Koen Verbeeck (4/30/2014)
If they run in paralllel, they are independent of each other.
However, this is not 100% true. Say you have 7 tasks which can run in parallel.
But only 4 of time can run at the same time (due to the number of processors available). If one crashes, the other 3 will continue to run. But no new data flow tasks will be started since the parent container (package, sequence container, for each loop, whatever) has already failed.
This makes sense