| Two-tier
Client/Server model The term Client/Server
has several definitions in several contexts. In
our context, two different applications run
either on the same machine or other machine and
both of these applications talk to each other.
One application acts as a server, and the other
machine as a client.
A typical
example for the server application would be a
database server. It listens requests from the
clients and sends the results back to the client.
Typically a server application is located on a
machine which has a lot more horse-power and
serves to lot of clients.
Another
application acts as a client application and
typically loaded on a normal machine. The client
application makes requests to the server
application. For example, a PowerBuilder
application sending a SQL query to the database
and displaying the results on the screen.
The OLE context
has a little different definition, but, we
won’t be going into OLE now. We won’t
be teaching the evoluation of computers, but, we
need to tell about the conventional computers in
few sentences.
Conventional
applications run in the same memory space. For
example, a CICS/COBOL application that runs on a
dumb terminal. You need to run that application
only on a IBM mainframe terminal. Even if you run
the application on your powerful penitium machine
using a terminal emulator, still, the actual
application runs on the mainframe, not on your
PC. That means, you can’t use your powerful
desktop machine, it is just useless when you run
the CICS application. On the other hand when
applications are designed as per the
Client/Server architecture, you don’t have
to through away your PC. In the Client/Server
architecture, as long as the client and the
server is able to talk, it doesn’t matter on
which platform the applications are running.
We can divide
an application into three logical parts,
Presentation, Business logic, data handling. In
conventional applications such as COBOL/CICS, all
these three parts are part of one application. In
Client/Server applications, the client
application contains the presentation logic, data
handing is done in the server application. The
business logic goes either in the client or in
the server or split between both the client and
server application.
When all the
business logic is put in the client application,
it is called FAT client model. On the other hand
when all the business logic is put in the server
application, it is called THIN client model.
Where exactly should the business logic should go
is the choice of the origanization.
|
