DOSAR Workshop Sept. 27-28, 2007 J. Cochran 1 The State of DOSAR Outline What Exactly is DOSAR (for...

20
1 DOSAR Workshop Sept. 27-28, 2007 J. Coch The State of DOSAR Outline hat Exactly is DOSAR (for the new folks) rief History oals, Accomplishments, & Strategy evelopments Since Last Workshop n the Horizon … onclusions

Transcript of DOSAR Workshop Sept. 27-28, 2007 J. Cochran 1 The State of DOSAR Outline What Exactly is DOSAR (for...

1DOSAR Workshop Sept. 27-28, 2007 J. Cochran

The State of DOSAR

Outline

What Exactly is DOSAR (for the new folks)

Brief History

Goals, Accomplishments, & Strategy

Developments Since Last Workshop

On the Horizon …

Conclusions

2DOSAR Workshop Sept. 27-28, 2007 J. Cochran

– Community and campus based grid organization

– Its primary goal is spreading/enabling grid use• DOSAR in Korean is the God of Martial Arts

• DOSAR evolved from the DØ Remote Analysis effort– Coordinated group efforts in DØ simulation/reconstruction/production

• One goal is to work closely with other disciplines

• Each group should play a leadership role in realization of computing grids in corresponding campuses and states

What Exactly is DOSAR ?

Distributed Organization for Scientific & Academic Research

(nucleation points …)

3DOSAR Workshop Sept. 27-28, 2007 J. Cochran

History

Ancient times: DØ Remote Computing Era

– SAM, DØ data management system, in place: pre-2001

– DØ Remote Analysis Model Proposed: Nov. 2001

– Proposal for RAC accepted and endorsed by DØ: Aug. 2002

– UTA awarded MRI for RAC: June 2002

– Formation of DØ Southern Analysis Region (D0SAR 1.0): Apr. 2003• D0SAR DØ MC Production begins

– Activation of 1st US (Regional Analysis Center) RAC at UTA: Nov. 2003

– Formation and activation of DOSAR Grid for MC: Apr. 2004

4DOSAR Workshop Sept. 27-28, 2007 J. Cochran

Desktop Analysis Stations

Institutional Analysis Centers

Regional Analysis Centers

Normal InteractionCommunication Path

Occasional Interaction Communication Path

Central Analysis Center (CAC)

DAS DAS…. DAS DAS….

IAC ... IAC IAC…IAC

RAC….

RAC

DØ Remote Analysis Model (DØRAM)

Fermilab

•Data and Resource hub•MC Production•Data processing•Data analysis

5DOSAR Workshop Sept. 27-28, 2007 J. Cochran

History

– Transition to Distributed Organization for Scientific and Academic Research,

DOSAR: Apr. 2005 (DOSAR 2.0)

• Active engagements with LHC experiments began

– DOSAR VOMS installed at UTA: May 2005

– DOSAR registered as a VO in OSG: July 2005

– ATLAS distributed production and analysis system,

Panda, implemented at OU and UTA: Jan. 2006

– All groups engaged in LHC experiments

Modern times: beyond DØ

6DOSAR Workshop Sept. 27-28, 2007 J. Cochran

Who is DOSAR ?

First Generation IAC’s University of Texas at Arlington Louisiana Tech University Langston University University of Oklahoma Tata Institute (India)

• Second Generation IAC’s

– Cinvestav, Mexico SPRACE - São Paulo Regional

Analysis Center, Universidade Estadual Paulista; Brazil

University of Kansas Kansas State University

• Third Generation IAC’s– University of Arizona, Tucson– Rice University Ole Miss, University of Mississippi

Iowa State University Louisiana State University Oklahoma State University …

7DOSAR Workshop Sept. 27-28, 2007 J. Cochran

Goals (preamble)

Many branches of science routinely reside at the boundaries of computing technology

With projects which span many institutions, computing resources must be pooled (GRID)

As with many resources, the needs often exceed what is available(even when carefully pooled)

The projections for many current & future projects fall in this category(esp LHC computing)

At the same time, we realize that at most institutions, the vast majority of computing cycles are idle a large fraction of the time

8DOSAR Workshop Sept. 27-28, 2007 J. Cochran

Goals

DOSAR seeks to bring together institutions which can/must use GRID technology to solve their computing problems (esp smaller institutions)

As needs increase faster than funding, the scientific community will need to tap into these un(under)-used resources

but, there are many barriers to utilizing these resources (technical, financial, manpower, …)

DOSAR should, as necessary, serve as a representative/liason for these institutions to the greater GRID community

DOSAR should help enable maximal use of available resources at the member institutions (all institutions ?)

