08.03
Adi Parmono
Distributed system is a system consisting of several computers and connected in a network where all computers contribute to each other to complete problem. In a distributed system, there is a sub-system capable of providing job to each system that wants to join and receive back the results. To build a distributed system, you should use something standard which has been implemented by various platforms. So the system can be made used extensively. In addition to the standard use of technology, consideration should also problem of funds to be spent to build a distributed system to be suppressed minimum.
The main thing that is very influential in the funding problem is the hardware. Good news that can be taken into consideration is the hardware (in this case a computer) that already out of date can be used to build a distributed system. Technology in the distributed system consists of several categories, including:
1) Socket
This technology is the forerunner of the distributed system. Generally, there are concepts a server that will open the socket and defines the various rules and clientclient perform their respective duties by following the rules. This technology very difficult and complex because it takes the ability to make such new protocol.
2) Remote Procedure Call (RPC)
In RPC there is a sub-system that defines the various rules in the form procedure, then other sub-systems is to call the procedure completing their respective duties.
3) Distributed Objects in a single programming language
Example implementation in Java Remote Method Invocation (RMI), DOPY, and Pyro.
4) The object is distributed without regard to differences in programming language
Example of implementation on Microsoft's DCOM and CORBA.
5) other technologies such as Message-oriented middleware, mobile agents, and so forth.
One of the standard technology that can be used in building a distributed system is the Remote Procedure Call (RPC). In the context of a distributed sytem, a sub-system defines some rules (in this case a procedure) and other systems possible join to solve the problem based on a defined procedure. Technology RPC has been considered more or less around 1976. This technology requires models addressing servers, transport protocols, and type marshalling. Several projects have been successfully distributed system which made them:
1) Abacus is a distributed system that combines 17 Pentium III computers with 2 processors each (total there are 34 CPUs) with Linux system used Internet-based multimedia applications. Abacus was built in 2000 and
rumored as one of the best computing devices in Hong Kong.
2) Berkeley Open Infrastructure for Network Computing (BOINC) developed by team from the University of California, Berkeley USA.
3) Einstein @ home is a project that uses distributed systems in performing the search for spinning neutron stars using data from LIGO and the detector GEO gravitational wave.
4) Climateprediction @ net is a distributed system used in the calculation climate change.
But behind the benefits, of distributed systems also raises new issues such as data concurrency, latency, and a variety of other possible errors. This is something that is considered reasonable in the development of information technology and a challenge for IT world to find a real solution and appropriate
1) Socket
This technology is the forerunner of the distributed system. Generally, there are concepts a server that will open the socket and defines the various rules and clientclient perform their respective duties by following the rules. This technology very difficult and complex because it takes the ability to make such new protocol.
2) Remote Procedure Call (RPC)
In RPC there is a sub-system that defines the various rules in the form procedure, then other sub-systems is to call the procedure completing their respective duties.
3) Distributed Objects in a single programming language
Example implementation in Java Remote Method Invocation (RMI), DOPY, and Pyro.
4) The object is distributed without regard to differences in programming language
Example of implementation on Microsoft's DCOM and CORBA.
5) other technologies such as Message-oriented middleware, mobile agents, and so forth.
One of the standard technology that can be used in building a distributed system is the Remote Procedure Call (RPC). In the context of a distributed sytem, a sub-system defines some rules (in this case a procedure) and other systems possible join to solve the problem based on a defined procedure. Technology RPC has been considered more or less around 1976. This technology requires models addressing servers, transport protocols, and type marshalling. Several projects have been successfully distributed system which made them:
1) Abacus is a distributed system that combines 17 Pentium III computers with 2 processors each (total there are 34 CPUs) with Linux system used Internet-based multimedia applications. Abacus was built in 2000 and
rumored as one of the best computing devices in Hong Kong.
2) Berkeley Open Infrastructure for Network Computing (BOINC) developed by team from the University of California, Berkeley USA.
3) Einstein @ home is a project that uses distributed systems in performing the search for spinning neutron stars using data from LIGO and the detector GEO gravitational wave.
4) Climateprediction @ net is a distributed system used in the calculation climate change.
But behind the benefits, of distributed systems also raises new issues such as data concurrency, latency, and a variety of other possible errors. This is something that is considered reasonable in the development of information technology and a challenge for IT world to find a real solution and appropriate
Here I will provide a solution to network programming using Python, for ease in network programming, in order to manage the network properly. Please download the following tutorial. May be useful for you all.
Posted in: Setting Network
0 komentar:
Posting Komentar