Infocube Recursive Loads

Hi Experts,
I´m facing a scenario where an infoobject inside an Infocube will have to be filled based on a master data table after the transactional load is done. This could be simply achieved by creating another infocube with this infoobject and loading data from the first one to the second filling this data in transformation.
However, due to the amount of data volume involved, I can´t duplicate all this data to fill up this information. So, the recursive load came up as an idea. An infocube can be the datasource to itself.
I have some doubts concerning this scenario:
1) Will all data be duplicated anyway ?
2) Can a DELTA approach be implemented based on the empty field to be filled ?
3) Comparing to a normal load scenario, is there any performance issues specific to this solution?
I´ve checked similar posts, but I haven´t found the answers for the questions above.
Any other ideas to work on this scenario?
Thanks!
Regards,

Outside of the box:
Think about creating a highly compounded InfoObject to use as a "key" in the InfoCube. Use this InfoObject as line item dimension in your InfoCube. Make your update field a mater data attribute of your highly compounded InfoObject. Update the highly compounded InfoObject. as master data.
Recusive load
Create an export data source from the InfoCube to an InfoSouce then use the InfoSource as a source for the InfoCube,
Add or stripp out key figures in the transformation.
Never tried it myself, let me know if it works!
John Hawk

Similar Messages

Reflexive / recursive loads

I read in the book <a href="http://www.sap-press.com/product.cfm?account=&product=H1941">Efficient SAP NetWeaver BI Implementation and Project Management</a> by Gary Nolan (appendix B.8.2, p. 357, BW Data Moel Review Checklist) that reflexive or recursive loads on transactional data should be avoided for performance reasons.
Does anyone have a better idea about the specific meaning of reflexive or recursive loads? Can you provide me an instance of recursive load?
Thanks, <a href="https://wiki.sdn.sap.com/wiki/display/profile/Davide+Cavallari">Davide</a>

I post the detailed explanation Gary Nolan - the author of the book - kindly sent me by email:
Recursive loads are created by creating an export datasource from an infocube, thus creating an extractor from that cube. The extractor is then loaded into the cube again. This creates a cube loading to itself. While this seems like an illogical setup, there are actually legitimate reasons to use it.
You may want to loop through all the transaction records to check something or to update something. For an easy example, perhaps you are tracking sales orders and have the salesperson stored in your infocube. The salesperson to customer assignment is stored only in a ztable. You read this ztable each time you load to assign the salesperson and put it in your InfoCube. This works fine for all new sales orders, or those that have changed for the day because you get a delta. Each delta record that comes through, you read the ztable and put the salesperson into the cube.
If a salesperson assignment changes in the ztable all new records would take on the new salesperson.
But, what about the records that have been loaded in the past? Those old sales orders are not going to come through on a delta, and thus would have the old (and incorrect) salesperson.
In order to fix this we could use a reflexive load to reload the infocube to itself and redetermine the salesperson for all records in the cube. It has obvious performance implications because of the volume, but it does work.
This should be considered the last resort for this type of data model. Best to handle it via a master data table and use attributes of that...but sometimes this cannot be avoided.

Recursive Load

Hi,
Please let me know what is Recursive Load.
Thanks and Regards
Asim.
Edited by: asim khan on Dec 23, 2009 8:16 AM

Hi ,
Recursive loads are created by creating an export datasource from an infocube, thus creating an extractor from that cube. The extractor is then loaded into the cube again. This creates a cube loading to itself. While this seems like an illogical setup, there are actually legitimate reasons to use it.
Please have a look to this thread , you should get all your answer.
Reflexive / recursive loads
-Vikram

Recursive Load Problem!

Hi Experts,
i've made a recursive load from Cube to Same Cube that means (recursive loads are created by creating an export datasource from an infocube, thus creating an extractor from that cube. The extractor is then loaded into the cube again.)
the reason why we do this is to pupulate a new field in the cube with values of another field in same cube.
is there any option via routine to update the records without creating new one. i mean every time i load the Delta-Upload from Datasource(Cube itself) to the cube again it loads everything. even if there were no changes in the cube and the delta-load was initialized.
thanks
Younes

Hi Subray,
the Logic is already in the recursive Update rule.
i do first a normal Full Load to the cube. and then start the recursive infopackage which has the Updaterule with the start routine that contains the logic to populate the field. that works fine but i wanted to test when i do the recursive one twice it loads all data again! i thought that th Delta Logic will only loads data which was added or modified!
Thanks
Younes

Ratios of a Infocube - erroneous load

I have a problem in the load of the ratios of a Infocube. When I make a querie of this Infocube the ratios that I have defined leave to 0.This Infocube to me load through a ODS and the ratios of the Infocube are loaded of direct form of ODS-Infocube and in the ODS these ratios have values, nevertheless in the Infocube these ratios leave me to 0.

