select distinct snapnum, FOF1 from MDPL.AvailHalos where FOF1 is not null
gets this query plan:
The query plan used to run this query: -------------------------------------------------- CALL paquExec('SELECT snapnum AS snapnum,FOF1 AS FOF1 FROM MDPL.AvailHalos WHERE ( FOF1 is not null ) ', 'aggregation_tmp_68800777')-- USE spider_tmp_shard-- SET @i=0-- CREATE TABLE multidark_user_kristin.2013-10-01T11:18:03:0401 ENGINE=MyISAM SELECT DISTINCT @i:=@i+1 AS row_id, snapnum,FOF1
FROM aggregation_tmp_68800777 -- CALL paquDropTmp('aggregation_tmp_68800777')
Looks to me, as if PaQu only pays attention to the DISTINCT-keyword after adding unique row ids, so it has no effect.
The DISTINCT keyword has no effect.
A query like this:
select distinct snapnum, FOF1 from MDPL.AvailHalos where FOF1 is not null
gets this query plan: The query plan used to run this query: -------------------------------------------------- CALL paquExec('SELECT snapnum AS
snapnum
,FOF1 ASFOF1
FROM MDPL.AvailHalos WHERE ( FOF1 is not null ) ', 'aggregation_tmp_68800777')-- USE spider_tmp_shard-- SET @i=0-- CREATE TABLE multidark_user_kristin.2013-10-01T11:18:03:0401
ENGINE=MyISAM SELECT DISTINCT @i:=@i+1 ASrow_id
,snapnum
,FOF1
FROMaggregation_tmp_68800777
-- CALL paquDropTmp('aggregation_tmp_68800777')Looks to me, as if PaQu only pays attention to the DISTINCT-keyword after adding unique row ids, so it has no effect.