Home | Previous Page | Next Page   The Database Server > Introducing the Database Server > Extended Parallel Server >

Extended Client/Server Operations

Extended Parallel Server extends the client/server architecture of Dynamic Server to the parallel-processing platforms and multiple coservers.

Connection Coserver

A connection coserver is the coserver that manages a client connection to a database server. The connection coserver determines if a client request can be satisfied with data that resides on that coserver alone or if it requires data that resides on other coservers.

Different clients can connect to different coservers. For example, Client A can connect to Coserver 1, and Client B can connect to Coserver 2. Coserver 1 is the connection coserver for Client A, and Coserver 2 is the connection coserver for Client B.

Participating Coserver

When a client database request requires access to data that resides in table fragments on other coservers, the other coservers are called participating coservers.

Figure 3 shows a client that is accessing a very large database that is fragmented across many coservers. Coserver 1 is the connection coserver. Coservers 1 through N are all participating coservers.

Figure 3. Client Query That Coservers Service
begin figure description - This figure is described in the surrounding text. - end figure description

Online Transaction Processing and Decision-Support Applications

Extended Parallel Server provides support for both OLTP and DSS applications.

DSS applications perform more complex tasks than OLTP, often including scans of entire tables, manipulation of large amounts of data, multiple joins, and the creation of temporary tables. Such operations involve many calculations and large amounts of memory. As a result, DSS execution times are typically far longer than OLTP execution times. However, Extended Parallel Server excels at handling the complex queries that are typical of decision-support applications.

Another approach to DSS operations, called a data mart, draws selected data from OLTP operations or a data warehouse to answer specific types of questions or to support a specific department or initiative within an organization.

Extended Parallel Server provides exceptional performance for DSS applications such as data marts and data warehouses.

For more information about data mart applications, refer to the IBM Informix: Database Design and Implementation Guide.

Extended Parallel Server does not support distributed data queries where a single transaction contains queries to more than one database across multiple database servers.

Home | [ Top of Page | Previous Page | Next Page | Contents | Index ]