Hi,
did you check whether your cube has the ratios? Check if it'S the right keyfigure (the same like in your ODS)? IS it filled from the ODS? Did yiu check in the query if there are some restrictions in characterisitcs that the query can't find ratios?
Try fisrt to see values in your cube with listcube. If there are no values check your transformation / update rule. If there are values in the cube check your query for restrictions, filters, conditions.
Regards
Jürgen

Recursive Load in BI

Hi Experts,
I am doing a recursive load in one of cubes. I just need to update one the flags in the cube from a external table table, So i have return a routine for that. So the source and target is the same cube. I have already done this, It is working properly in Quality and Development. Is this not a best practise. Please let me know.
Thanks,
Jasmine

If you are doing loads from cubes - then you will need the indices to be up - especially if you are using delta loads from the cube which is based on the request number.... But then if you have indices down then the extraction performance might be slower ( not too slow - since the F Table is partitioned by request ) and then bring up your indices after the load.
If you are extracting from the cube with selections and without indices then your load will take more time.... because given your data volumes - reading the data from the cube without indices will take a lot of time...

How can an Infocube time load be calculated?

Dears,
I have to make a large number of InfoCube copies in my system, basically I will copy all the information from an original InfoCube and load it in a mirror InfoCube, really simple, but, I need to pre-know the time the loadings are going to take place. Each infocube is different from the other, different number of characteristics,ratios and total registers.
I can calculate the number of registers and space taken by each InfoCube, but how can rearrange some kind of formulation in order to find out an estimated time the loadings are going to spend in its execution?.
Thanks to all and regards,
Katy Rozic.

Hi,
the attributes and/or texts that an infoobjects has are time-dependent. So it could be that an attribute is time-dependent and another one is not.
If you change an infoobject and goto "Master data/texts" tab, you see an option to select texts and also if these texts should be stored time-depenent.
If you goto the "Attribute" tab you see all the attributes of that infoobject (or click on Detail/Navigation Attribute). For each attribute you have the option to select if this attribute should be stored time-dependent.
regards,
Raymond Baggen
Uphantis bv

Reloading Vs Recursive load

hi guys
I am working in one BI for financial domain.Here,if you are aware some times data will be reloaded by deleting old data when new rates are introduced...(Rate means...Clients wanna compare actuals data for a year in ACT_2010 RATE or BUD_2010 rate like that for comparison)....sometimes instead of deleting and reloading,we are doing recursive loads manually(cube has transformation to itself and one DTP)....
In what situations do we need to prefer one over other?
Thanks for your replies.
Rgds,
P Olin.

Hi,
Let us say there was a change in plan data for year 2010. The user wants to see the new data. Usually plan datasource is not delta enabled and hence we will do a full load to cube. SO we will do a selective deletion of 2010 plan data and then do a reload of plan data for 2010. Instead of this you can do a recursive load. This transformation will be a one to one mapping except that the keyfigures are reversed. So the total is negated. After this load you can do a full load for plan data for 2010 and hence your cube shows new value.
Regards,
Raghavendra.

Deleting Data from Infocube before loading

We are doing full loads by "Plan Version" into our CCA cube. We do these full loads each night. Eventually a "Plan Version" will stop changing and we can discontinue loading.
My question is there a way we can delete data in the cube by "Plan Version" before the new request is loaded. Right now it deletes overlapping request after the load but this is inefficient.
We would like to delete the data before we load.
We are doing this using a process chain in order to drop indexs first then reload and update statistics.
Thanks
Richard
Message was edited by: Richard Carrier

Richard,
You can do full replacement load of your plan data by changing your plan data load infopackage which persumably have the selection to load plan data i.e. 0vtype 20-90.
In the Data target tab of the infopackage, select the update data target for the cube(s) in question and click on the automatic loading of identical/similar requests from the infocube (the column to the right of the pencil) and click on delete existing requests button and tick
a) full or initial data load and selection are the same or more comprehensive. I typically like to tick also the
boxes for infosources are the same, datasources are the same and source system are the same.
This will load the infopackage as you normally do and after that immediately delete any same or encompassed requests. i.e. if current value type selection is 20-90, it would delete a request for 20 as well as a value type for 40 and one for 50...
Hope this helps,
Mary
Hope this helps.

Infocube data loads fail with UNCAUGHT_EXCEPTION dump after BI 7.01 upgrade

Hi folks,
We're in the middle of upgrading our BW landscape from 3.5 to to 7.01. The Support pack we have in the system is SAPKW70103.
Since the upgrade any of the data loads going to infocubes are failing with the following
ABAP dump UNCAUGHT_EXCEPTION
CX_RSR_COB_PRO_NOT_FOUND.
Error analysis
    An exception occurred which is explained in detail below.
    The exception, which is assigned to class 'CX_RSR_COB_PRO_NOT_FOUND', was not
     caught and
    therefore caused a runtime error.
    The reason for the exception is:
    0REQUEST is not a valid characteristic for InfoProvider
Has anyone come accross this issue and if so any resolutions you adopted to fix this. Appreciate any help in this regard.
Thanks
Mujeeb

hi experts,
i have exactly the same problem.
But I can`t find a test for infoobjects in rsrv. How can I repair the infoobject 0REQUEST`?
Thx for all answers!
kind regards
Edited by: Marian Bielefeld on Jul 7, 2009 8:04 AM
Edited by: Marian Bielefeld on Jul 7, 2009 8:04 AM

