Validation of the MONARC Simulation Tools

1
Validation of the MONARC Simulation Tools

• Y. Morita
• KEK Computing Center (CERN IT/DB)
• for the MONARC Collaboration

2
MONARC

• Models Of Networked Analysis
• at Regional Centres
• http//www.cern.ch/MONARC/
• Primary goal is to identify baseline Computing
  Models that could provide viable solutions
  meeting the data analysis needs of LHC
  experiments.
• Boundary conditions are network bandwidth,
  computing power, distributed database systems,
  data processing capabilities available at the
  start of LHC (year 2005).
• See H.Newman's talk on Feb/10 plenary session.

3
Working Groups

• Architecture WG
• Baseline architecture for regional centres,
  Technology tracking, Survey of computing model of
  current HENP experiments
• Analysis Model WG
• Evaluation of LHC data analysis model and use
  cases
• Simulation WG
• Develop a simulation tool set for performance
  evaluation of the computing models
• Testbed WG
• Evaluate the performance of ODBMS, network in the
  distributed environment

4
MONARC Simulation

• Process Oriented Discrete Event Simulation
• assign active tasks in the system to threads in
  the Java program, (re-)calculate the time of
  elementary operations such as CPU, dirk I/O,
  network transfer
• developed by I. Legrand
• See K.Sliwa's (I.Legrand's) talk on Feb/8 (F148)

5
Data Server Model

• Each component in the model is an abstraction of
  the real system
• Database architecture follows the Objectivity
  implementation
• Time response function for the interacting
  components are modeled in the simulation

6
Performance Evaluation

• Develop a model of the system and simulate of the
  performance
• Validate the simulation results with the actual
  measurements
• Elaborate the simulation model and measurements
• Iterate the cycle to the required "level of
  detail"

7
Testbed DB Application

• Object Model inAtlas Simulated Raw Events (Atlas
  0.0.24 in 1TB milestone)

8
LAN measurements
(1)

• Machine A
• Sun Enterprise450 (400MHz 4x CPU)
• Machine B
• Sun Ultra5 (270MHz) -- Lock Server
• Machine C
• Sun Enterprise450 (300MHz 2x CPU)
• Tests
• (1) Machine A local (2 CPU)
• (2) Machine C local (4 CPU)
• (3) Machine A (client) and Machine C (server)
• number of client processes 1, 2, 4, ..., 32

(2)
(3)
9
Measurements
job execution time
Aggregated CPU
10
Modeling of the Client Job
Event
Time
Job on Machine A
CPU 17.4 SI95 I/O 207MB/s _at_ 54MB file
Job on Machine C
CPU 14.0 SI95 I/O 31MB/s _at_ 54MB file

• TI/O and TCPU can be extracted from single client
  test on two different machines with SPECint95
  ratings and the disk I/O speed measurements
• Same job parameters are put into the simulation
  of multiple clients

11
Disk I/O

• Measured with write() read() system calls with
  various block sizes and file sizes
• File size MB 54 864
• Block size MB 54 1024 54 1024
• Machine A write MB/s 14.8 16.2 13.5-9.2 15.9
• " read MB/s 207 207 18.5 22
• Machine C write MB/s 18.3 25.9 20.7-8.0 27
• " read MB/s 31 44 22.8 28.3
• Reading 54MB file fits on the disk cache on
  Machine A

12
Modeling of AMS

• AMS page transfer latency is modeled into the
  simulation
• Physical Bandwidth B
• Effective Bandwidth Beff
• ?T ?t(transfer) ?t(handshake)
• unit_size / B RTT
• Beff unit_size
• ------ -----------------------------
• B unit_size B RTT

13
Other Modeling Details

• Client SMP architecture is modeled with the same
  number of CPU farm with high speed network
  connection

Network
14
Simulated Results
15
Job execution time

• Job execution time competing for the same resource

16
Conclusions

• Concurrent job profile of Objectivity testbed
  application is extracted and reproduced by the
  MONARC simulation program.
• Objectivity AMS client/sever performance is
  evaluated and simulated.
• Other part of components have to be validated.
  Next major component will be the hierarchical
  mass storage system.
• FTP data transfer model instead of AMS will be
  used for the study of regional center.