on 05-23-2016 12:23 PM
Hello Everyone,
We have a scenario in normal SAP BI which we want to implement in HANA and see if we achieve any improvement .
Scenario:
Currently we have a source S1 which gives sales against profit center e.g.
company code | Profit center | sales | month |
C1 | P1 | 1000 | 4 |
We have another table T1 where these same values are split along multiple profit centers and ratio are stored e.g.
Profit center | Company code | sales | month | split ratio |
A1 | C1 | 600 | 4 | 3 |
A2 | C1 | 400 | 4 | 2 |
The report shows the values as . The values in the source S1 is split on profit center line based on T1.
company code | Profit center | sales | month | split ratio |
C1 | A1 | 600 | 4 | 3 |
C1 | A2 | 400 | 4 | 2 |
Current solution which we have: We refresh all the records from S1 into the target multiple times a day which consumes lot of time and resources.
The amount of data in S1 is huge and historical records needs to be adjusted everytime any ratio changes in T1.It takes hours to load the data currently in BI
Requirement:
We need to replicate the same scenario in Hana based on projections etc. and understand how fast these refresh will be and how can we implement it.
Thank you for your suggestions/Solutions.
Please reply back if the question is not clear.
Hi Ajeet,
There are many options or ways to improve your BI landscape by HANA.
1. ERP on anydb -> BW on HANA: your query results will improve 20-100 times upon scenario. However data extraction will only improve apprx 20% if you have standard extraction, but DSO activation will improve dramatically, thanks to coloumn store and HANA optimized re-design of fact and dim tables.Also you can replicate data table to table from ERP to BW with no latency by SLT
2. ERP on HANA -> BW on HANA : Addition to all improvements in 1st option you have chance get real time data without replication, by using analitic and calculation views. Simply creating a remote cube based on HANA views.
Just make a quick search in googe "BW on HANA" or goto
Cheers,
sarhan
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
what is the link between S1 and T1
how do you know A1 and A2 belong to P1?
P1 is profit center group and A1/A2 are profit center. So ideally, T1 should have entries like below
PG PC Ratio Month Sales
P1 A1 3 4 600
P1 A2 2 4 400
Please confirm that.
Regards,
Sandeep
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
t1
cc pg sales month
c1 p1 1000 4
c1 p2 500 4
t2
pc pg sr
a1 p1 6
a2 p1 4
q1_t1 = ce_column_table("t1",["cc","pg","sales","month"]);
q2_t2 = ce_column_table("t2",["pc","pg","sr"]);
q3 = CE_LEFT_OUTER_JOIN(:q1_t1,:q_t2,["pg"]);
t3
pc pg sr cc sales month
a1 p1 6 c1 1000 4
a2 p1 4 c1 1000 4
p2 c1 500 4
var_out = ce_projection(:q3,["pc","pg","sr","sales",CE_CALC('if(isnull("pc","sales","sales"*"sr"/100,decimal(5,2)))') as "new_Sales","month"]);
var_out
pc pg sr cc sales new_sales month
a1 p1 6 c1 1000 600 4
a2 p1 4 c1 1000 400 4
p2 c1 500 500 4
Create CV with sql
Begin
q1_t1 = ce_column_table("t1",["cc","pg","sales","month"]);
q2_t2 = ce_column_table("t2",["pc","pg","sr"]);
q3 = CE_LEFT_OUTER_JOIN(:q1_t1,:q_t2,["pg"]);
var_out = ce_projection(:q3,["pc","pg","sr","sales",CE_CALC('if(isnull("pc","sales","sales"*"sr"/100,decimal(5,2)))') as "new_Sales","month"]);
end;
Regards,
Sandeep
Hello Ajeet,
Greeting of the day,
Good question, my suggestion would be following:-
1) Get the HANA view which shows the Sales per profit center (should be standard) and then copy this and create a calcualtion view with your ratios
2) Or you can use the HANA live content and to get the ratio, use another table and put the processing logic in the reporting tool of BO.
3)In case you are on S/4 system, the source S1 should be updated real time (need to have embedded BW) and then use this.
Hope this helps, personally i like the option 1,
Check Sales Analysis Fiori app and the back end view it is using
You should get your answer
Regards
Rishab
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
User | Count |
---|---|
101 | |
12 | |
11 | |
6 | |
6 | |
4 | |
3 | |
3 | |
3 | |
3 |
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.