Why CTL file is not separate for loading data into Columnar Tables?

[former_member182302]

Hi Guys,

We have 2 kinds of tables i.e Row and Column supported by HANA.

And on Memory side the way the tables are stored is different.

If i see CTL file, i specify the row-wise delimiter and hence i am assuming that is also one of the reasons for faster loading into "Row Tables"

Want to confirm if my assumption is correct? if yes will in future will we have some option to specify whether it is a row or column table in CTL files?

Request your inputs on this.

Regards,

Krishna Tangudu

[Former Member · Accepted solution]

Hi Krishna,

This is not the direct reason for the column-store tables to be slower for data loading.

Inserting data into an 'n' column tables is like inserting data into 'n' 1 column tables and then maintaining a relationship for that record. Hence it is generally slower.

With regard to the CTL file, it will always be a row wise delimiter that will be considered because as a at the core, one needs to identify one record singularly and complex operations like aggregation comes later.

--

Shreepad Patil