SAP BW on HANA Advanced DSO – Part 1
SAP BW on HANA 7.4 SP10 new changes has happened, will create Advanced DSO, composite provider and Open ODS view will create only in BW molding tools not in SAP GUI
Look forward will use only 3 objects BW data models
1. Advanced DSO – Persistent layer
2. Open ODS view – Virtual layer
3. Composite Provider- Joins/unions
This blog will describe Advanced DSO.
Advanced DSO have persistent object is combination of below objects
1. Field based structure of PSA ( filed based modeling)
2. No activation required for like WDSO (update property)
3. 3 tables have (active, new, change log table ) like standard DSO (over write property)
4. Info cube model – select setting – all char are keys ( addition property)
Below is the changes are happened in object level
System version BW 7.40 SP10
Advanced DSO tables and purpose to use
Below tables are generated while activating the Advanced DSO (ADSO Name – ZODS_PAN1)
In this documents will discuss about, I can say Advanced DSO first model.
While creating the ADSO will select the active/compress data with check box write change log data.
If you select above options it will act as the standard DSO and it will generate 3 tables + 2 tables.
Modeling ADSO in Eclipse using BW modeling tools
Go to HANA STUDIO -> BW Modeling tools
Select the info area -> right click -> select Advanced DSO object
Below screen will appear
Ex ADSO Name – ZODS_PAN1
Create form templates
- None – will not select any object
- Data Source – if you select the data source need to give the source system name and data source name -> will create ADSO using the data source but data source has to create with respective source system.
- Info provider -> can use the existing info providers.
Click on finish then below screen will appear
Modeling properties
Active/Compress data
Write change log table -> if you select data will insert into table
Keep inbound data, extract form inbound table -> if you select will keep the data in bound table and available for reporting
Unique data records – if you select duplicate records not allowed.
All char are keys, reporting on union of inbound and active tables – if you select this option ADSO will act as CUBE.
Extended table – The extended table’s concept relates to warm data and optimize the main memory resource management in SAP HANA by using extended tables
I will discuss about first model – write change log table with activate/compress data
Go to details tab -> below screen will appear
Add the info objects and key figures
Then select the keys for ADSO according to Keys data will transfer.
Selected customer as key fields in ADSO will use 120 key fields as maximum
Then create the data source in BW system and map the transformation form data source to ADSO
Aggregation type – overwrite and summation
ADSO Manage screen
Request generated as timestamp nothing but RTSN (Request Transactional Serial Number) while RTS Number we can find easily when it was loaded.
Another advantage for request generation in standard DSO while loading it will generate the BI_REQ but ADSO have Request TSN – Request transactional serial number.
When you activate a transformation, the system checks whether it can be executed in SAP HANA.
Whether this will be possible basically depends on the operations and objects used.
The following operations and objects are not supported for executing the transformation in SAP HANA:
ABAP routines are not supported HANA database and it will process in application layer.
(rule type, routine, routines for characteristics, start, end, and expert).
only one to one mapping objects process in HANA data base.
Next Blog will discuss ADSO 2nd Model.
Thanks,
Phani.
Pani.....useful blog... when will be next blog?
Great Blog Phani...
a quick question.. I see you are loading from a FF to ADSO....and you created ADSO at the HANA Studio level.
Lets say I want to load from ADSO1 to ADSO2...and when I create ADSO2 will I get any option to "Pick" the source as ADSO1? or do I need to manually Create ADSO2 with all the fields I need and then go into BW to create a Transformation and map the fields?
Yes if you are loading the data from ADSO1 to ADSO2 need to create the transformation and map the fields and execute DTP.
now you will the data to ADSO2.
Thanks,
Phani.
Hi Phani,
I want t load data from ecc datasaource (2lis_11_vahdr) to adso. is it possible ?
Yes you can load, same info object insert into ADSO mapping with transformation load the data from 2LIS_11_VAHDR to ADSO.
When we create ADSO from data source how do we handle delta?Do we still run info package as normal to get delta into the data source or do we have a separate mechanism and we no longer require info package and delta mechanism is handled automatically? Does this apply to ODS view also.
Hello Former Member,
Hope the below given information clears your doubt, if i am wrong please correct me
ADSO – Advanced DSO
When we are creating ADSO, we can specify whether the ADSO should act as DSO or Write Optimized DSO or Info Cube.
Hence we can say that ADSO will behave based on the type we defined during the creation.
We have,
ADSO type Classical DSO
ADSO type Write optimized DSO
ADSO type Info Cube
While creating any object, it may be DSO or Write Optimized DSO or Info Cube, in the underlying data base HANA, 3 tables and 2 views will be generated automatically.
Naming convention of Tables and Views:
3 tables:
/BIC/A<Object name>1 – Activation queue/New data table
/BIC/A<Object name>2 – Active data table
/BIC/A<Object name>3 – Change log table
2 Views:
/BIC/A<Object name>6 – Extraction view (to feed data to next level)
/BIC/A<Object name>7 – Reporting
We will consider 3 Cases while creating ADSO,
Case 1. ADSO type Classical DSO
Case 2. ADSO type Write optimized DSO
Case 3. ADSO type Info Cube
Case 1. ADSO type Classical DSO
Let us Assume, we are creating ADSO type Classical DSO(ZC_DSO).
