Vous êtes sur la page 1sur 7
“OBIEE 11g Working with PARTITIONS” Description Change Author
“OBIEE 11g Working with PARTITIONS” Description Change Author
“OBIEE 11g Working with PARTITIONS” Description Change Author

“OBIEE 11g Working with PARTITIONS”

Description Change Author
Description Change
Author

This is another document from OBIEE11g Beginner’s Guide Series. This document briefs you the step by step approach to understand Partitioning for OBIEE to speed up query performance. Learn from expert

0.1 1 st Review 0.1 Initial Draft History: Version
0.1 1 st Review
0.1 Initial Draft
History:
Version

Publish Date 15 th Sep 2011 15 th Sep 2011

Hitesh Mankar

Amit Sharma

Partitions
Partitions

OBIEE 11g Beginners Guide

|

|

learnhyperion.wordpress.com

1

A

partition

is

a division of a logical

or its constituting elements into distinct independent

parts. Database partitioning is mostly used to manage,

increase

and also speeds up

performance of OBI server. Using partitioning we splits big tables into small tables and more manageable form to get high performance from database and also from OBIEE.

2. Level Based Partitioning: - If the same facts (measures) are stored in different or separate
2. Level Based Partitioning: - If the same facts (measures) are stored in different or separate
table’s at different levels of aggregation. The data inside level based partitioned table is always calculated
to a specific level of aggregation.
indexes
tables
If the metadata is built using multiple sources OBI Server handles all the navigation to appropriate
source. So OBI decides on its own to access which source for faster and correct result to meet the rusers
request.
For a single logical table source in a Business Model data is often partitioned into multiple physical
sources. And when a logical table source does not contain the entire set of data at a given level, need to
specify the partition of the set that it contains.
Different types of partitioning used to increase performance of OBI server.
1. Fact Based Partitioning: - if partitioning is done like data related to fact is stored in different tables.
For example Yearly Amount sales are stored in different table other then Fact_sales table.
Below is the look of detail and partitioned fact.
Partitioning for OBIEE:-
Partitioning enables user to store one logical object a
can take
transparently in several independent
table
OBIEE 11g Beginners Guide
|
http://learnoraclebi.wordpress.com
|
learnhyperion.wordpress.com
2
Partitioning can provide
because
partitioning. The partitioning feature Database enables to partition stored
as
physical
and
segments.
for easier management and improved performance.
elimination (pruning) capabilities, but also because
data
performance improvements
parallel
execution plans
great
segments
such
advantage of
of partition

This technique allows to mix two

with two different grains (the level based partitioning).

Below is the look of Level Based Partitioned table.

This technique allows to mix two <a href=facts table with two different grains (the level based partitioning). Below is the look of Level Based Partitioned table. 3. Value Based Partitioning: - According to the values of the data partitioning is possible, can partition data into separate tables. Depending of a value of column, we can split the query against a table or another. Value based partitioning can create complexity within query processing. Creating partitions value based partitions number of tables will increase, but help to get faster result. Below is the look of value based partitioned tables. Multiple value based partitioned tables. Follow below steps to create request for partitioned tables. 1. Import metadata into physical layer. OBIEE 11g Beginners Guide | http://learnoraclebi.wordpress.com | learnhyperion.wordpress.com 3 " id="pdf-obj-2-12" src="pdf-obj-2-12.jpg">
This technique allows to mix two <a href=facts table with two different grains (the level based partitioning). Below is the look of Level Based Partitioned table. 3. Value Based Partitioning: - According to the values of the data partitioning is possible, can partition data into separate tables. Depending of a value of column, we can split the query against a table or another. Value based partitioning can create complexity within query processing. Creating partitions value based partitions number of tables will increase, but help to get faster result. Below is the look of value based partitioned tables. Multiple value based partitioned tables. Follow below steps to create request for partitioned tables. 1. Import metadata into physical layer. OBIEE 11g Beginners Guide | http://learnoraclebi.wordpress.com | learnhyperion.wordpress.com 3 " id="pdf-obj-2-14" src="pdf-obj-2-14.jpg">

3. Value Based Partitioning: - According to the values of the data partitioning is possible, can partition data into separate tables. Depending of a value of column, we can split the query against a table or another.

Value based partitioning can create complexity within query processing. Creating partitions value based partitions number of tables will increase, but help to get faster result.

Value based partitioning can create complexity within query processing. Creating partitions value based partitions number of

Below is the look of value based partitioned tables.

