This content has been marked as final. Show 5 replies
Try the search function, we (I) have answered this before !
Basically set up logical primary key, ODI needs the PK as it uses it to join the Journalizing view back to the source table. If your using Streams / Logminer the PK columns also determine what gets created as the unconditional log group, etc.
I would suggest you set a PK, then simulate the Journal Start and go through the output steps with a fine tooth comb to really understand where the PK fields are used - if you get it wrong, you can get Duplicate rows presented to ODI when it reads the journal data store.
For example :
Re: Configure CDC on tables not having Primary Key
Re: CDC without source primary key
Edited by: PeakIndicators_Alastair on Nov 6, 2012 2:33 PM
Thanks for you reply.
I had made search and already found the link you mentioned.
But I couldn't figure out how to create a primary key in Model in Data store on view.
I double clicked on view in model and tried to find option to make a primary key.
But I could not found.
Can you please advice.
Expand the Datastore, right click 'Constraints' and insert a new one (of type Primary Key).
Thanks I found it. I appreciate your help...
Glad to help,1 person found this helpful
Which JKM are you using?
If you feel the thread is answered please mark as closed so other users can search for answered Q's in the future , mark any helpful / correct posts also please.