STAR C OMPUTING Plans for Production Use of Grand Challenge Software in STAR Torre Wenaus BNL Grand Challenge Meeting LBNL 10/23/98.

Slides:



Advertisements
Similar presentations
Copyright © 2007 Ramez Elmasri and Shamkant B. Navathe Slide
Advertisements

Chapter 9. Performance Management Enterprise wide endeavor Research and ascertain all performance problems – not just DBMS Five factors influence DB performance.
Batch Production and Monte Carlo + CDB work status Janusz Martyniak, Imperial College London MICE CM37 Analysis, Software and Reconstruction.
O. Stézowski IPN Lyon AGATA Week September 2003 Legnaro Data Analysis – Team #3 ROOT as a framework for AGATA.
EventStore Managing Event Versioning and Data Partitioning using Legacy Data Formats Chris Jones Valentin Kuznetsov Dan Riley Greg Sharp CLEO Collaboration.
Week 2 IBS 685. Static Page Architecture The user requests the page by typing a URL in a browser The Browser requests the page from the Web Server The.
Magda – Manager for grid-based data Wensheng Deng Physics Applications Software group Brookhaven National Laboratory.
Chapter 1 Introduction to Databases
Migrating to EPiServer CMS 5 Johan Björnfot -
VxOware Progress Report August How to create a new section? Configure section –Create metadata structure (template) –Create elements map for web.
Introduction to Databases and Database Languages
IT Essentials: PC Hardware and Software 1 Chapter 7 Windows NT/2000/XP Operating Systems.
Database testing Prepared by Saurabh sinha. Database testing mainly focus on: Data integrity test Data integrity test Stored procedures test Stored procedures.
Practical Database Design and Tuning. Outline  Practical Database Design and Tuning Physical Database Design in Relational Databases An Overview of Database.
STAR C OMPUTING ROOT in STAR Torre Wenaus STAR Computing and Software Leader Brookhaven National Laboratory, USA ROOT 2000 Workshop, CERN February 3, 2000.
Alexandre A. P. Suaide VI DOSAR workshop, São Paulo, 2005 STAR grid activities and São Paulo experience.
Irina Sourikova Brookhaven National Laboratory for the PHENIX collaboration Migrating PHENIX databases from object to relational model.
A User’s Introduction to the Grand Challenge Software STAR-GC Workshop Oct 1999 D. Zimmerman.
Grid Status - PPDG / Magda / pacman Torre Wenaus BNL U.S. ATLAS Physics and Computing Advisory Panel Review Argonne National Laboratory Oct 30, 2001.
Summary of 1 TB Milestone RD Schaffer Outline : u Goals and assumptions of the exercise u The hardware constraints u The basic model u What have we understood.
Chapter 16 Practical Database Design and Tuning Copyright © 2004 Pearson Education, Inc.
Grand Challenge MDC1 Plans Doug Olson Nuclear Science Division, Berkeley Lab for the HENP-GC Collaboration RCF Meeting September 24, 1998.
What is a schema ? Schema is a collection of Database Objects. Schema Objects are logical structures created by users to contain, or reference, their data.
PHENIX Simulation System 1 December 7, 1999 Simulation: Status and Milestones Tarun Ghosh, Indrani Ojha, Charles Vanderbilt University.
STAR C OMPUTING STAR Computing Infrastructure Torre Wenaus BNL STAR Collaboration Meeting BNL Jan 31, 1999.
NOVA Networked Object-based EnVironment for Analysis P. Nevski, A. Vaniachine, T. Wenaus NOVA is a project to develop distributed object oriented physics.
Magda Distributed Data Manager Status Torre Wenaus BNL ATLAS Data Challenge Workshop Feb 1, 2002 CERN.
Grand Challenge and PHENIX Report post-MDC2 studies of GC software –feasibility for day-1 expectations of data model –simple robustness tests –Comparisons.
Introduction to Database Systems1. 2 Basic Definitions Mini-world Some part of the real world about which data is stored in a database. Data Known facts.
