Hi Folks,

I have a simple SSIS package -> It reads a local text file which has 10 rows of data ( id, name, telephone # ) and puts it into a table.

It uses the "SSIS Flat File source" to read and a "SQL Command" to insert into the table. I can see that it reads line by line and puts each line into one row in my table.

Now, my production data is over 5 GiG of mainframe data and it seems their data is arranged in some hierarchical form.. so the position or arrangement of data in that file is important.

I pulled the data using my package and as far as I can see , my SSIS package pulled one line at a time ( from the flat file) and pushed it into my table. For each row, I also created an identity column in my table to be able to identify the positional arrangement of the hierarchical data and then use relational mappings to suit our business needs.

In all of this, my assumption is -

"SSIS reads one line at a time, inserts to my table and goes down to the next line .

It does NOT read a snapshot of rows from the flat file so as to write them into the table using internal ordering methods based on that particular snapshot
"

My question is .. is my assumption correct ?

Thanks as always!

Warm Regards,
RanjitSHans

Correct.What do you mean by "internal ordering methods"? Internal to SSIS or the hierarchical database? How would SSIS guess the internal ordering method of the hierarchical data?

You will have to store soma data to relate the new line to the context of the previously processed line(s). When you write it to table, you must use that context data to convert the hierarchical data to data suitable for a relation data model.