how to get reason for change information

Hi all:
I have a basic question: our customer table has about 200 columns, some of them are SCD 1 and some are SCD 2. and when customer information is updated, it normally involve a few different fields, they could be scd1 and/or scd2. customer information come from a few different source tables.
so my question is how to get reason for change information? for example customer may update their address (scd2), first name (scd 1) and credit detail (scd 2) at the same time, I know its wrong to compare every field to figure out it is a scd 1 change or scd 2 change and which fields are changed. is it any other solution for that?

thanks
Howard

Hi - if you need to know which columns have changed then the only way is to compare the old and new versions of every column, there's no way round it.
Obviously there are many ways of doing this, some of them dependant on the DB/ETL tool that you are using, but you will probably only find the best way by (informed) trial and error

Regards

If this is something you need to provide, as well as support reporting against, your best bet would be to implement a logging fact table that tracks what changes were applied. It would be a by-product of the process that applies the updates. It is real messy to do using most ETL tools, (one row coming in, an indeterminate number of rows going out) but it can be done.

hi, thanks guys, hmmmm, looks like i need dig into the ETL tool!