what is identity column in TD?
Answers were Sorted based on User's Feedback
Answer / surendra reddy duggireddy
Identity Columns generate a system-generated number for
every row as it is inserted
in the table. It is also knows as a DBS Generated Unique
Primary Index. Identity
Columns are used if you want to guarantee row uniqueness in
a table and they can
simplify the initial port from databases other than
Teradata. They are often used as the
Primary Index so they can guarantee even distribution. In
some cases they are utilized to
optimize and simplify the first port from another database
that uses generated keys.
Identity columns do NOT need to be part of the Primary
Index, but they are often used
for that purpose because they provide even data
distribution. The biggest rule is that
you can only specify one Identity Column per table
Example:
-------
Create Table Employee_Table
( Employee_No INTEGER
GENERATED ALWAYS AS IDENTITY
(START WITH 1 INCREMENT BY 5
MAXVALUE 1000000
NO CYCLE)
, Dept_No INTEGER
,First_Name CHAR(20)
,Last_Name CHAR(20)
,Salary DECIMAL (10,2)
) PRIMARY INDEX (Employee_No);
| Is This Answer Correct ? | 8 Yes | 0 No |
Answer / yuvaevergreen
Identity columns are system generated values. Generally
these columns are used for unique values. Some restrictions are
1. Only 1 column can be specified as identity column.
2. PPIs, Join indexes, hash indexes cannot have identity
columns as indexes.
3."Default" and "Always with cycle" option in identity
columns may result in duplicates.
| Is This Answer Correct ? | 3 Yes | 0 No |
Hi All, can anyone tell me how many types of Skew exists? whats the difference between amp skew,data skew,cpu skew, io skew, delta amp cpu ?
Why MLOAD needs Work Tables?
How many types of index are present in teradata?
What are the joins in teradata?
How can we build data marts from FSLDM. is there any standard approach for this?Is FSLDM supports place ?
why use references rather than pointers in the public api, particularly for arguments which are modified?
How many tables can you join in v2r5?
What is the difference between SI and JI?
We are migrating an oracle table into teradata, the volume of data is huge and partitioned (year wise list partition). How do i simulate the same in teradata.
How do you create a table with an existing structure of another table with data and with no data?
What tools would you use for that?
Difference between Teradata V2R5 and Teradata V12 versions?