RE: Seeking a faster response to Where EXISTS or WHERE IN

I have created indexes and iNav has zero suggestions.



I have not tried the join. Duh. I’ll try that next.



Thanks



Darryl



From: Paul Therrien [mailto:paultherrien@xxxxxxxxxxxxxxxxxx]
Sent: Thursday, December 22, 2016 9:49 AM
To: RPG programming on the IBM i (AS/400 and iSeries)
Cc: D Freinkel (a4g)
Subject: Re: Seeking a faster response to Where EXISTS or WHERE IN



how about:

select a.* from filea a

join (select disinct field_aa from from file_B) b

on a.field_a = b.field_aa

(syntax may not be exact)

This might be quicker.



Paul





On 2016-12-22 09:22, D Freinkel (a4g) wrote:

I am doing extracts from files based on records in file A



Select * from FILE_A where field_A in (Select FIELD_AA from FILE_B).



Or



Select * from FILE_A where field_A exists i (Select FIELD_AA from FILE_B
where FIELD_A = FIELD_B).



File_A and FILE_B both have about a million records.



Problem:

When running the program it takes a few minutes to build the sub-select.



Using CHAINS is instant, but I would prefer to use SQL.



Can anyone help with performance improvements?



TIA



Darryl Freinkel