Its been noticed many times that some of the processor in EDQ are mixing up values in the results browsers, that too primarily in the Upper processor.
For Example.
If you have an Email = John.smith@xyz.com , and you pass it thru Upper processor, then the results of Email.upper = david.cooper@abc.com , a totally new values from a different record set.
You have to be careful with some processors, sometimes though it shows the wrong values in the results browser but if you check the final staged data there may not be any wrong values. It’s always better to check results of each processors and validate.
Oracle says its memory management issue you have to allocate appropriate memory during installation if not Merge processors will cause such mix ups. Though it’s a DBA’s job, developer has to be careful of such issues so that it don’t reach the staged data you are writing in the end.
As an immediate work around, you can re-run the process and see if the wrong values still appear in results window.
Best practice is to use such transformation processors in the beginning of the process and also keep the Merge processors in the end, which could cause a memory issue as per oracle.
Comments
Post a Comment