Skip to Content
avatar image
Former Member

Impacts on Infocube due to data load every 10 minutes

Hello experts,

We have a user requirement to load transaction data every 10 minutes during day. Till sometime back, we were using a Day cubes to load the delta requests during day and load the entire data to Night cubes during night and empty day cube. Reasons for this approach were:

1. Time taken to delete and create index for the Night cubes.

2. Less housekeeping activity as the number of requests loaded to the night cubes were less. So the system table size was intact.

We recently migrated our system to HANA and hence we dont require any index creation/deletion. Now we are planning to remove the Day cubes and load data to a Single cube during day and night.

Could you please help me in understanding the impacts caused by this approach? Few impacts that could be thought were:

1. Increase in the system tables due to the increased number of requests loaded.

2. There will be increase in number of data because of the Changed records coming from DSO.

Could you please give any other impacts + some precautionary measures to be implemented in the system?

Regards,

Madhu,

Add comment
10|10000 characters needed characters exceeded

  • Get RSS Feed

1 Answer

  • Jan 17, 2015 at 12:40 PM

    Hi Madhu,

    With a large number of requests it is advisable to run compression frequently. With a HANA optimized InfoCube there are several partitions: partition 1 with uncompressed data (Request > 0) and partition 2 with compressed requests (Request = 0). This can have a positive impact on the loading performance (i.e. the merge time).

    This information I found in document Aspects of the HANA-optimized InfoCube.

    Best regards,

    Sander

    Add comment
    10|10000 characters needed characters exceeded