WSP07134 - Laserfiche WebLink

Home Browse Search

e e ~e 003H2 few combinations of centralized/distributed data processing. When processing centrally, a single machine or set of machines (e,g, a single database server) is used to store data and process requests to fetch and modify this data (Figure 2), The main advantage of this approach is that there is only one copy of the data (i.e. there is no data duplication); thus, there is a high probability of maintaining data integrity. The main disadvantage of this approach is that all requests for data come to the database server therefore creating the possibility of a bottle-neck in processing the requests. The centralized data/centralized processing concerns a configuration used predominantly during the 1960's and 1970's (e.g, time sharing and .batch processing). There are several commercially available database management systems that support centralized data management and processing The option of distributed processing allows data to be stored and processed on many machines (possibly over a wide area). Neither the users nor the application programs have to be aware of where the data is physically stored. Data in a distributed database (Figure 3) can be distributed with or without duplication. An advantage of distribution with data duplication is that if one of the database servers is down, back-up copies of the data on other servers exist. However, this could lead to the risk of data being accessed that is out of date. This problem can be addressed by having management protocols in place that assure a high probability of distributing updates to the database at reasonable time increments, An advantage of distribution without data duplication is less storage and greater data integrity. However. if one of the servers goes down, some of the data will not be accessible, There are a few commercially available database management systems that support distributed databases without data duplication, There are no geographic information systems that support distributed data. Data exchange is an important pan of the system configuration since there is a clear need for users from remote sites to have access to data and programs. There are a number of options for moving data/programs between users: media transfer, batch file. and interactive data communications. Media transfer involves physical media such as tapes and diskettes. Data and! or programs are copied onto the physical media, The physical media are then physically moved (e.g, via US Mail. etc.) to location(s) where they are needed, At those sites data and programs are then copied onto the local machine(s); thus, there is duplication of data. Data being moved on the physical media must be in a format that the destination machine can interpreI (e.g, ASCII). The obvious problem with this method is that moving data from one machine to another can be a slow process(e,g, a minimum turn around time of a few hours to several days). Centralized Processing Distributed Processing Data Exchange I I I I I I , DAMES& MOORE/CA DSWES -25