Introduction: To define and explain a distributed system by defining it properties
Within the system is found single entities which survive on its own with its individual capabilities and components that make up its own properties however it has the capabilities to communicate with others alike by passing messages from point to point. The system and all the single entities maintain one common goal, which is to work together for a relative commonality yet still work independently of one another.
The attribute of this article is to explain two broad distributed system models and defining their properties.
There are two principal reasons for using distributed systems. Firstly, the very nature of the application may necessitate the use of a communication network that links several computers.
Secondly is a network capable computer that works individually, independently working off the network however still benefiting in practical on the network scenarios.
A distributed system is beneficially more reliable more cost effective and can obtain a healthy level of performance by using a collection of numerous low-end computers, in comparison with a single high-end computer.
The hybrid distributed system or a monolithic uniprocessor system and otherwise known as a Fault tolerant cell array. Which is a system containing an array of computers with individual aspects. Supper fast multithreading and multi core processers running servile programs at the same time, less bottlenecks and it will support massive parallel data processing architectures into a single monolithic entity.
The workstation server
A server is an application or device that performs service for connected clients as part of client server architecture. It can also be a computer system that has been designated for running a specific server application. A server can also serve applications to users on an intranet. (Compair, 2011)
Servers will keep files organized and provide applications and share points, and connect those in the system to the internet.
Workstations themselves are typically designed for specific high-end computers found in graphics design, audio and video editing typically speaking.
A typical organization such as the Government will have a large number of computer workstations, laptops and client desktops dispersed over wide geographical area and bridging across the globe. The premise would be to have each unit configured with an operating system that may include similar and dissimilar features hardware and software configurations pending on the geographical area in which the unit would be deployed with one or more common applications that will interact with the main entity (the server) by using the World Wide Web or an Intranet.
The providing advantages the implementations of multiple unites spread across the globe allows information to be gathered from each location and transmitted back to the main server.
The shared processer pool allows an assigned partial partition to a specific user or group of users. This concept is done through the logical makeup of a server or server array. These processes can also be implemented logically into a global distribution system also. Commonly could be labeled as a Share drive and or server where one or more drive partitions in a single drive or physical drives are assigned. Each group or single unit will be allocated to that assigned space. Assigned logical partitions are basically configured shared processor pools. You can assign a logical partition to a shared processor pool at the time that you create the logical partition, or by reassigning existing logical partitions.
Common workload issues
IBM has been powering processer pools for a long time since the POWER5 and now POWER6 processors came along. They were designed to handle multiple shared processor pools. Unlike UNIX and some other distributed systems are likely dedicated to single applications, designed to handle the demand of high and low peaks of services or common workload. Units using the pooling system will take advantage of having less traffic throughout its connection to the server so the only high demand will be among those using that partition. By creating separate shared-processor pools and putting the applications that have the same or similar software needs together provides a differential licensing-cost reduction. They allow greater control of the percentage of processor resources. They can be used to isolate the resources that are available to the units within the shared pools. They also can be used to separate workloads.
Maintenance and security can be most difficult and make things seem like having multitudes of units a disadvantage. Within a closed system or an open system such as globally deployed units the security risks are high, granting the open units security is higher. The obvious disadvantages of this type are replacements, problematic fixes, enforcing codes, updates, viruses, and things like theft. All of the disadvantages can be addressed and remedied by implementing policy, secure logon apps like “Smart off” and “Tokens”, and general operating and security classes by addressing proper behavioral aspects needed while using all government computers.
This is primarily used in high work load operations whereas many are in the same organization yet having different jobs with all the information leading back to a main controlled server. Nodes may makeup in part of any type of server along with separate nodes known as print servers.
The needs the wants and the things that are deployed into a system great or small are allocated by the nature or environment of which a system will operate.
Compair. (2011). Retrieved 11 18, 2011, from Diffen: http://www.diffen.com/difference
Andrew Tanenbaum. (2007). Distributed Systems. Upper Saddle Rive, NJ, USA: Pearson Prentice Hall.
Tony Sintes. (2002, 08 23). App server, Web server: What's the difference? Retrieved 11 18, 2011, from JavaWorld.com, 08/23/02: http://www.javaworld.com/javaqa/2002-08/01-qa-0823-appvswebserver.html?page=2