set hive.fetch.task.conversion=more; DESCRIBE FUNCTION to_unix_timestamp; DESCRIBE FUNCTION EXTENDED to_unix_timestamp; create table oneline(key int, value string); load data local inpath '../../data/files/things.txt' into table oneline; SELECT '2009-03-20 11:30:01', to_unix_timestamp('2009-03-20 11:30:01') FROM oneline; SELECT '2009-03-20', to_unix_timestamp('2009-03-20', 'yyyy-MM-dd') FROM oneline; SELECT '2009 Mar 20 11:30:01 am', to_unix_timestamp('2009 Mar 20 11:30:01 am', 'yyyy MMM dd h:mm:ss a') FROM oneline; SELECT 'random_string', to_unix_timestamp('random_string') FROM oneline; -- PPD explain select * from (select * from src) a where unix_timestamp(a.key) > 10; explain select * from (select * from src) a where to_unix_timestamp(a.key) > 10;