-Re: How to write Block of queries in Hive?
Aniket Mokashi 2012-01-05, 06:13
[moving discussion to hive user list]
I would suggest you to send your discussion to hive user list in order to
reach a broader audience.
As per my understanding, in the query- map_script and reduce_script are
custom scripts that run as a streaming jobs. You are asking hive to run
map_script as mapper job on 3 columns to generate 3 new values- c1, c2, c3.
After this, hive will sort your records on c1 and c2 and distribute them to
reducers based on c3 values. 'reduce_scripts' will consume these 3 records
and generate 2 records to store in pv_users_reduced.
Hope it helps.
On Wed, Jan 4, 2012 at 8:55 PM, Bhavesh Shah <[EMAIL PROTECTED]>wrote:
> I am new to hive. I want to write block of queries in Hive so that one
> query give result to another one like in SQL.
> I have also visited one link given below:
> In above link I am looking for functions but I get below one and I dont
> understand following things:
> USING 'map_script'USING 'reduce_script'
> in following block:
> FROM (
> FROM pv_users
> MAP ( pv_users.userid, pv_users.date )
> USING 'map_script'
> AS c1, c2, c3
> DISTRIBUTE BY c2
> SORT BY c2, c1) map_output
> INSERT OVERWRITE TABLE pv_users_reduced
> REDUCE ( map_output.c1, map_output.c2, map_output.c3 )
> USING 'reduce_script'
> AS date, count;
> Pls can anyone tell what is the use of scripts and how to write block
> of queries in hive?
> Bhavesh Shah