Abdul, My questions here are: 1. Can I write logic in SSIS to update if the record already exists and certains columns are same, if so how can I do that. 2. If I incorporate logic into SSIS, can the performance be same as simple import process. Because the simple import process in SSIS uses the bulk insert mechanism with batch inserts of records. The answer to your question isn't as simple as you may think. In the first place, your use of the phrase "can I" adds to the problem because it seems that you are very much an SSIS novice. so let me approach this in anothe way. 1. An SSIS package can be created that would handle your problem. The trick is to use the MERGE statement that is available in T-SQL. You can also use a pair of SQL statements such as those that you provided in your most recent post. Howver, your SQL code may be impacting our performance somewhat because you are using the IN (SELECT....) instead of using the INNER JOIN approach to determining which rows to update. 2. A straight forward bulk load is always going to be the fastest choice; however, since you need to UPDATE some rows and INSERT others, that option isn't available to you,so there's no point in worrying about that. ;-) That being said, though, if this is your first SSIS package, I would suggest that you create a temp table into which you the INSERT both updated versions of the existing rows and then the new rows (with the IDENTITY INSERT ON, so that you can keep the IDs of the existing rows), and then truncate the existing table and reload it from the temp table. ;-)
