Transcript title
Phase 3 Letter of Intent (1/2)
Short: N Pages
May Refer to MONARC Internal Notes to Document Progress
Suggested Format: Similar to PEP Extension
Introduction: deliverables are realistic technical options and the associated
resource requirements for LHC Computing;
to be presented to the experiments and CERN, in support of
Computing Model development for the Computing TDRs.
Brief Status; Existing Notes
Motivations for a Common Project --> Justification (1)
Goals and Scope of the Extension --> Justification (2)
Schedule: Preliminary estimate is 12 Months from completion of
Phase 2, that will occur with the submission of the final Phase 1+2
Report. Final report will contain a proposal for the Phase 3
milestones and detailed schedule
Phase 3A: Decision on which prototypes to build or exploit
MONARC/Experiments/Regional Centres Working Meeting
Phase 3B: Specification of resources and prototype configurations
Setup of simulation and prototype environment
Phase 3C: Operation of prototypes and of simulation; analysis of results
Phase 3D: Feedback; strategy optimization
December 10,1999:
MONARC Plenary Meeting
Harvey Newman (CIT)
Phase 3 Letter of Intent (2/2)
Equipment
Needs (Scale specified further in Phase 3A)
MONARC Sun E450 server upgrade
TB RAID Array, GB memory upgrade
To act as a client to the System in CERN/IT, for distributed system
studies
Access to Substantial system in the CERN/IT infrastructure
consisting of a Linux farm, and a Sun-based data server over Gigabit
Ethernet
Access to a Multi-Terabyte robotic tape store
Non-blocking access to WAN links to some of the main potential RC
(e.g. 10 Mbps reserved to Japan; some tens of Mbps to US)
Temporary use of a large volume of tape media
Relationship to Other Projects and Groups
Work in collaboration with CERN/IT groups involved databases and large
scale data and processing services
Our role is to seek common elements that may be used effectively in the
experiments’ Computing Models
Computational Grid Projects in US;
Cooperate in upcoming EU Grid proposals
US other National Funded efforts with R&D components
Submitted to Hans Hoffmann for Information on our Intention to Continue
Copy to Manuel Delfino
December 10,1999:
MONARC Plenary Meeting
Harvey Newman (CIT)
Phase 3 LoI Status
Monarc has met its milestones up until now
Progress Report
Talks in Marseilles: General + Simulation
Testbed Notes: 99/4, 99/6, Youhei’s Note --> MONARC number
Architecture group notes: 99/1-3
Simulation: Appendix of Progress Report
Short papers (Titles) for CHEP 2000 by January 15
December 10,1999:
MONARC Plenary Meeting
Harvey Newman (CIT)
MONARC Phase 3: Justification (1)
General: TIMELINESS and USEFUL IMPACT
Facilitate the efficient planning and design of mutually
compatible site and network architectures, and services
Among the experiments, the CERN Centre
and Regional Centres
Provide modelling consultancy and service to the
experiments and Centres
Provide a core of advanced R&D activities, aimed at LHC
computing system optimisation and production prototyping
Take advantage of work on distributed data-intensive computing
for HENP this year in other “next generation” projects [*]
For example in US: “Particle Physics Data Grid” (PPDG) of DoE/NGI;
+ “Joint “GriPhyN” proposal on Computational Data Grids by
ATLAS/CMS/LIGO/SDSS. Note EU Plans as well.
[*] See H. Newman, http://www.cern.ch/MONARC/progress_report/longc7.html
December 10,1999:
MONARC Plenary Meeting
Harvey Newman (CIT)
MONARC Phase 3 Justification
(2A)
More Realistic Computing Model Development
(LHCb and Alice Notes)
Confrontation of Models with Realistic Prototypes;
At Every Stage: Assess Use Cases Based on Actual Simulation,
Reconstruction and Physics Analyses;
Participate in the setup of the prototyopes
We will further validate and develop MONARC simulation system
using the results of these use cases (positive feedback)
Continue to Review Key Inputs to the Model
CPU Times at Various Phases
Data Rate to Storage
Tape Storage: Speed and I/O
Employ MONARC simulation and testbeds to study CM variations,
and suggest strategy improvements
December 10,1999:
MONARC Plenary Meeting
Harvey Newman (CIT)
MONARC Phase 3 Justification
(2B)
Technology Studies
Data Model
Data structures
Reclustering, Restructuring; transport operations
Replication
Caching, migration (HMSM), etc.
Network
QoS Mechanisms: Identify Which are important
Distributed System Resource Management and Query
Estimators
(Queue management and Load Balancing)
Development of MONARC Simulation Visualization Tools
for interactive Computing Model analysis (forward reference)
December 10,1999:
MONARC Plenary Meeting
Harvey Newman (CIT)
MONARC Phase 3: Justification (3)
Meet Near Term Milestones for LHC Computing
For example CMS Data Handling Milestones: ORCA4: March 2000
~1 Million event fully-simulated data sample(s)
Simulation of data access patterns, and mechanisms used
to build and/or replicate compact object collections
Integration of database and mass storage use
(including caching/migration strategy for limited disk space)
Other milestones will be detailed, and/or brought forward to meet
the actual needs for HLT Studies and the TDRs for the
Trigger, DAQ, Software and Computing and Physics
ATLAS Geant4 Studies
Event production and and analysis must be spread amongst
regional centres, and candidates
Learn about RC configurations, operations, network bandwidth,
by modeling real systems, and analyses actually with
Feedback information from real operations into simulations
Use progressively more realistic models to develop future
December 10,1999: MONARC Plenary Meeting Harvey Newman (CIT)
strategies
MONARC: Computing Model
Constraints Drive Strategies
Latencies and Queuing Delays
Resource Allocations and/or Advance Reservations
Time to Swap In/Out Disk Space
Tape Handling Delays: Get a Drive, Find a Volume,
Mount a Volume, Locate File, Read or Write
Interaction with local batch and device queues
Serial operations: tape/disk, cross-network, disk-disk
and/or disk-tape after network transfer
Networks
Useable fraction of bandwidth (Congestion, Overheads): 30-60% (?)
Fraction for event-data transfers: 15-30% ?
Nonlinear throughput degradation on loaded or poorly configured
network paths.
Inter-Facility Policies
Resources available to remote users
Access to some resources in quasi-real time
December 10,1999:
MONARC Plenary Meeting
Harvey Newman (CIT)