There is a natural limit to how many dataflows you can run in parallel in SSIS. Regardless of whether your limit is on the source or destination side, you will eventually reach those limits.
You might have set up all your package orchestration in a way that made perfect sense at that time, but over time, some tables grow faster than expected and others don’t grow at all. Due to foreign key relationships, you may not be able simply to shuffle the dataflow tasks around to maximize throughput. Manual reengineering along these lines would potentially be very time consuming, and even worse, the result would be obsolete shortly thereafter.

This session is about using the Business Intelligence Markup Language (Biml) to monitor and control your orchestration patterns. By automatically analyzing the results in ETL logs, we’ll be able to automate our staging orchestration!

Prerequisites: Good understanding of dataflows with SSIS, especially with higher volumes of data
ETL   SSIS   tuning