multi-tier Modeling in Ohio Attempts to Balance Run Time and Forecast Granularity Gregory Giaimo, PE The Ohio Department.
Week 6 Operating Systems.
Tanenbaum 8.3 See references
Source: NHI course on Travel Demand Forecasting (152054A) Session 10 Traffic (Trip) Assignment Trip Generation Trip Distribution Transit Estimation & Mode.
Types of Operating System
Introduction to HP LoadRunner Getting Familiar with LoadRunner >>>>>>>>>>>>>>>>>>>>>>
Computer System Architectures Computer System Software
UNIX Unbounded 5 th Edition Amir Afzal Chapter 1 First Things First.
Calculating Transportation System User Benefits: Interface Challenges between EMME/2 and Summit Principle Author: Jennifer John Senior Transportation Planner.
Input/OUTPUT [I/O Module structure].
Windows 2000 Advanced Server and Clustering Prepared by: Tetsu Nagayama Russ Smith Dale Pena.
Module 9: Configuring Storage
Planning and Designing Server Virtualisation.
Increasing Precision in Highway Volume through Adjustment of Stopping Criteria in Traffic Assignment and Number of Feedbacks Behruz Paschai, Kathy Yu,
Development of the Graphical User Interface and Improvement and Streamlining of NYMTC's Best Practice Model Jim Lam, Andres Rabinowicz, Srini Sundaram,
TRANSIMS Version 5 Application Concepts January 20, 2011 David Roden – AECOM.
Utilizing Multi-threading, Parallel Processing, and Memory Management Techniques to Improve Transportation Model Performance Jim Lam Andres Rabinowicz.
Course ILT Basics of information technology Unit objectives Define “information technology” (IT), distinguish between hardware and software, and identify.
NTERFACING THE MORPC REGIONAL MODEL WITH DYNAMIC TRAFFIC SIMULATION INTERFACING THE MORPC REGIONAL MODEL WITH DYNAMIC TRAFFIC SIMULATION David Roden (AECOM)
Hardware Trends. Contents Memory Hard Disks Processors Network Accessories Future.
Processes and OS basics. RHS – SOC 2 OS Basics An Operating System (OS) is essentially an abstraction of a computer As a user or programmer, I do not.
1 Selecting LAN server (Week 3, Monday 9/8/2003) © Abdou Illia, Fall 2003.
Block1 Wrapping Your Nugget Around Distributed Processing.
Guy Rousseau, Modeling Manager, Atlanta Regional Commission Atlanta Travel Forecasting Methods: Traditional Trip-Based & Activity-Based Model AMPO Travel.
Lecture 3 Process Concepts. What is a Process? A process is the dynamic execution context of an executing program. Several processes may run concurrently,
Parallelization of the Classic Gram-Schmidt QR-Factorization
DONE-08 Sizing and Performance Tuning N-Tier Applications Mike Furgal Performance Manager Progress Software
Paul Metaxatos Urban Transportation Center University of Illinois at Chicago CTS-IGERT – Weekly Seminar May 14, 2009 Some Issues in Transportation Demand.
OCR GCSE Computing © Hodder Education 2013 Slide 1 OCR GCSE Computing Chapter 2: CPU.
Using GIS Networks to Represent Model Networks May 19, 2009 Hoyt Davis & Chunyu Lu Gannett Fleming, Inc.
Exploring Cube Base and Cube Voyager. Exploring Cube Base and Cube Voyager Use Cube Base and Cube Voyager to develop data, run scenarios, and examine.
YONSEI UNIVERSITY Korea Emme Users’ Conference 21 April 2010 Prof. Jin-Hyuk Chung.
PARALLEL APPLICATIONS EE 524/CS 561 Kishore Dhaveji 01/09/2000.
Presented to Time of Day Subcommittee May 9, 2011 Time of Day Modeling in FSUTMS.
1 FSUTMS-Voyager: Transit Standards within Evolving FSUTMS Summary Presentation Florida Model Task Force Tampa, Florida December 13 th, slides.
Presented to MTF Transit Committee presented by David Schmitt, AICP November 20, 2008 FSUTMS Transit Model Application.
11 CLUSTERING AND AVAILABILITY Chapter 11. Chapter 11: CLUSTERING AND AVAILABILITY2 OVERVIEW  Describe the clustering capabilities of Microsoft Windows.
TRB Planning Applications May 2009, Houston,TX Changing assignment algorithms: the price of better convergence Michael Florian and Shuguang He INRO.
May 4, 2007 The Corradino Group SE Florida Model Users Group 1 SERPM-6 and Cube Cluster Corradino’s Initial Experience.
Installation of Storage Foundation for Windows High Availability 5.1 SP2 1 Daniel Schnack Principle Technical Support Engineer.
Memory Hierarchy: Terminology Hit: data appears in some block in the upper level (example: Block X)  Hit Rate : the fraction of memory access found in.
1 Lecture 1: Computer System Structures We go over the aspects of computer architecture relevant to OS design  overview  input and output (I/O) organization.
Module 9 Planning and Implementing Monitoring and Maintenance.
PROOF tests at BNL Sergey Panitkin, Robert Petkus, Ofer Rind BNL May 28, 2008 Ann Arbor, MI.
PROOF Benchmark on Different Hardware Configurations 1 11/29/2007 Neng Xu, University of Wisconsin-Madison Mengmeng Chen, Annabelle Leung, Bruce Mellado,
FroNtier Stress Tests at Tier-0 Status report Luis Ramos LCG3D Workshop – September 13, 2006.
Lab Activities 1, 2. Some of the Lab Server Specifications CPU: 2 Quad(4) Core Intel Xeon 5400 processors CPU Speed: 2.5 GHz Cache : Each 2 cores share.
Wouter Verkerke, NIKHEF 1 Using ‘stoomboot’ for NIKHEF-ATLAS batch computing What is ‘stoomboot’ – Hardware –16 machines, each 2x quad-core Pentium = 128.
Transportation leadership you can trust. presented to Third International Conference on Innovations in Travel Modeling presented by Thomas Rossi Cambridge.
1© Copyright 2015 EMC Corporation. All rights reserved. NUMA(YEY) BY JACOB KUGLER.
Unit 2 Technology Systems
Travel Modelling Group Technical Advisory Committee
Understanding and Improving Server Performance
Windows Server 2008 and SQL 2008 Windows Server 2008.
DEVELOPMENT OF A FLEXIBLE ZONING SYSTEM
Reducing Model Run Times Using Distributed Computing
CCR Advanced Seminar: Running CPLEX Computations on the ISE Cluster
Support for ”interactive batch”
Ohio Traffic Forecasting Manual
Presentation transcript:

