Normalization: is a process of organizing data and minimizing redundancy
De-normalization: is a technique to move from higher to lower normal forms of database modeling in order to speed up database access.
Stored Procedure: is a named group of T-SQL statements which can be created and stored in Database as an object.
Primary Key: is a unique identifier of a row in a DB table, [it can’t be NULL]
Unique key: forces uniqueness to a respective table column, [it can be NULL]
Foreign Key: a foreign key in 1 table refers to the primary key in other table, Used to force referential integrity.
Inner join: exists in both tables
Left Outer join: all records from left side table + matched rows from right side table (totals number of rows will be same as left table)
Right Outer join: all records from right side table + matched rows from left side table (totals number of rows will be same as right table), it’s a mirror image of left outer join
Full Outer join: all records from left side table + all records from right side table, weather matched or not
Cross join: returns [left table rows * right table rows], a Cartesian product of both tables
Self join: when table joins to itself using diff aliases to avoid confusion
Union: selects only distinct records from both tables
Union all: selects all records from both tables
View: is a subset of a table, can be used to retrieve data, insert or Update data. Can contain multiple select statements inside
Trigger: A trigger is a SQL procedure that initiates an action when an event (INSERT, DELETE or UPDATE) occurs.
Cursor: is a database object used to loop trough records on row by row bases.
Index: pointers to data records, represents structure of how data get stored physically in a table
||Non clustered index
|Reorders physical data stored in table
||It contains pointers to data rows
|A table can have Only 1 clustered index
||A table can have one OR many non-clustered index
|Table is having By default a clustered index
|Leaf nodes contains data
||Leaf nodes contains reference to data
Linked server: is a concept of adding other remote server to a group to query DB’s of both servers together
Collation: set of rules that determines how data stores & compares in database
Collation types: case sensitive, accent sensitive, kana sensitive, width sensitive
Data ware housing:
- Record should Never delete from DB
- All records must be linked
- Once committed records should be read-only
- All changes made must be tracked with time
User defined function (UDF): is a bunch of T-SQL statements which accepts 0 or more parameters and returns a scalar data value or table.
DDL: data definition language – e.g. TRUNCATE command is a DDL command
DML: data manipulation language – e.g. INSERT, UPDATE & DELETE are DML commands
Read Full Post »