Posted by aknoch on 08/21/07 23:33

My basic situation is this - I ONLY want duplicates, so the opposite
of DISTINCT:

I have two tables. Ordinarily, Table1ColumnA corresponds in a one to
one ratio with Table2ColumnB through a shared variable. So if I query
TableB using the shared variable, there really should only be on
record returned. In essence, if I run this and return TWO rows, it is
very bad:

select * from TableB where SharedVariable = 1234

I know how to join the tables on a single record to see if this is the
case with one record, but I need to find out how many, among possibly
millions of records this affects.

Every record in Table1ColumnA (and also the shared variable) will be
unique. There is another column in Table1 (I'll call it
Table1ColumnC) that will be duplicated if the record in Table2 is a
duplicate, so I am trying to use that to filter my results in Table1.
I am looking to see how many from Table1 map to DUPLICATE instances in
Table2.

I need to be able to say, in effect, "how many unique records in
Table1ColumnA that have a duplicate in Table1ColumnC also have a
duplicate in Table2ColumnB?"

Thanks if anyone can help!

-- aknoch

• Next in forum: Accessing one db by two sql server installations (eg. native vs vmware)
• Prev in forum: Re: An expression of non-boolean type specified in a context where a condition is expected, near ')'.
• Thread view: Need to find instances of duplicates within a column; joining 2 tables.

[Reply to this message]