On which column will you take primary index???
Answers were Sorted based on User's Feedback
Answer / yuvaevergreen
Generally, PI on an unique column would be recommended.
This is because, the unique primary index helps TD to
distribute the row evenly among amps.
| Is This Answer Correct ? | 29 Yes | 0 No |
Answer / saidulu
In PI(Primary Index) also we have 2 types ie (unique
PI ,Non unique PI)
how can we say unique column as primarIndex.If it is
non unique PI means it allows duplicates.
Now i am comming to which column as Primary Index
1.Frequntly using column.
2.Not regularly changing column
(Not at all changing values /rarely)
3.By default it takes as first column as PI
| Is This Answer Correct ? | 7 Yes | 1 No |
Answer / sailendra
Actually we have two types of primary index in teradata
1.Unique Primary Index and
2.Non unique primary index
unique primary index are generated on the unique columns
and by default it will take the first column if you have
not specified it as unique primary index.
In non unique primary index we can have two rows with same
values.
But the main purpose of selecting a column as primary index
is the uniqueness of the column and the frquency of the
column used in the queries and in the join conditions.
| Is This Answer Correct ? | 4 Yes | 0 No |
Answer / malathi
Primay key is a constraint used on a table and used to
impose the rule of uniqueness for the application.
Whereas primary index is the way tera data internally
distributes rwos in a table. Primary index can be created
on any column which may mot be a primary key. In teradata
each table should have a primary index (be default first
column in the table is primary indexed).
| Is This Answer Correct ? | 3 Yes | 6 No |
what is identity columns in TD?
Can we load a Multi set table using MLOAD?
How many tables can you join in v2r5?
What is meant by a Highest Cost Plan?
In your project what type of issues and errors you facing in Teradata
Is multi insert ansi standard?
write lock is compatiable with which type of lock?
How do you eliminate duplicates?
How to cast date "2015-03-03-012204-000000" to timestamp
What are the updated features of teradata?
How to run the Teradata jobs in unix environment ???
What is a node in teradata? Explain