Hi, I would do a single capture to save CPU(s) on source side and use 2 different propagations with filtering rules to appropriate destinations. --romas On Wed, Sep 17, 2008 at 12:40 PM, Stalin <stalinsk@xxxxxxxxx> wrote: > All, > > Wondering if anyone could chime in with the experience in implementing > streams. We have a table wherein each subset of data needs to be > replicated to different databases for e.g rows of type x to database > A and rows of type y to database B and so on.