Editorial / Best Answer
nishant2200
1. Create a connection. This requires a connection string, which can be given declaratively or put in a well defined place like the .config files. Advantage of keeping in .config files is that it enables use of Connection Pooling by .Net framework, else even one small change in connection string will cause CLR to think it's not the same connection and will instantiate new connection for other request.
2. Open the connection and keep it open until done, typically done as using (con) { //use }
3. If using connected data model, create a SqlCommand object, decorate it with desired command, command type (stored procedure for eg), add any parameters and their values to the command, and then consume the command by using ExcuteReader or ExecuteScalar. In case of ExecuteReader, we will get back a handle to a fast-forward, read only pointer to the recordset. We can also decorate Command object with multiple recordsets in 2.0 and execute one by one (MARS - Multiple Active Record Sets)
4. If using disconnected data model, create a DataAdapter object, decorate it with desired SELECT, INSERT, UPDATE, DELETE commands, add parameters as necessary and then fill up a DataSet or DataTable using the DataAdapter. Subsequent SQL can be executed using insert, update, delete commands on the dataset.
What are the steps to connect to a database?
Editorial / Best Answer
nishant2200Profile Answers by nishant2200 Questions by nishant2200
1. Create a connection. This requires a connection string, which can be given declaratively or put in a well defined place like the .config files. Advantage of keeping in .config files is that it enables use of Connection Pooling by .Net framework, else even one small change in connection string will cause CLR to think it's not the same connection and will instantiate new connection for other request.
2. Open the connection and keep it open until done, typically done as using (con) { //use }
3. If using connected data model, create a SqlCommand object, decorate it with desired command, command type (stored procedure for eg), add any parameters and their values to the command, and then consume the command by using ExcuteReader or ExecuteScalar. In case of ExecuteReader, we will get back a handle to a fast-forward, read only pointer to the recordset. We can also decorate Command object with multiple recordsets in 2.0 and execute one by one (MARS - Multiple Active Record Sets)
4. If using disconnected data model, create a DataAdapter object, decorate it with desired SELECT, INSERT, UPDATE, DELETE commands, add parameters as necessary and then fill up a DataSet or DataTable using the DataAdapter. Subsequent SQL can be executed using insert, update, delete commands on the dataset.
Related Answered Questions
Related Open Questions