Data Source Key Field error for DXC

[Former Member]

Hi All,

While trying to activate the data source 2LIS_11_VAHDR and 0MATERIAL_TEXT in the BW System for DXC consumption I get the following error -

Error in HANA DB system: Unable to modify DSO table '2LIS_11_VAHDR': no key field defined (at least one of the defined columns must be a key column)

How can the data source keys be maintained.

Pl assist. Thanks.

[Former Member · Accepted solution]

Hi Saurabh,

These notes need to be applied in the Netweaver server where you're trying to implement DXC. BASIS team will be able to help you with these notes.

1701750 DataSource: Secondary Index on the PSA

If your DataSource is missing a key, apply this note to any BW systems connected to the SAP

Business Suite system you are using with DXC.

1677278 DataSource: Changing the Key Definition (A version)

Provides a report where you can define a semantic key for any DataSources that are missing

keys. DataSources without keys will cause an error when you try to generate the In-Memory

DataStore Object in SAP HANA.

Before applying this not to your SAP Business Suite system, first apply SAP note 1701750 to

any BW systems connected to the SAP Business Suite system you are using with DXC.

1710236 SAP HANA DXC: DataSource Restrictions

Lists specific DataSources not supported by DXC.

Hope it helps.

--

Shreepad

[Former Member · Accepted solution]

Thanks Shreepad.

Am getting this checked. DXC seemed to be working fine for the data source 0CO_PC_10 (this was done as part of COPA RDS installation).

I can see the key fields in RSA2 for this extractor in ECC. Why does it work for one data source and not others, with the same system configuration. Pl advise. Thanks.

[Former Member · Accepted solution]

Hi Shreepad,

I checked with my Basis team and these notes have already been applied on our systems. Thanks.

[Former Member · Accepted solution]

Hi Saurabh,

As you know about how the DXC works by redirecting the flow of Data Source from PSA to an IMDSO in HANA.

So when this happens a set of tables has to be created in HANA based on the structure of Data Source. When the Data Source itself has no keys then the system throws an error while creating a table with no PK.

Hence some of the Data Sources(With Keys) are directly supported, but others need some modifications in their structures which is given by these Notes.

--

Shreepad

[Former Member · Accepted solution]

Hi Saurabh,

As the notes are installed from BASIS, you should be able to perform the below steps:

1. SE38

2. Execute Report : 'ROSOA_CHANGE_KEY'

3. Select P_CHNG

4. In P_DS Enter your extractor name.

5. Make Changes to the structure by having keys.

6. Then redo the steps of Activation/Replication of Data Source.

7. Then try the replication of data again.

Hope it works this time

--

Shreepad

[Former Member · Accepted solution]

Hi Shreepad,

This works, thanks. I have some queries on the steps you mentioned above -

1. Where do we run the the program ROSOA_CHANGE_KEY, on the source or on the BW System. I did both and it works in both cases, but what is the right approach.

2. If this is done in the source and if the source feeds a HANA System (using DXC) and a BW System (using the normal extraction), will this affect the extraction into BW System.

Pl advise. Thanks.

[Former Member · Accepted solution]

Hi Saurabh,

Glad it works!!!

1> Running the code only on source system should be enough.

2> Though I have not tried it, but I think it should not affect the extraction to 'another' BW System as this program just adds semantic key(s) and does not change the physical structure of the extractor.

Another point here would be the Embedded BW System in the source system is of no use once DXC is configured.

--

Shreepad

*Kindly mark answers as Helpful/Correct if they solve your queries.

[lakshminarasimhan_n4 · Accepted solution]

Excellent solved our issue..