That’s “Block Change Tracking” if you’re not familiar with the acronym – a feature that appeared in the 10g with its own background process (CTWR – change tracking writer) to help rman do faster incremental backups on very large datafiles.
It’s very useful if you’re using large tablespaces (i.e. the “one file per tablespace”) option – but, as with many things Oracle produces, when you push the limits or move into combinations of features odd errors start to appear.
Saibabu Devabhaktuni has written a blog note about using BCT with physical standby databases – a relatively new enhancement to BCT, but possibly one you need to be cautious about for reasons he explains.
