We have got a SSIS package loading millions of rows to Master Data Services database. Basically, it is trying to load data for a particular target model and particular version. We have to load it for 7 different versions for a model. By individually running the package for every version, the package totally takes 23 hours. But, if we put a For Loop Container to avoid manual intervention and execute the same set of tasks, it is taking 36 hours. Is it something related to the package being held in memory for long time or the connection being held for long time in the target server ? Can you please guide in improving the performance of the package ?Venkataraman R

It's very hard to say what might be causing the time difference. I'd suggest using Package Logging to look at timings of events occurring in the package to determine what part(s) are slower in the looped version. If you're concerned that it's lingering memory issues, then divide your package into a parent/child set, and run the child package inside the loop with an Execute Package Task, where you run it out-of-process. That will start the child with a new chunk of clean memory. Talk to me now on