Would this apply to copying data from one cluster to another, assuming I do a rolling drain and shutdown ? Thanks Le 9 janv. 2012 16:32, "Brandon Williams" <dri...@gmail.com> a écrit :
> On Mon, Jan 9, 2012 at 9:14 AM, Brian O'Neill <b...@alumni.brown.edu> > wrote: > > > > What is the fastest way to copy a column family? > > We were headed down the map/reduce path, but that seems silly. > > Any file level mechanisms for this? > > Copy all the sstables 1:1 renaming them to the new CF name. Then > create the schema for the CF. > > -Brandon >