Hi All, I'm sure this question has been answered a million times in pieces but I can't find one succinct answer to what the ideal settings should be. I have a package that does the following with precedence constraints: 1. Truncate Table 2. Drop Indexes 3. Load table 4. Index Table 5. Update Table (using a proc inside execute sql task) If I run the proc natively in SSMS < 2 mins, if I run just the execute sql task ~ 2mins, if I run the whole package the proc can take up to 8mins. There are no other processes running on the server, I have tweaked NOFMT, NOCOUNT and Transaction type of a) the task b) the package c) all tasks plus package, but nothing is making the difference Suggestions please, getting to the point where if someone says "don't bother using SSIS its bad at this because of XYZ" then I'll listen. I hope not..... Thanks David

Before anyone asks "why do the update statement afterwards and not in the dataflow?", it would end up being a complex lookup on date between X and Y join which I consider to be inferior in SSIS, or it would be an enormous lookup

I think that each SQL Task validates against the source, opening a new connection. Open a profiler to check it. Vctor M. Snchez Garca (ES) (BI) Hope this help. Please vote if you find this posting was helpful. if this is an answer to your question, please mark it. http://bifase.blogspot.com | http://twitter.com/atharky