cancel
Showing results for 
Search instead for 
Did you mean: 

Delta from ODS to Cube

Former Member
0 Kudos
52

Hi,

We have a scenario where data is coming from Application 11 for SD. Data Flow is as follows :-

2lis_11_VASCL,& 2lis_11_V_SCL --> DSO1 - Delta to DSO2->Delta to Cube

Till August 2011 all the data loads were Full and hence in the process chain all the change logs were deleted on daily basis. But from August it was changed to Delta from DSO1 onwards. But there was no change in the process variant which was deleting the changelogs of DSO 1 and DSO 2. Now we have found that many of the values in the cube are not from August onwards.

If we do a full load from data source to the DSO 1 and then delta to DSO 2 and then again Delta to Cube will this cause any problems? Will it resolve the issue? Any other suggestions regarding the situation will be appreciable.

Obviously we have now change change log deletion variant and now it only deletes the change logs which are more than 7 days older.

Navesh

Accepted Solutions (1)

Accepted Solutions (1)

Former Member
0 Kudos

Hi Navesh,

A change log is not generated until and unless there is a change in the record.If you Do a full load to First level DSO and Then the delta to Second level DSO ..the change log of second level DSO may not carry the FULL LOAD Details because already the data is available in the active table of the second level DSO.

As far as i understand from your scenario

You are having correct data till second level DSO ...PLease proceed with below two ways

1)Delete the data in the cube and have a full load from the Second level DSO....and initialize data with out data tranfer(Set the data in Source fetched) in DTP.........Agree that its a big task if the data which you are maintaining is huge...but this is the Easy one

2)Check for an object in the cube which captures the date of loading or any other field which differenciates data

Assume that we are having the loading date

a)Perform selective deletion from August to till date( this is to avoid Duplicate/Doubling of data if this data is already available

in Cube)

b) Perform Full load from August to till date

c)Initialize data with out data tranfer(Set the data in Source fetched) in DTP (This avoids the further update of data which we have updated through the Full load)

Please let us know for furhter information

Regards,

MADhu.

Answers (2)

Answers (2)

former_member188080
Active Contributor
0 Kudos

Hi,

Looking at your email I would suggest just go ahead and reload the DSO and cube again..

or else

1. select the records which are different in cube in value compared to source

2. Then go to DSO1 and DSO2 and see where there is difference

3. then if there is no difference ..no need to delete the data from these DSOs.

4. then go to cube delete the data and reinit and load data again..

Thanks and regards

Kiran

anindya_bose
Active Contributor
0 Kudos

Hello Navesh

What Madhu is saying is absolutely right ; couple of month back I tested this in system.

Here is the link .

If you are doing full load to DSO1 , then you can do that after deleting the content. That will generate new request id in change log table while activation. Then a delta load should bring all from DSO1 to DSO2. But cube data needs to be deleted and reload.

Regards

Anindya