Chapter 14
Distributed Computing
Manuel Delfino
Distributed computing is an established discipline in computer science and engineering. It has evolved over the past 40 years to be one of the most important
methodologies for implementing the data processing services needed by almost
every activity in society. Distributed computing is still evolving rather rapidly,
with major innovations introduced every few years. The aim of this chapter is to
introduce the reader to the basic concepts of distributed computing in the context
of particle physics, allowing a better understanding of what is behind the scenes
when using distributed systems, and providing starting points in order to seek further
information on the subject.
14.1 Usage by Particle Physics
Distributed computing, i.e. the coherent use of many computers to accomplish
a given task, is extensively used in particle physics. It is particularly suited to
simulating and analyzing data from particle collisions, where processing the data
of one collision is largely independent from processing data of all other collisions
and hence very little communication between the different computers is necessary.
This type of distributed computing is called “High Throughput Computing” (HTC), 1
because there is a very large number of quasi-independent tasks to accomplish,
1 Livny, Miron, et al. “Mechanisms for high throughput computing.” SPEEDUP journal 11.1
(1997): 36–40.
M. Delfino ()
Departament de Física, Facultat de Cièncias, Universitat Autònoma de Barcelona, Barcelona,
Spain
e-mail: manuel.delfino@uab.cat
© The Author(s) 2020
C. W. Fabjan, H. Schopper (eds.), Particle Physics Reference Library,
https://doi.org/10.1007/978-3-030-35318-6_14
613
Distributed Computing
Manuel Delfino
Distributed computing is an established discipline in computer science and engineering. It has evolved over the past 40 years to be one of the most important
methodologies for implementing the data processing services needed by almost
every activity in society. Distributed computing is still evolving rather rapidly,
with major innovations introduced every few years. The aim of this chapter is to
introduce the reader to the basic concepts of distributed computing in the context
of particle physics, allowing a better understanding of what is behind the scenes
when using distributed systems, and providing starting points in order to seek further
information on the subject.
14.1 Usage by Particle Physics
Distributed computing, i.e. the coherent use of many computers to accomplish
a given task, is extensively used in particle physics. It is particularly suited to
simulating and analyzing data from particle collisions, where processing the data
of one collision is largely independent from processing data of all other collisions
and hence very little communication between the different computers is necessary.
This type of distributed computing is called “High Throughput Computing” (HTC), 1
because there is a very large number of quasi-independent tasks to accomplish,
1 Livny, Miron, et al. “Mechanisms for high throughput computing.” SPEEDUP journal 11.1
(1997): 36–40.
M. Delfino ()
Departament de Física, Facultat de Cièncias, Universitat Autònoma de Barcelona, Barcelona,
Spain
e-mail: manuel.delfino@uab.cat
© The Author(s) 2020
C. W. Fabjan, H. Schopper (eds.), Particle Physics Reference Library,
https://doi.org/10.1007/978-3-030-35318-6_14
613