Information Technology Reference
In-Depth Information
Metacomputing systems today primarily take the form of networks of workstations
(NOWs) [3], that allow people to pool together existing and mostly idle workstations in
their own local or institution-wide networks and use them to do parallel processing without
having to purchase an expensive supercomputer [4]. More recent metacomputing projects
have started going beyond the bounds of individual institutions, and have started address-
ing the problems of security, reliability, and scalability in nation-wide and global networks.
These projects include Legion [5], Globus [6], NetSolve [7], and Ninf [8]. Since then, a lot
of metacomputing research have been focused on creating “grid computing” [9] environ-
ment, which aim at providing seamless access to computing resources around the country
and the world.
VC is a form of metacomputing that focuses on maximizing the ease with which people
can have their machines be part of a metacomputer [1]. While other forms of metacom-
puting such as grid computing seek to make “using” compute servers on the network easy,
“setting up” these compute servers is currently not as easy and requires helps of experts.
Thus, metacomputing systems still have complex setup requirements that prevent ordinary
users from participating in these systems. The idea behind VC is to eliminate these setup
requirements, and make it so easy to join a computing network that anyone, even a casual
computer user without any technical knowledge, can do it. By allowing casual users to
contribute their computers' processing power in this way, VC makes it possible to harness
the power of thousands, or even millions, of computers in very little time.
In recent years, there are two big reasons why VC should be focused on [2]. The first
reason is that the prices of personal computers have dropped drastically. The price drops
have made it possible for many households to have one or more computers. In addition to
family computers, many college students have their own computers and the large number of
colleges have many computers in labs and libraries for student use. A lot of these computers
are connected to the Internet and many computers remain turned on 24 hours a day, only
being used for part of the day and sitting idle during the rest of the day. These computers
have the potential to be used for VC during those idle periods.
The second is that there is a trend of putting more than one core on a CPU so that
personal computers are now coming with multiple processing cores instead of a single one.
While personal computers with these multi-core CPUs are now truly capable of doing multi-
ple tasks at the same time, people still have trouble doing multiple tasks at once. Therefore,
we expect that there will be an increase in the number of CPU cycles that are used by the
system' s idle process. Other people than the computers' owners could potentially use these
CPU cycles without noticeably affecting the computers user' s experience. These changes
have the potential to really increase the amount of CPU cycles donated to VC projects.
1.2.
Problems and Motivations
While there has been a rapidly growing interest in VC as a ultrahigh performance comput-
ing environment, VC still has a mandatory issue for reliable computing. Different from the
grid computing system which shares managed reliable computers within or among organi-
zations, computing resources of VC system are owned and managed by volunteer partici-
pants. Those resources may behave erratically due to hardware/software failures or virus
infection, or may behave maliciously to falsify the computation, each of which results in
