Answer / sat!sh

The data distribution of table among AMPs is called Skew Factor

Generally For Non-Unique PI we get duplicate values so the
more duplicate vales we get more the data have same rowhash
so all the same data will come to same amp, it makes data
distribution inequality,

One amp will store more data and other amp stores less
amount of data, when we are accessing full table, The amp
which is having more data will take longer time and makes
other amps waiting which leads processing wastage

In this situation (unequal distribution of data)we get Skew
Factor High

For this type of tables we should avoid full table scans

ex:
AMP0 AMP1
10000(10%) 9000000(90%)

in this situation skew factor is very high 90%

Answer / sricharan

It is a number.It tells you how the data is distributed
among Processors or Amps.Skew factor varies about 0-100.
If the data is distributed evenly among the processor the
skew factor is ZERO whether it is smaller or large table.
Skew factor doesn't depends on size of table but it only
depends on distribution of data.

If the skewfactor is more we have to access the table on
only primary index columns but not whole table.

Answer / yuvaevergreen

Skew Factor is the indication of how evenly the data is
spread across the AMPS.

A skew factor of 0 indicates that the data is perfectly
distributed across all the AMPS.

Answer / navaneeth reddy

Skew factor is distribution of rows of a table among the
available no.of AMP's.
If your table has a chance of using unique primary index,it
is always better to use UPI which ensures the skew factor
around 0%.
If there is no chance of having unique values column in a
table choose a column as PI(primary index) which has less
duplicate values which inturn results in less skew factor.
That is the data will be distributed almost(not exactly
equal percentage) equally to all AMP's.

Answer / yuvaevergreen

Below query can be used to find the distribution by amps.
SELECT HASHAMP(HASHBUCKET(HASHROW(index or column)))
,COUNT(*)
FROM TABLENAME GROUP BY 1 ORDER BY 2 DESC;

Answer / srinu

The Table is having too many Duplicate rows, The Skew table
will hadle the duplicate rows up to limitations,Once its
cross the no of Rows it will effect on Performence.The Dba
people will the SKEW tables.

I think this Answer helps u

what is differences between Fastload and Multiload as per DBA aspect ?

1 Answers   Wipro,

1. generally how many node system will be used by the client. 2. while running mload if the job is aborted/failed at any phase what we have to look after to run the job successfully? 3.i wanted to know how to write the fastload script for table creation as well as table loading in a single script?

3 Answers  

How to improve the query performance without taking the collect statistics? In explain plan i have 'NO CONFIDENT'.

2 Answers  

one table have input no name 10 rao 20 siva 30 srinu 10 rao i want to ouput like this way no name 20 siva 30 srinu 10 rao how it posible in only sql server query?not oracle?

3 Answers   IBM,

Pls any body reply for this question. Which circumstance does the optimizing choose a product join? Thanks for advance

6 Answers   Cap Gemini,

What is a sparse index?

0 Answers  

What is the command in bteq to check for session settings ?

0 Answers  

List the logical and conditional operators that are used with teradata along with their meanings?

0 Answers  

suppose one Cylinder split data blocks are moved from where?

1 Answers  

How would you load a very large file in teradata in general? What utility would you use for it? Why?

0 Answers  

What are the different methods ot loading a dimension table? A fact table etc?

0 Answers  

In general, how do you optimze any sql in teradata?

0 Answers