Table Partitioning in HANA

[Former Member]

Hi

We have a big table that we are loading to Hana every month during system refresh. It is taking lot of time to load as the table is very big and its not partitioned. Wanted to know few basic things before I start working on the Table Partitioning.

1. How do we check which Partition Specification is suitable for my Table ( Hash, Round Robin or  Range ). Do I need to check with Developer as which option is correct.

2. Do I need to create the table in HANA before I do the replication with Partitions. If yes, can I copy the same table structure from our ECC system and create the table.

3.  If I use Hash, I need to specify columns as partitioning columns. How do we find which columns we need to specify. 

I have already gone through the SAP documents and your personnel responses would be appreciated.

Mahesh Shetty

[Former Member]

Partition based on your development requirements. A simple way to go ahead is using range partition based on date, this would suit simple scenarios. You should partition the table before loading data into it.

you can refer to the following link for some details.

http://scn.sap.com/community/hana-in-memory/blog/2013/08/04/various-table-partitioning-options-in-sa...

[SK-EA]

Mahesh,

Can you provide details on the usage of SAP HANA in your case? Is it a Datamart or BW on HANA scenario?

For later scenario, please look at 2019973 - BW on HANA - Handling BIG data tables OSS note once.

Srinivas K

[Former Member]

Srinivas

We are using HANA to replicate some of the Large FICO ECC Tables. We are not currently using BW on HANA.

Mahesh Shetty