1 GCA Application in STAR GCA Collaboration Grand Challenge Architecture and its Interface to STAR Sasha Vaniachine presenting for the Grand Challenge.
STAR C OMPUTING STAR MDC1 Experience and Revised Computing Requirements Torre Wenaus BNL RHIC Computing Advisory Committee Meeting BNL December 3-4, 1998.
Analysis trains – Status & experience from operation Mihaela Gheata.
STAR Event data storage and management in STAR V. Perevoztchikov Brookhaven National Laboratory,USA.
Linux+ Guide to Linux Certification, Third Edition
STAR C OMPUTING STAR Analysis Operations and Issues Torre Wenaus BNL STAR PWG Videoconference BNL August 13, 1999.
1D. Olson, SDM-ISIC Mtg, 26 Mar 2002 Scientific Data Management: An Incomplete Experimental HENP Perspective D. Olson, LBNL 26 March 2002 SDM-ISIC Meeting.
STAR Collaboration, July 2004 Grid Collector Wei-Ming Zhang Kent State University John Wu, Alex Sim, Junmin Gu and Arie Shoshani Lawrence Berkeley National.
Online Databases General Functions Database Categories Tools Conditions & Configurations DB Technical Issues Activities of Interest & Need R. Jeff Porter.
STAR C OMPUTING STAR Computing Status and Plans Torre Wenaus BNL STAR Collaboration Meeting BNL Jan 31, 1999.
Grand Challenge in MDC2 D. Olson, LBNL 31 Jan 1999 STAR Collaboration Meeting
1 fileCatalog, tagDB and GCA A. Vaniachine Grand Challenge STAR fileCatalog, tagDB and Grand Challenge Architecture A. Vaniachine presenting for the Grand.
STAR C OMPUTING The STAR Databases: From Objectivity to ROOT+MySQL Torre Wenaus BNL ATLAS Computing Week CERN August 31, 1999.
1 Grid in STAR: Needs and Plans A. Vaniachine STAR Grid Components in STAR: Experience, Needs and Plans A. Vaniachine Lawrence Berkeley National Laboratory.
STAR C OMPUTING Introduction Torre Wenaus BNL May ‘99 STAR Computing Meeting BNL May 24, 1999.
SPI NIGHTLIES Alex Hodgkins. SPI nightlies  Build and test various software projects each night  Provide a nightlies summary page that displays all.
Magda Distributed Data Manager Prototype Torre Wenaus BNL September 2001.
M. Oldenburg GridPP Metadata Workshop — July 4–7 2006, Oxford University 1 Markus Oldenburg GridPP Metadata Workshop July 4–7 2006, Oxford University ALICE.
Grid Status - PPDG / Magda / pacman Torre Wenaus BNL DOE/NSF Review of US LHC Software and Computing Fermilab Nov 29, 2001.
Presented By:. What is JavaHelp: Most software developers do not look forward to spending time documenting and explaining their product. JavaSoft has.
1 Efficient Data Access for Distributed Computing at RHIC A. Vaniachine Efficient Data Access for Distributed Computing at RHIC A. Vaniachine Lawrence.
Building Preservation Environments with Data Grid Technology Reagan W. Moore Presenter: Praveen Namburi.
SQL Advanced Monitoring Using DMV, Extended Events and Service Broker Javier Villegas – DBA | MCP | MCTS.
October 19, 2010 David Lawrence JLab Oct. 19, 20101RootSpy -- CHEP10, Taipei -- David Lawrence, JLab Parallel Session 18: Software Engineering, Data Stores,
1 Copyright © 2008, Oracle. All rights reserved. Repository Basics.
The HENP Grand Challenge Project and initial use in the RHIC Mock Data Challenge 1 D. Olson DM Workshop SLAC, Oct 1998.
You Inherited a Database Now What? What you should immediately check and start monitoring for. Tim Radney, Senior DBA for a top 40 US Bank President of.
Magda Distributed Data Manager Torre Wenaus BNL October 2001.
Compute and Storage For the Farm at Jlab
Data Virtualization Demoette… Logging in CIS
Practical Database Design and Tuning
Module 11: File Structure
ALICE analysis preservation
Oracle Solaris Zones Study Purpose Only
Introduction of Week 3 Assignment Discussion
Grid Canada Testbed using HEP applications
Practical Database Design and Tuning
Data Persistency Solution for LHCb
OO-Design in PHENIX PHENIX, a BIG Collaboration A Liberal Data Model
Offline framework for conditions data
Presentation transcript:

