Analytical Views Vs joins in IDT

[Former Member]

Hello ,

I was wondering on what are the diffrences between creating analytical views ( these would do all joins in Hana studio) & doing the joins in IDT ( Bo Tool ) which one is recommended

Thanks,

Kiran

[Vitaliy-R · Accepted solution]

If you are looking for the performance - you should be doing Analytic (and Calculation) Views in HANA Studio, which in turn are executed by HANA natively. Regards. -Vitaliy

[Former Member]

There are some striking differences between creating views in HANA Vs IDT:

1. Views created in HANA are columnar based views. This leads to fast retrival of data( only required attributes and measures are accessed by reporting tool) On the other hand if we used IDT to create views, during reporting whole table will be treated as row-based table (the whole table will be read based on Filters and brought into cache, it is only during presentation the unrequired attribute will not be accessed)

2. IDT can join data from multiple sources at universe layer unlike HANA (which has to get all Data in HANA before creation of any kind of views).

3. Calculation Performance : Calculation performance on HANA will be strikingly fast as compared to IDT due to Hardware and architecture of HANA

4. In my experience, HANA is only used for high-volume of data. Depending on type of analysis user may want to perfom, there can be considerable latency added to overall time of reporting.

[Former Member]

Hi,

Analytical & Calculation views should be done in HANA for performance reasons. However, if you will use filters I recommend them to be done in the front end to be more visible to end users/developers, unless everything is well documented

Cheers

Tansu

[Norman_May]

Hallo,

filters on columns views will be pushed into the column store engine. Thanks to early filtering of data this leads to better performance than predicate processing int the client will do. So if possible, filters should also be processed on top of the HANA views.

Regards, Norman