With application deployments, clients either have a direct data connection to each data source or they connect to the Data Server. In most production scenarios it is best to use the Data Server. The Data Server uses EII and XML technologies to gather, federate and distribute information from disparate data sources based on information currently in demand. It also caches the data so that multiple demands are delivered to any number of clients - without need of subsequent data queries. These important factors greatly enhance processing speed.

However, in some small scale implementations the direct data connection may be the best choice. For instance, in order to rapidly implement Enterprise RTView for prototyping and testing purposes. When testing is complete, it can be ported to a served data deployment for the production environment.

Ultimately, this decision would be determined by weighing the advantages that the Data Server brings; performance, scalability, security, easier setup, and lower impact on backend data sources, against the cost advantage of the direct data connection.

NOTE: Refer to the Data Sources section of this documentation to see if deployment with a Direct Data Connection is supported by your data source.

The pros and cons of the two scenarios, Served Data and Direct Data Connection, are described below.

Pros and Cons

Issue	Served Data	Direct Data Connection
Setup	Server:  Requires Data Server setup Client:  Requires less setup since clients do not need to meet all system requirements for each data source	Server:  Does not require Data Server Client:  Requires more setup since each client must meet all system requirements for each data source
Performance	For a very small deployment it is slower, but for all other deployments it is faster	For a very small deployment it is faster, but for all other deployments it is slower
Security	More secure since backend data applications are only accessed by the Data Server	Less secure since backend data applications are directly accessed by clients
Scalability	More scalable due to data requests being federated and caching by the Data Server	Less scalable since each client makes individual data request
Cost	Large Deployment: Client maintenance more costly, hardware less costly Small Deployment: Hardware cost equal to Direct Data Connection but client maintenance more costly	Large Deployment: Client maintenance less costly, hardware more costly Small Deployment: Hardware cost equal to Served Data but client maintenance less costly

Application with Served Data
The Application with Served Data deployment involves providing access to the Display Viewer Application for all clients which need to view displays.  This could be done by either doing individual installs on each client, sharing an install with a central platform, or via Java Web Start. The Data Server is installed on a server to provide access to all defined data sources and deliver the federated and cached data via XML to the appropriate Display Viewer Applications (see Figure 6). When you deploy Enterprise RTView as an Application with Served Data, depending on the command, commands are executed either on the server or the client. See command descriptions for information on where commands are executed.
  Figure 6: Application with Served Data Deployment Overview

Choose Served Data Deployment

Application with Direct Data Connection
The Application with Direct Data Connection deployment involves providing access to the Display Viewer Application for all clients which need to view displays. This could be done by either doing individual installs on each client, sharing an install with a central platform, or via Java Web Start.  Each individual client must be configured so that it can have direct access to data sources. This may involve the installation of additional software such as database drivers or middleware components depending on the data source types used (see Figure 7).
  Figure 7: Application with Direct Data Connection Deployment Overview

Choose Direct Data Connection Deployment