Delete Duplicates In a Transaction Fact table

[sleezy]

I have a large table consisting of 68 million rows. I have duplicates in the Fact table
Whats the best way of deleting the duplicate rows.
I left out the Transkey because thats the only thing that uniquely identifies a record.
Whats the best way of deleting the duplicate rows on a large table like this one ?

tbFactF2005 fields as listed below
'LoanKey
,BookKey
,ClientKey
,EmployerKey
,LoanRefId
,ProductKey
,HierarchyKey
,TransGroupKey
,TransactionCode
,WhRef
,Account
,Amount
,LoanRef
,Client
,TransactionType
,ValueDate
,ProcDate
,Company
,ReverseInd
,BatchNumber
,Reference1
,Reference2
,Narative1
,Narative2
,User
,Time
,Date
,Source
,CashInd
,CommInd'

[skhanal]

You need to run the delete in smaller sets. How many rows out of 68 million you expect to be deleted?. You may need to run more frequent COMMITs depending on how many you delete.

You can use

SET ROWCOUNT nnnn

or WHERE clause to get smaller set.