TPB Models Development Status Report Presentation to the Travel Forecasting Subcommittee Ron Milone National Capital Region Transportation Planning Board (TPB) January 23, 2009 tfs_ _modelsDevStatus_cubeCluster5.ppt

Models development status report, 1/23/09 2 Version 2.3 Model Development Activities in motion since last meeting: –2007/8 HH travel survey file cleaned –Updated (~3,700) TAZ system remains under review –Project for improving the TPB’s use of GIS technology to facilitate network development remains in progress –Approaches for reducing Version 2.3 model execution times have been explored

Models development status report, 1/23/09 3 Speeding up model executions Four approaches identified: 1.Faster hardware  2.New traffic assignment solution algorithms under development by Citilabs, Inc.  3.Decrease the number of speed feedback iterations 4.Implementation of distributed processing (DP) capability that currently exists in Cube Voyager

Models development status report, 1/23/09 4 Reducing Speed Feedback Iterations in Version 2.3 Investigation: Pre-existing model outputs summarized by iteration (year: 2030) –VMT by facility type –Total transit trips by trip purpose

Models development status report, 1/23/09 5 Version 2.3 VMT by iteration

Models development status report, 1/23/09 6 Ver.2.3 Transit Trips by Iteration

Models development status report, 1/23/09 7 Conclusions based on iteration summaries Global metrics summaries indicate that results of iteration 3-4 are not substantially different than those of iteration 6 Finer detail iteration summaries should be analyzed (e.g., screenline, link level, possibly i/j level)

Models development status report, 1/23/09 8 Implementing Distributed Processing Current hardware/software specifications used by TPB Preliminary DP work: Identifying the most time-consuming modeling steps Overview & key points regarding the deployment of DP with Cube Voyager (a.k.a. Cube Cluster) Experience gained by TPB thus far

