cancel
Showing results for 
Search instead for 
Did you mean: 

Physical Partitioing Calender Year

Former Member
0 Kudos
45

HI Gurus.

Cube is physically partitioned on 0Calmonth.

Report is running on Calender Year Example - 2007 and 0calmonth2 Example - 005 not on 0Calmonth Example - 2007005.

Does phy part still improve performance.

Thankyou Gurus

Message was edited by:

George M

Accepted Solutions (1)

Accepted Solutions (1)

Former Member
0 Kudos

Table partitions are physically separated tables, but logically they are linked to one table name. PSA tables and non-compressed F-fact table are partitioned by the system (by request ID). The (compressed) E-fact table can be partitioned by the user by certain time characteristics. For range-partitioned InfoCubes, the SID of the chosen time characteristic is added to both fact tables.

When using range partitioning, query response time is generally improved by partition pruning on the E fact table: all irrelevant partitions are discarded and the data volume to be read is reduced by the time restriction of the query. Note that all F fact table partitions have to be read.

The number of (range) partitions is one more than defined. This additional partition keeps all time characteristics higher than the upper limit. Example: monthly partitions 01/2007 – 12/2007. This creates 12+1 partitions. The first one contains records from 01/2007 and earlier, then there are 11 partitions containing the months 02 – 12/2007. An additional partition for all records after 12/2004 is created.

In ORACLE, report SAP_DROP_EMPTY_FPARTITIONS can help you to remove unused or empty partitions of InfoCube or aggregate fact tables. Unused or empty partitions can emerge in case of selective deletion or aborted compression and may effect query performance as all F fact table partitions are accessed for queries on the InfoCube.

Hope it Helps

Chetan

@CP..

Answers (2)

Answers (2)

Former Member
0 Kudos

Hi,

As your cube is partitioned for the year 2007 and query is running for calmonth 02 it will directly retrive data from that partition only so it will improve the performance. when you partitioned your cube based on 0calmonth system will partition according to that like 12 + 2 partitions, when ever data comes it will be stored in the respective partitions and will be retrived from them at the time of query execution.

Regards

Sankar

Former Member
0 Kudos

I think it will not improve performanc as it is partitioned for 2005 and not for other years. If the partition is done on calmonth for all years teh it would improve performance i believe.

Former Member
0 Kudos

venkat,

I corrected my question.

Former Member
0 Kudos

It will do then.