Automation Infocube History load

Hi,
Scenerio : i have 3 infocubes in which we load yearly data...say 2004, 2005 and 2006 data in each cube....now to load 2007 data, we have to delete contents of 2004 yr and dump thr 2007 records so 3 yrs of history will b maintained....now if 2008 comes, 2005 data will b deleted n so on ....
i know, i can do this manually deleting and populating records at starting of each yr ..
can ne1 has idea if i can automate the process and maintain this globally, like i can use the same code or method for other cubes as well....
Can we implement this using decision making process types in process chains???
waiting for ur reply guys.......
Regards
Ankur Shah

Hey,
i have 3 cubes and records are loaded from ODS.
I require to load data from ODS to cube an its a delta load on daily basis. So for the current yr, it will b to say cube 1.For the yr 2005 and 2004, i have loaded the data to cube2 and cube3.okies..
Now when 2007 comes in i require to delete contents of cube3 (2004 yr data) and load it with 2007 yr...and it will on daily basis from ODS..
Suppose i included this flow in my Process chain, then too how i can trigger it to say particular cube...i mean now if 2008 comes, thn how the process chain knows which cube to delete n whr to load data...once it knows which cube to delete the data, i can obviously fill the records...
Regards
Ankur

"Recursive Load Attempt for namespace:AgServiceStatus" error

When opening LR5 on my HP desktop this morning I received the following error in the library module. Lightroom did a backup last night before closing. Has anyone come across this problem before?

Have you installed any 3rd party plugins?

How to load historic values of a char in production

Hi,
I have a requirement where some vendor names are missing for few CO orders.Delta is running succesfully for past 2 years.
I will enhace the data source to pull the missing values from diff.tables but how to pull those values into the cube.
Scenario1: Deleting the data in the cube
Scenario 2: without deleting the data in the cube.
I am more interested in not deleting the data.
Please suggest in both the scenarios.
Thanks
Sri

Check out these links...This should solve your problem
Recursive on infocube: Possible?
Infocube Recursive Loads
Re: Recursive Load in BI
Reflexive / recursive loads
Edited by: Ananda Theerthan on Apr 28, 2010 7:31 PM

Data not loading for new Fields InfoObjects from DSO to InfoCube

Hi Gurus -
               I have a DataSource that is providing data to existing DSO then to Infocube. My client asked me to added couple fields to DataSource and get the data to DSO and then DSO to InfoCube.
Here is the Old scenarion:   DataSource -> DSO -> InfoCube.
Here is the New Scenation:
     DataSource ( Added new fields) -> DSO (added new InfoObjects for Corresponding DataSource FIelds) -> InfoCube( Added new InfoObjects to mapped from DSO)
I added the new fields to DataSource, added the corresponding InfoObjects to DSO and InfoCube.
I successfully loaded data from DataSource to DSO. Data is populating for the new Fields/InfoObjects in DSO.
But when I load data from DSO to InfoCube, I don't see any data for New Fields InfoObjects in the InfoCube.
Data from DSO to InfoCube is loading fine for the Old InfoObjects Fields but not for the New InfoObjects I added in InfoCube.
-Sonali

Hi,
Why dont u debug the load through DTP debugging and check what happens to the source field and target field once it passes through the transformation. You can easily trace back where the fields are becoming blank.
The loads which you have mentioned earlier did it have values in Added Records/Transferred Records column for the cube.
Regards,
Mani

Process Chain taking long time in loading data in infocube

Dear Expert,
We are loading data thru PC in AR cube it takes data frm
PSA-> DSO->Activation->Index Deletion->DTP(load infocube)->IndexCreation->Create Aggregates.
In Index creation everyday its taking long time around 9 to 10 hrs to create it
when we go in RSRV and repair the infocube thr loading of data happens fast. We are doing it(RSRV) everyday. In DB02 we have seen dat 96% tablespace is used.
Please tell permanent solution.
Please suggest its BI Issue or Basis.
Regards,
Ankit

Hi ,
We are loading data thru PC in AR cube it takes data frm
PSA-> DSO->Activation->Index Deletion->DTP(load infocube)->IndexCreation->Create Aggregates.
In the above steps insted of Create Aggregates it should be Roll up Process of aggregates.
You can ask the basis team to check the Table space in the transaction db02old/db02.
Check if there is long running job in SM66/SM50 kill that job.
check there should be enough Batch process to perform the steps.
Hope this helps.
"Assigning points is the ways to say thanks on SDN".
Br
Alok

Infocube Recursive Loads

Similar Messages

Maybe you are looking for