I read an article about partitioning in Hive which said that "Partitioning in Hive distributes execution load horizontally".
I'm curious about what horizontal (as opposed to vertical) load distribution means - if that makes any sense. And can load be distributed "vertically"?
In hive when you partition a table that means you create different directories for same kind of data. for eg. you have a data which have city related details and you want detail related to a city lets assume pune.
without partition: you will have to scan whole table for only pune related details.that means vertically whole table.
with partition : if you have created partition on cities then the data will be distributed over the partitions means for eg pune related data goes to pune partition and when you try to access the the data related to pune then you dont have to scan whole table.
