In our project , we had a three level product hierarchy structure with length of the each level being 5,5 and 8 respectively . Based on this, the characteristics fields in COPA were defined using the standard characteristics PAPH1 ,PAPH2 and PAPH3 . When doing this in CO-PA , SAP created domains RKEG_PAPH1 ,2 and 3 with similar length as the levels in structure PRODHS.
After this was completed , the product hierarchy structure has been changed in PRODHS to four levels with length as 4,4,4,4 respectively. As a result of that the domains RKEG_PAPH1 , 2 and 3 are not matching with the new structure and the PAPH1 to 3 characteristics in COPA have been made unusable.
To fix this SAP help is suggesting two things:
1) First change the number of characters in domains/data element RKEG_PAPH1 to 3 to match the new structure in PRODHS
2) Find out the list of tables/structure where RKEG_PAPH1 to 3 is used and perform a data conversion.
From the where used list , we are able to see that this are used in 21 structures which are in turn used in numerous programs, structures, BAPIs etc.
We would like to know :
1) What is impact of doing this on numerous programs, structures, BAPIs etc ?
2) How these data conversion is done and what utilities are used for the same? Is it required to do this as our project is still not live? Only Dev and Quality systems now running.
3) Potential issues that we may face?
4) Any other alternative solution.
First impressions: Since you are not live yet, the impact is much lesser than if you were live. So thats saving grace.
I gather from your post that you either don't wish to leave these Chars as they are (ignore them and create new ones instead). Have you explored the possibility of using these for some other 5-char or 8-Char Characteristics?
The impact is massive, but restricted to BAPIs and custom programs you may have built based on these Chars. Whatever method you follow (new Chars or change these existing ones), you will still need to modify your reports to reflect the new data structures. But then, do you have an option?
You could use custom programs for data conversion, mapping your lagacy data to the new fields, or simply use Excel spreadsheets - that also depends a lot on the source of your legacy data.
Is it required? You don't data convert until your first cutover and that is mostly into a Regression Test server and then into the P server. So do you really need to bother about data conversion to the Dev or the Quality servers?
You may need to explore the possibility of deleting these Chars from the Op Co and including new ones. While this is not a recommended option, that way you can still transport only the required Chars to the P server and ignore the rest. Your Test or Quality servers will go for a toss (data mismatches etc.), but you could try and not let it affect your golden client.
Alternatives: explore the possibility of ignoring these 3 Chars and creating 4 new ones instead. That, by far, could be your best option.
All the best.