Ultralight/PLaNetS Tutorial

Tuesday, Dec. 2nd, 2008

Florida International University
Kovens Conference Center

3000 NE 151 Street

Miami, Florida 33181

Hosted by CIARA

   

Time

Title

Description

Speaker

7:00-8:00
(60 min)

Breakfast

Location:  Alcove South
                                             

 

8:00-9:00
(60 min)

Dynamic Circuit Networks

Location: Bayview Ballroom South 214A

This tutorial provides an introduction to the use of dynamic circuit networks in data intensive research.

Today's data intensive applications, like High Energy Physics research at the LHC, rely on efficient bulk data transfers of large datasets over distances spanning several continents. Such transfers, in order to proceed efficiently, require a guaranteed bandwidth from source to destination, something that is hard to achieve on the general-purpose network like the Internet.

Other examples of research applications using dynamic circuit provisioning are gravitational wave detectors (LIGO) and radio astronomy. Both rely on guaranteed bandwidth channels between sites distributed around the globe.

Dynamic circuit networks are designed to provide such bandwidth guarantees by creating virtual circuits upon request from the user or user application.

In this tutorial we will cover the concepts of dynamic circuit networks and their interoperation for multi-domain provisioning, necessary for creation of end-to-end circuits on large geographical scale. We will discuss applications such as LambdaStation and TeraPaths, making use of dynamic circuit provisioning.

A live demonstration will show the actual process of creating a light-path from a user/application perspective.

Dr. Artur Barczyk
California Institute of Technology
and
Mr. Azher Mughal

9:10-9:50
(40 min)

Configuring Servers To Do Maximum Throughput Bandwidth

Location: Bayview Ballroom South 214A

A presentation of how to select hardware for optimum disk and network bandwidth. A set of common tools used to benchmark disk and network hardware in order to detect where bottlenecks occur will be demonstrated.

 

Mr. Michael Thomas
HEP Systems Administrator
California Institute of Technology

10:00-10:20
(20 min)

AM Break

Location:  Alcove South

10:20-11:15
(55 min)

End System Tuning to Maximize Data Transfer Performance

Location: Bayview Ballroom South 214A

A hands-on tutorial showing some of the common ways to measure and tune an end host to maximize data transfers.

Mr. Michael Thomas
HEP Systems Administrator
California Institute of Technology

11:25-12:00
(35 min)

Virtualization and Cloud Computing

The Enterprise Cloud - an example of Cloud Computing

Mr. Kurt W Robinson
Sales Engineer
Terremark Worldwide

Mr. John Zima
Terremark World Wide Inc. NAP of the Americas

Corporate
Sponsor

12:00-1:00
(60 min)

Lunch

Location:  Bayview Ballroom North 214B

 

1:00-2:30
(90 min)

 

MonALISA Monitoring

Within the large distributed LHC infrastructure monitoring is vital to pinpoint anomalies and quickly respond to them, but also to identify trends. This tutorial will discuss the distributed monitoring framework MonALISA. You will learn how to setup MonALISA services, publish parameters, deploy dynamic web pages displaying various monitoring parameters, and setup alerts. In this tutorial we will deploy MonALISA services on hosts at FIU, Caltech and U. Florida. Attendees with access to a computer can participate by installing their own MonALISA services.
The only system requirements are java 1.4.2 or greater and internet connectivity through a set of configurable ports. A recent version of python is recommended as well. The tutorial has been tested on Linux but other platforms should work.

Dr. Jorge Luis Rodriguez,
Florida International University

Dr. Frank van Lingen
Grid Job Submission
California Institute of Technology, CERN

2:30

 

PM Break

Location:  Bayview Ballroom North 214B
(Participants will be provided access to refreshments during the meeting.)

2:30-4:00
(90 min)

Grid Job Submitting

Location: Bayview Ballroom South 214A

Submitting a job to the Grid is one of the two fundamental tasks a user engages in when utilizing grid resources. In this tutorial you will learn how this is done on the world wide distributed infrastructure known as the Open Science Grid (OSG). We will begin by describing the authentication and authorization mechanism used to allow users onto resources and then illustrate the process of submitting jobs. We will do this with a series of example beginning with a simple "Hello world" submission, and then we will submit jobs that will interact with monitoring agents and retrieve status information. Finally we will demonstrate process used by the researchers on the Compact Muon Solenoid experiment one of the larger Virtual Organizations on the OSG and one of the LHC experiments.

Dr. Jorge Luis Rodriguez,
Florida International University

Dr. Frank van Lingen
Grid Job Submission
California Institute of Technology, CERN

4:10-5:10
(60 min)

FDT Tool

Location: Bayview Ballroom South 214A

This hands-on tutorial will introduce the Fast Data Transfer (FDT) tool and show how it can be used to transfer data between end hosts in an optimized manner. At the end of the tutorial we will have a brief introduction to the dCache storage system and show how FDT has been integrated with dCache. Users who have access to an existing dCache system will be encouraged to follow along.

Mr. Michael Thomas
HEP Systems Administrator
California Institute of Technology

5:10

Program ends