When we create ADSO ZC_DSO Type Classical DSO, following tables will be created in the HANA database.
3 tables:
/BIC/AZC_DSO1 – Activation queue/New data table
/BIC/AZC_DSO2 – Active data table
/BIC/AZC_DSO3 – Change log table
2 Views:
/BIC/AZC_DSO6 – Extraction view
/BIC/AZC_DSO7 – Reporting
Now when we execute the DTP, data will be first moved to /BIC/AZC_DSO1 – Activation queue/New data table, After activation Data will be moved to /BIC/AZC_DSO2 – Active data table and data will be deleted in /BIC/AZC_DSO1 – Activation queue/New data table.
We can do repoting on /BIC/AZC_DSO2 – Active data table.
If we need FULL/INIT – Data need to be extracted from /BIC/AZC_DSO2 – Active data table.
If we need DELTA – Data need to be extracted from /BIC/AZC_DSO3 – Change log table
If we need data from views, we can extract data from /BIC/AZC_DSO6 – Extraction view.
Case 2. ADSO type Write optimized DSO
Let us Assume, we are creating ADSO type Write Optimized DSO(ZWO_DSO).
When we create ADSO ZWO_DSO Type Write Optimized DSO, following tables will be created in the HANA database.
3 tables:
/BIC/AZWO_DSO1 – Activation queue/New data table
/BIC/AZWO_DSO2
/BIC/AZWO_DSO3
2 Views:
/BIC/AZWO_DSO6
/BIC/AZWO_DSO7
Even though all the tables are created in the HANA DB, utilization of those tables depends on the type of ADSO.
Write optimized DSO is used only for data storage(Corporate memory, data staging purpose)
Hence, data will be loaded to /BIC/AZWO_DSO1 – Activation queue/New data table
Case 3. ADSO type Info Cube
Let us Assume, we are creating ADSO type Info Cube(Z_ICUBE).
When we create ADSO Z_ICUBE Type Info Cube, following tables will be created in the HANA database.
3 tables:
/BIC/AZ_ICUBE1 – Activation queue/New data table
/BIC/AZ_ICUBE2 – Active data table
/BIC/AZ_ICUBE3 –
2 Views:
/BIC/AZ_ICUBE6 – Extraction view
/BIC/AZ_ICUBE7 – Reporting
But in case of Info cube we need to consider
/BIC/AZ_ICUBE1 – Activation queue/New data table as F-table &
/BIC/AZ_ICUBE2 – Active data table as E-Table.
Whenever we load data, first it will be moved to /BIC/AZ_ICUBE1 – F-table, and after COMPRESSION, data will be moved to /BIC/AZ_ICUBE2 – E-Table.
For reporting, data will be obtained from /BIC/AZ_ICUBE1 U (union) /BIC/AZ_ICUBE2
Thanks & Best Regards,
Srinewas.
Thanks a lot Former Member
I have one doubt here.
"If we need data from views, we can extract data from /BIC/AZC_DSO6 – Extraction view".
I have only 2 options to extract data in DTP 1) Full 2) Delta in 7.5 SP03.
If I use Delta --> change log is used.
If I use Full --> Active table is used.
how to take advantage of above extraction view ? Or which DTP settings I have to use?
Hi Phani,
nice blog, thank for the info.
We have one Problem with a aDSO. Everything looks fine and is working, but when we use the aDSO in Reporting (Analysis for Office and Lumira) some fields are missing.
Do you have an idea what is wrong there?
As you see in the hardcopy, we have e.g. a field CONTRACTSTARTDATE the aDSO, but in the data preview and in the frontend we can't see this....
Thanks in advance and best regards
Thorsten
Awesome 🙂
Excellent content on SAP BI
Hi Phani,
nice blog!
I work on "BW on HANA" (Release: 750 Final Release, Patch Level: 6, Hotfix 2) for an BW opearation team.
May I have a question related with ADSOs:
How can I change the display layout of ADSO something like normal DSO or InfoCube layout?
From operational point of view, it is now unusable.
Normal InfoCube:
Reques ID / Reportable / Compression status / Req.status / Monitor / DTP/Infopackage / Req.Date / Upd.Date / Selection Conditions / Transfered Rec. / Added Rec. / Typ. Data Upd. / Source/InfoSource / Name of Source / DataSource / DS name / Source Syst. /Source Syst.Descr. / Req.Time / Upd.Time / Req.Number
ADSO:
Load stat. / Load Req / Load monitor / Act.Status / Act. Req. / Act.rec of Req. / Err.Proc.Log
This is DEPROLABLE!!!
Do we have any possibility or do you have any idea?
Thank you in advance,
Istvan VERES
Hi Phani,
Nice blog , Kindly share ADSO 2nd Model.