Models development status report, 1/23/09 9 Travel model server specifications Hardware (cogtms002) –High-end workstation (not a true server) –Two CPUs, each with four cores = 8 cores Intel Xeon X5365; Chip speed: 3.0 GHz System bus: 1,333 MHz; L2 Cache: 8 MB per CPU –Memory: 4 GB RAM –Hard drive: 2.27 TB direct-attach storage array Software –Server operating system, so that the computer can operate like a server (Windows Server 2003) –Citilabs Cube Base 5.0.2; Citilabs Cube Voyager Note: We have moved from TP to Voyager for the Ver. 2.3 model Server (cogtms002) is shared by 4-6 modelers

Models development status report, 1/23/09 10 Cube Cluster: Preliminary work Identify the most time-consuming modeling steps (Ver. 2.3 model) Model execution time by iteration (cogtms002)

Models development status report, 1/23/09 11 Cube Cluster: Preliminary work Model execution time by modeling step (iteration 6) Minutes  Model Step

Models development status report, 1/23/09 12 Cube Cluster: Overview Spread the computing load across –Multiple computers connected via a LAN; –Multiple CPUs within one computer, or multiple cores within a CPU or a set of CPUs in a computer (the current approach being tested by TPB staff); or –Both Each processor or core is referred to as a “node” There is generally a main process and one or more sub- processes Cube Cluster works with Voyager, not with TP+ Two flavors of distributed processing in Voyager –Intra-step distributed processing (IDP) –Multi-step distributed processing (MDP)

Models development status report, 1/23/09 13 Cube Cluster: Overview Intra-step distributed processing (IDP) –IDP breaks up zone-based processing of vectors or matrices into zone groups that can be processed concurrently on multiple computing nodes –Works for only two modules: MATRIX, HIGHWAY Multi-step distributed processing (MDP) –More general than IDP –Can be used to break up processing conducted by any module in Voyager, as well as any user-written program (e.g., Fortran) –Caveat: the distributed blocks and the mainline process must be logically independent of each other. For example, you cannot run path skimming procedures before you update the speeds on the network that will be skimmed. However, you can assign peak and off-peak networks concurrently in most models, since these steps are generally independent of each other.

Models development status report, 1/23/09 14 Cube Cluster: Key points Because of the zone-independent requirement on IDP and the step- independent requirement on MDP –it requires careful planning and setup by the user to implement DP Cube Cluster has limited error-handling capabilities –It uses a file-based signaling method to communicate between the main process and the sub-process(es) –If a sub-process crashes, the main process will wait indefinitely –Best to use DP on a model that has been cleaned of syntax errors In general, DP works well for computationally intensive applications (e.g. doing hundreds of matrix computations for each zone in a mode choice step), but will result in less time savings for disk intensive procedures (e.g. combing 3 matrix files into one matrix file)

Models development status report, 1/23/09 15 Cube Cluster: TPB experience AECOM did some work in this area, while under contract with us –Work is undocumented, but we have reviewed some of their work –Lesson: Things can get complicated TPB staff –Tested IDP (for MATRIX and HIGHWAY) on the highway assignment step, since this is the most time-consuming step

Models development status report, 1/23/09 16 Cube Cluster: TPB experience Adding code for DP Global control of DP options (in your script): –DISTRIBUTE INTRASTEP=T MULTISTEP=F Initiate IDP of the current MATRIX or HIGHWAY step (in your script): –DistributeINTRASTEP ProcessID='mwcog', ProcessList=1-4 Open up one or more cluster nodes –Interactively: Cube Base > Utilities > Cluster Node Management –Command line Voyager –start Voyager.exe mwcog1.script /wait -Pvoya Cluster utility –start Cluster [ProcID] [ProcList] [Start/Close/List] [Exit] These commands would generally be in the batch file used to launch your script. Up to now, we have used the interactive approach and the Voyager command.

Models development status report, 1/23/09 17 Cube Cluster: TPB experience Highway assignment: Running IDP with 3-4 sub-processes results in a 50% time savings 83 minutes should now take about 42 min. Time savings –42 x 7 = 294 min (4.9 hours) –So 18.5 hours becomes 13.6 hours (25% savings)

Models development status report, 1/23/09 18 Conclusions Running time can be reduced by: –Reducing the number of speed feedback steps –Implementing DP –New hardware (?): We need to use benchmarks to assess speed, not simply the rated speed of the CPU in GHz Future steps –Optimizing the use of IDP and/or MDP in executing traffic assignment and other steps (mode choice and transit fare building) –Investigate improved traffic assignment algorithms These methods reach a higher level of convergence using a smaller number of iterations, so they have the potential to save us time