I have for example a dtsx package, with multiple data flows, and i wish to know by the 'end' of the process how many rows were retrieved/analyzed from all the existing sources there. I want to know even if the package fails how many rows were more or less processed. Is there any smart solution to retrieve this rowcount?

you can get number or rows on each data path of data flow with row count transformation into a variable, and then log that variable value. And for find BAD Rows in data flow you can CONFIGURE ERROR OUTPUT on source/transformation/destination and redirect bad data rows to another destination for example flat file destination and after running the package you can look at that flat file to find bad data rows and troubleshoot them, also you can use row count transformations on error data path to fetch number of bad data rows into variables and log them too.http://www.rad.pasfu.com My Submitted sessions at sqlbits.com

let's say i have a Dataflow with a source A and a destination B, what i want to count is all the rows that were extracted from A, then i have 2 possible situations: 1. A -> B goes with succes so i can use a variable and a rowcount 2. A -> B goes with failure.. but i still want to know how many rows where extracted until that moment the point is that i have a master dtsx which runs a lot of packages which inside them have a lot of dataflows, so to customize every each package, every each dataflow, each in a very specific way sounds like there must be another solution :)

Your only other solution is to turn on SSIS Package Logging with the OnPipelineRowsSent event logged. This will generate a LARGE volume of logging events, as each component will report the number of rows it processes as it processes them. Talk to me now on

If you want to just fetch number of rows and you don't need to fetch exact bad data rows exactly, and in other hand you don't want to change child packages and each data flow separately, then Todd's suggestion is your only way. But if you want to fetch error data rows exactly you should CONFIGURE ERROR OUTPUT and there isn't any other way , you should do it in each data flow which you want to monitor bad data rows.http://www.rad.pasfu.com My Submitted sessions at sqlbits.com