(strength in collaboration)

(member institutions should serve as nucleation points for regional/statewide GRIDS)

9DOSAR Workshop Sept. 27-28, 2007 J. Cochran

Accomplishments

→ LTU brought in Super-Mike and additional resources for DØ & ATLAS

→ UTA plays leadership role in HiPCAT, Texas Grid community; Leading BioTex grid, working with chemists, geologists and medical professionals

→ LTU Working as a leading institution in LONI, Louisiana Grid community

→ OU and LU have been working toward creating “The Oklahoma state grid”, working closely with OSCER

→ SPRACE leads Brazilian national grid effort w/ funds!!!

10DOSAR Workshop Sept. 27-28, 2007 J. Cochran

DØ Production

11DOSAR Workshop Sept. 27-28, 2007 J. Cochran

Breakdown of DOSAR Contribution to DØ Production

12DOSAR Workshop Sept. 27-28, 2007 J. Cochran

DOSAR Weekly Production by site(common scale)

13DOSAR Workshop Sept. 27-28, 2007 J. Cochran

DOSAR Weekly Production by site

14DOSAR Workshop Sept. 27-28, 2007 J. Cochran

DØ Data Reprocessing

Substantial DOSAR contribution ?

15DOSAR Workshop Sept. 27-28, 2007 J. Cochran

Some Successes in DOSAR Funding

• Funds from NSF MRI for UTA – RAC: 2002– Construction of the first U.S. university based RAC

• EPSCoR + University funds for LTU – IAC: 2003– Increase IAC compute resources

• Brazilian National Funds for SPRACE: 2003 & 5– Construction of an extensive RAC for

• EPSCoR funds for OSU, OU & LU: 2004– Compute resources for IAC– Human resource for further development– Renewal submitted

• LTU at part of LONI wins support from State of LA: 2005

• OU, LU and UTA, together with UNM, won a joint ATLAS Tier 2 site: 2005

• LTU won a joint MRI funds: 2006

16DOSAR Workshop Sept. 27-28, 2007 J. Cochran

DOSAR Strategy

• Maximally exploit existing software and utilities to enable as many sites to contribute to the DØ and LHC experiments– Continue to participate in DØ MC and Reco activities in opportunistic

manner– Focus on LHC experiments– Support OSG move into the global grid framework– Engage in and contribute significantly in OSG

• Engage in realization of computing grid beyond HEP– Work closely with campus and state computing people to bring grid onto

campuses• Bring DOSAR specific computing jobs to the grid• Want to make grid usage more commonplace

17DOSAR Workshop Sept. 27-28, 2007 J. Cochran

Working with OSG

• We accomplished virtually everything we wanted in DØ– J. Snow (LU) is the leader in MC production

• Bring fresh expertise to OSG

• Expertise in monitoring solutions such as Ganglia and MonALISA UTA working on Panda ATLAS monitoring together with CMS experiment

• LTU working on first implementation of HAOSCAR (High Availability Open Source Cluster Application Resources) w/ OSG stack

18DOSAR Workshop Sept. 27-28, 2007 J. Cochran

Contributions to OSG

• Testing of framework, middleware, and user interfaces.

• Active participation in OSG integration and deployment activities.

• Partner with high-speed optical network initiatives

• Help implement and utilize grid computing technology for educational use.

• Participate and test grid-based HEP data analysis and disseminate the experience to OSG

19DOSAR Workshop Sept. 27-28, 2007 J. Cochran

Developments Since Last Workshop

Potential expansion of DOSAR “in to Africa” …

Participation of DOSAR in NSF SUCCEED Proposal to develop GRID Outreach and Educational Opportunities

Representation of DOSAR on the OSG Council

On the Horizon …

Imminent expansion of OSG(+ATLAS) to MS Windows platform - this is an extremely significant development (congrats to OU group!)

ATLAS testing on DOSAR sites - use of Tier3 sites in ATLAS prod

20DOSAR Workshop Sept. 27-28, 2007 J. Cochran

Conclusions

• DOSAR is an example of a successful grid organization– Critically important asset to DØ

• All groups actively engaged in LHC experiments

• Using DOSAR resources for DØ and LHC data analyses and production

• Closely engaged in OSG activities

• Continue Desktop Cluster deployment, CONDOR pools

• Provide leadership role in state-wide grid efforts

• Expanding membership and scope