STAR C OMPUTING Plans for Production Use of Grand Challenge Software in STAR Torre Wenaus BNL Grand Challenge Meeting LBNL 10/23/98

STAR C OMPUTING Torre Wenaus, BNL Grand Challenge Meeting 10/98 Congrats and thank you! First, congrats on a successful test in MDC1 of a software apparatus  very well focused on the priority needs of STAR and RHIC  of immediate use to STAR for day to day operations  on target to meet STAR’s needs for physics analysis Thank you for building rapidly a very useful facility!

STAR C OMPUTING Torre Wenaus, BNL Grand Challenge Meeting 10/98 STAR MDC1 Production Simulated data:1.7TB of 200GeV/n Au-Au HIJING (217k events) Reconstructed DSTs: 600GB XDF files from above (168k events) Objectivity event store: 50GB, stopped when disk space exhausted Want to load all 600+ GB into Objectivity  Event database files HPSS resident  Management via GC software in STAF  In about 2 weeks, after a schema and code update Already the 50GB federation is migrated to HPSS (no space), so the sooner we can offer GC-managed access to CAS STAF users the better  Applications ready or nearly ready to use the event store DB: l Event by event scale correlation analysis (SCA) l DST QA evaluation code l Strangeness analysis

STAR C OMPUTING Torre Wenaus, BNL Grand Challenge Meeting 10/98 Schema and Structure Update Changes to database schema and code before rebuilding the federation:  Individual container per collection in header database to minimize lock contention  Correction to cluster header: tag objects should not be referenced  Add event ID to cluster header  Storage of run info, summary in collection header must be fixed; problem with BaBar inheritance in StOdbEventCollection  Replace persistent strings to control object pollution  New vertex table from strangeness group (reduced ev0_aux)  Review and update of run specification and run configuration specification in DST tables l Specification of detector/parameter specification via configuration key

STAR C OMPUTING Torre Wenaus, BNL Grand Challenge Meeting 10/98 Database File Organization Top-level metadata and tag files in $STAR_DB/stardb (sol disk)  Event collections and headers  Tag headers and tags Single directory containing event data $STAR_DB/stardb/dst (disk1)  Can live with single disk directory for the moment, but will need multi-directory capability soon (when we split event data into >1 cluster)  Event database files correspond to reconstruction jobs with same name as XDF file HPSS migration by moving event data files to /starreco/stardb/dst in HPSS  Compatible with GC? All DB loading so far done serially  BaBar has seen problems in parallel loading; we need to test

STAR C OMPUTING Torre Wenaus, BNL Grand Challenge Meeting 10/98 Near term Use of GC Software “Production” usage mode: independent activity at the physics analysis group level  At the physicist level if we can get away with it (performance) Management of the HPSS event store for STAF CAS users is the critical path item  Retrieval by collection will be initial usage mode Beyond that, would like to have available to users  Query estimation and retrieval by query  Retrieval by event list  Logging, statistics, HPSS performance info  Tagdb visualizer Need direct API from TagDB to index builder, no intermediary file  Simple ‘rebuild the index’ procedure

STAR C OMPUTING Torre Wenaus, BNL Grand Challenge Meeting 10/98 Event Store Database on Linux Only operational platform for Objy event store at present is Sun/Solaris No STAR Objectivity operations on Linux  Biggest problem: BaBar/STAR database software port to Linux l No attempt made yet to build on Linux l At least some BaBar software reportedly needs egcs or 2.8.1, so we may have to wait for the next Objy/Linux version l Rogue Wave reportedly does compile and work but not officially supported l Will have a shot at building the BaBar essentials with gcc and see whether encountered problems are surmountable  If we get past that, until we get Objy version with Linux platform index overflow fixed, could build duplicate federations for Solaris and Linux So, timescale for wanting to use GC from Linux unclear

STAR C OMPUTING Torre Wenaus, BNL Grand Challenge Meeting 10/98 Use of ROOT Assumption:  ROOT usage of event store and ROOT usage of GC software are the same problem; when we solve the former we will have solved the latter  No specific issue for GC ROOT+Objy not touched yet. Lower priority than getting everything operating well under STAF.

STAR C OMPUTING Torre Wenaus, BNL Grand Challenge Meeting 10/98 MDC2 Wish List (Preliminary) Multiple separately stored clusters per event Retrieval of non-Objectivity components  Objy Event store stores file reference, not the data itself  We still pass an event ID and (set of) cluster ID’s to be retrieved, but you recognize cluster ID’s that refer to to non-Objy files  Uses: l XDF fall-back for Objectivity! l ROOT-format components l Raw data retrieval Retrieval of components from multiple COS’s Dynamic updating of index Ability to specify (at administrator level) ‘keep this data on disk’  Ability to manage what material we keep on disk