What are the issues related to the deployment of data access application?

Some of the basic issues related to the deployment of the data access applications are:

• Can we suppose the required database exists at the target location?
• If the database does exist then how can we get the connection string of it? Should we get it at the installation time? Or at the first run?
• If the database does not exist then how it can be created? Is it the responsibility of human installing our application? Or the application should create it at the time of installation?
• If the database is to be created by the application installation, then how does the installation setup know where (on which machine) to create the database? And what user name and password the setup should use to create the database?
• Once, the database is created or its connection string is found at the setup time, then how to store the connection string for later use (regular execution)?
• What if the database address (connection string) is changed during the application life cycle then what should be the application behavior? How can the application be aware of the new connection string?
• If the database schema is hard-coded in the code (like table and field names) and the DB schema is changed (table/field name or data type is changed or new field/table added or some fields/tables deleted) because of any reason, then how to fix this problem?

You can see from the above points that the basic issue is the identification of the address (or connection string) of the database server. The most important point to remember here is that the application must perform in any condition and must not be crashed because of any condition and most of the problems (if not all) should be handled without any change in the code or any update patch installation

Back