Value based partitioning can create complexity within query processing. Creating partitions value based partitions number of
This technique allows to mix two <a href=facts table with two different grains (the level based partitioning). Below is the look of Level Based Partitioned table. 3. Value Based Partitioning: - According to the values of the data partitioning is possible, can partition data into separate tables. Depending of a value of column, we can split the query against a table or another. Value based partitioning can create complexity within query processing. Creating partitions value based partitions number of tables will increase, but help to get faster result. Below is the look of value based partitioned tables. Multiple value based partitioned tables. Follow below steps to create request for partitioned tables. 1. Import metadata into physical layer. OBIEE 11g Beginners Guide | http://learnoraclebi.wordpress.com | learnhyperion.wordpress.com 3 " id="pdf-obj-2-30" src="pdf-obj-2-30.jpg">
This technique allows to mix two <a href=facts table with two different grains (the level based partitioning). Below is the look of Level Based Partitioned table. 3. Value Based Partitioning: - According to the values of the data partitioning is possible, can partition data into separate tables. Depending of a value of column, we can split the query against a table or another. Value based partitioning can create complexity within query processing. Creating partitions value based partitions number of tables will increase, but help to get faster result. Below is the look of value based partitioned tables. Multiple value based partitioned tables. Follow below steps to create request for partitioned tables. 1. Import metadata into physical layer. OBIEE 11g Beginners Guide | http://learnoraclebi.wordpress.com | learnhyperion.wordpress.com 3 " id="pdf-obj-2-33" src="pdf-obj-2-33.jpg">
Multiple value based partitioned tables.
Multiple value based partitioned tables.

Follow below steps to create request for partitioned tables.

1. Import metadata into physical layer.

OBIEE 11g Beginners Guide

|

|

learnhyperion.wordpress.com

3

2. Create Physical Joins.

2.

Create Physical Joins.

2. Create Physical Joins.
2. Create Physical Joins. Add logical table sources in Business Model and Mapping Layer, as perhttp://learnoraclebi.wordpress.com | learnhyperion.wordpress.com 4 " id="pdf-obj-3-15" src="pdf-obj-3-15.jpg">
2. Create Physical Joins. Add logical table sources in Business Model and Mapping Layer, as perhttp://learnoraclebi.wordpress.com | learnhyperion.wordpress.com 4 " id="pdf-obj-3-17" src="pdf-obj-3-17.jpg">
Add logical table sources in Business Model and Mapping Layer, as per requirement.

Add logical table sources in Business Model and Mapping Layer, as per requirement.

3.
3.

OBIEE 11g Beginners Guide

|

|

learnhyperion.wordpress.com

4

4.

4.

4. Check Business Model is proper or not. 5. 6. 7. Drag and drop business modelhttp://learnoraclebi.wordpress.com | learnhyperion.wordpress.com 5 " id="pdf-obj-4-9" src="pdf-obj-4-9.jpg">

Check Business Model is proper or not.

5. 6. 7.

5.

6.

7.

4. Check Business Model is proper or not. 5. 6. 7. Drag and drop business modelhttp://learnoraclebi.wordpress.com | learnhyperion.wordpress.com 5 " id="pdf-obj-4-23" src="pdf-obj-4-23.jpg">

Drag and drop business model into presentation layerCheck global consistency and save the rpd.

Configure rpd and start services to deploy into answers Login into answers.

Create a request

Configure rpd and start services to deploy into answers  Login into answers. Create a request

Instance 1).Select columns to make a request from criteria tab into Answers to get result from FACT based

4. Check Business Model is proper or not. 5. 6. 7. Drag and drop business modelhttp://learnoraclebi.wordpress.com | learnhyperion.wordpress.com 5 " id="pdf-obj-4-41" src="pdf-obj-4-41.jpg">
partitioned tables . 8.
partitioned tables .
8.

Click on Result tab to check result, check is it desired result or not.

(Result is correct now, check the query log for query)

4. Check Business Model is proper or not. 5. 6. 7. Drag and drop business modelhttp://learnoraclebi.wordpress.com | learnhyperion.wordpress.com 5 " id="pdf-obj-4-50" src="pdf-obj-4-50.jpg">

OBIEE 11g Beginners Guide

|

|

learnhyperion.wordpress.com

5

Instance 2) A filtered request for Value Based Partition.

Check Result is desired or not. Check Query Log.

Check Result is desired or not.

Check Result is desired or not. Check Query Log.

Check Query Log.

Check Result is desired or not. Check Query Log. Instance 3) A filtered request from get
Check Result is desired or not. Check Query Log. Instance 3) A filtered request from get
Check Result is desired or not. Check Query Log. Instance 3) A filtered request from get
Check Result is desired or not. Check Query Log. Instance 3) A filtered request from get

Instance 3) A filtered request from get result from Level Based Partitioned tables.

Check Result is desired or not. Check Query Log. Instance 3) A filtered request from get
Instance 2) A filtered request for Value Based Partition. Check Result is desired or not. Checkhttp://learnoraclebi.wordpress.com | learnhyperion.wordpress.com 6 " id="pdf-obj-5-31" src="pdf-obj-5-31.jpg">

OBIEE 11g Beginners Guide

|

|

learnhyperion.wordpress.com

6

Check Result is desired or not.

Check Query Log.

Check Query Log.

Check Query Log.
Check Query Log.
Check Query Log.
Check Query Log.
Check Result is desired or not. Check Query Log. OBIEE 11g Beginners Guide | <a href=http://learnoraclebi.wordpress.com | learnhyperion.wordpress.com 7 " id="pdf-obj-6-20" src="pdf-obj-6-20.jpg">

OBIEE 11g Beginners Guide

|

|

learnhyperion.wordpress.com

7