I would create separate partitions, one for each day worth of data, and then drop the partitions that are no longer needed.
On Mon, Jan 31, 2011 at 11:56 AM, Cam Bazz <camb...@gmail.com> wrote: > Hello, > > I understand there is no way to delete data stored in a table. like a > `delete from table_name` in the hive language. > All this is fine, because all the query results are inserted into > another table, and the previous data in it is overwritten. > > When we need to store data collectively in to a table, like add more > rows to it with each run, how do we do it? > > Best Regards, > -C.B. > -- Adam J. O'Donnell, Ph.D. Immunet Corporation Cell: +1 (267) 251-0070
