Data Difference for different modes of update

[Former Member]

Hello Experts,

We see difference in data updated to the target when load is tried in delta mode and full mode.

The number of records planned for load remains the same and in the end routine we do not mention any check/filters which should act accordingly for delta/full update.

Though the filter conditions remains same, missing entries are found.

-> We do not have semantic groups defined in our transformation

-> We do direct assignment. If the target field is made constant will resolve our issue ?

Are the semantic groups not defined is the cause for missing entries ?

Please clarify.

Thank you very much for your help in advance.

Regards,

Kavya S

[Former Member]

Hi All,

When I try to execute a delta load,

For example

Datapacket	Document Number
Packet 1	1234
Packet 2	1234

Same document number(key field) which has different datapacket IDs should actually get consolidated and single record should be populated.

Say I have 5 document numbers like above scenario, when in delta mode, out of 5 document numbers 3 provides consolidated data whereas other 2 take any one datapacket ID value.

When in full mode, I get the expected consolidated results for all the 5 document numbers.

Hope this clarifies.

Regards,

Kavya S

[Former Member]

Hi Kavya,

Are you loading data to a DSO? if yes then according to Key records should automatically get consolidated during activation.

But if you require all same key data records in same package then you require symentic groups in DTP. In deltas it is common that record with same key come in different packages. So to include them in one package you have to mention symentic groups in DTP so that all 5 records come is same package. With this you will be able to process all of them in your end routine.

Hope this helps.

Thanks

Amit

[former_member186399]

Hello Kavya,

There is a difference when you run the DTP in delta mode and Full mode. Also it depends on whether the data moves from PSA  or from DSO to a data target

1) when you run a full mode system collects the data from every PSA together and then passes on to the data target

2) When you run a delta mode system picks it up request wise and therefore you can have the same document number coming in different data package. This can be cause am issue if you are modifying some data using in routine. Your approach in this case has to be different for your routines.

Regards

Gajesh

[RamanKorrapati]

Hi,

please explain your issue with some examples.

What is your source and target?

Full  and delta loads never be the same number of records.

Delta volume will be less compared to full load. But its depend on client business.

Thanks

[sachin_kulshrestha2]

Dear Kavya,

You need to elaborate more on you EDW layer. To what i understood

1. If you upload data from DSO To DSO  in overwrite mode kyf in your transformation than the number would remain the same

2.  If you upload data from DSO To DSO  in append mode kyf in your transformation than the number would differ since the similar key field would append in the already existing record in the dso

3. Cube is aggregation with similar characteristic information will be aggregated into one single record for each load request.

Hope these scenario would help you further analyse at your end.

[fcorodriguezl]

Hi Kavya,

Are you try load data of ods1->ods2?

If yes, check in your dso2 if have old request maybe that are a reason with difference, is very strange your issue.

Regards.

[ccc_ccc]

Hi Kavya,

You mean to say that, same number records are updating either update mode full/delta?

If yes, please let me know how many transfer and how many added and more detail.

Regards,

Nanda