DSXchange

Welcome aboard. You need to set up an ODBC DSN (Excel driver) with the pathname of the worksheet. When importing table definitions you must check the System Tables check box; for some reason best known to Microsoft, Excel worksheets are presented to the ODBC driver as system tables.

Version 8.0 combines DataStage and QualityStage in one Designer, so there are QualityStage stages that perform the QualityStage tasks (investigation, standardization, matching and survivorship). It is about these that the OP is asking. My gripe is that the OP expects us to read the manuals for him/h...

How did you capture the score? What was in the score in each case? Describe your job design and the corresponding score.

Look more closely. It's there. Select the Aggregations folder and look in Available properties to add.

Welcome aboard. Two adjacent Modify stages are not supported (this is in the documentation). Put a Copy stage between them.

How soon is the interview?

You probably need parameter markers :1 through :22 (to correspond with the 22 columns in your Columns grid) rather than :2 through :23.

Of course it can. It's the surrogate key that can not. If some attribute of the customer (maybe name) changes, in a Type 2 SCD there will be a new row with the same CUST_ID, but with a new surrogate key.

Why not prove it either way? Run the job with Force set and not set, capturing and examining the Score. Which one has no copy operator? (Note that the Copy stage must be a straight-through copy (no changes) to be eligible to be optimized out.)

Try using explicit Sort stages on the inputs to the Join stage, even if they have Sort Mode set to "Don't sort (previously sorted)".

Rather than removing the file, why not use the target Sequential File stage to append to the "black hole" (/dev/null in UNIX, .\NUL in Windows)?

Care to put the line numbers on that? I don't have the time.

You have them, so you have the documentation and help files. Why don't you tell us? (The TXmap stage is not new in version 8.x.)

What partitioning are you using on the inputs to your Join stage?

No. What you are doing is not best practice (obviously). The IP address is only available while a particular user is logged in - it is not preserved in the DS_AUDIT table (which, I guess, expects that each user will have a separate login ID).