What steps required to load data again to DSO and CUBE after changing the routine

[pavneet_rana]

Hi to all,

Can any one suggest some steps/solution of how to load the data to DSO and CUBE after changing the routine.

Example: There was some code error in end routine written between two DSO (DSO1 and DSO2).

               Due to this data was coming incorrect to DSO2 and above CUBE.

The routine was changes between DSO1 and DSO2, to get correct result.

Now after changing the end routine code between DSO1 and DSO2.

should i need to do selective deletion for incorrect data in CUBE and DSO2 .

and reload the particular data (remove via selective deletion) again, from PSA to DSO1 and to DSO2 and then to CUBE.

or i need to take some other step for that.

Please share some light on it.

Regards

Pavneet Rana

[pavneet_rana · Accepted solution]

Thanks for reply to all,

I have some more doubt.

1. If i knew the record that are incorrect in DSO and CUBE, then i can go for selective deletion from CUBE and DSO2. and reload them after chaning the routine between DSO1 and DSO2.

Does it will disturbed the delta of DSO2 / DSO1 .

If yes then how to avoid this.

2. If i don't knew the record which are incorrect in DSO2 and CUBE. then how to what step should i take after changing the routine between DSO1 and DSO2.

Then how to proceed furthure.

Please can any one share some light on it.

I shall be thankful to you for this.

Regards

Pavneet Rana

[Former Member · Accepted solution]

Just to be clear, I'm assuming that the design of your Infoproviders is stated below:

SOURCE == Delta (Overwrite) ==>> DSO1 (Standard) == Delta (Overwrite) ==>> DSO2 (Standard) == Delta (Additive) ==>> Cube (Standard)

With scenario 1, you can just do a full upload for both DSO1 and DSO2 for the erronoeous records. These will create the delta records to change the erroneous data. Then you can do a normal delta from DSO2 to Cube.

For the 2nd scenario, If you can further narrow down which part of the data is causing the discrepancy (Month, Customer, Material, etc) then you do the suggestion as stated above.

All else fails, do a full deletion for ALL infoproviders then just reload the data again. This is a LAST OPTION and should be done during off hours or/and you should request for quiet time.

I'm not sure what will happen if DSO1 is a write-optimized DSO. It messes up the delta in my opinion.

[Former Member]

Why not try doing a filtered full load for the affected records from DSO 1 (Staging/Propagation) to DSO 2 (Transformation)?

The data will be overwritten in DSO 2, then it will create delta entries from DSO 2 to CUBE.

I used to do the deletion of both CUBE and and DSO 2 but running a million records with 5 lookup routines takes almost me a day to reload. Doing full loads between DSO 1 and DSO 2 save me a lot of time and it won't affect users running reports.

[pavneet_rana]

Thanks for reply to all,

Yes, i knew the data which are incorrect in CUBE and DSO2.

1. So can i go for selective deletion for incorrect record in CUBE and DSO2.

And reload again data (deleted via selective deletion) from DSO1 to DSO2 and CUBE after changing the end routine.

2. OR i need to delete the all record in CUBE and DSO2. and reload all record to DSO2 and CUBE again from DSO1.

3. or any other step i need to do.

please  share some light on it.

Regards

Pavneet Rana

[former_member186082]

Hi Pavneet,

As you say that changes done to end routine, make sure that entire data is not affected. If you are really so sure which data is incorrect, go with selective deletion and reload using the same selections. This really saves us time.

Reload from DSO1 -> DSO2 and then to CUBE

Regards
Chandu

[ravi_chandra3]

Hi

Better to delete the entire request and then trigger the DTP .

[Former Member]

Hi Pavneet,

Yes, go for selective deletion in DSO2 and cube, only if you are very certain that you know what are all the records that needs modification in the routine.

Better practice would be to delete entire contents in the target infoproviders and load again.

This is a time consuming effort, but you can ensure stability.

Br,

H