Heck, you could refine it this way as well.
select pm.*
from purchasemaster pm
inner join itemmaster im
on (pm.transactionid = im.transactionid and im.itempriority = 2)
This might reduce memory usage. I am not sure which will turn out to perform better in a resource friendly environment.
Tom Veldhouse
PS There is a lot of HTML email on this list. Is this generally considered good? Or should I convert my replies to text?
----- Original Message -----
Sent: Friday, June 29, 2001 10:58 AM
Subject: Re: [GENERAL]
select pm.*
from purchasemaster pm
inner join itemmaster im
on pm.transactionid = im.transactionid
where im.itempriority = 2
Try this instead.
Make sure the appropriate indexes (on pm.transactionid, im.transactionid and im.itempriority) exist if the tables are of any appreciable size.
Tom Veldhouse
----- Original Message -----
Sent: Friday, June 29, 2001 7:40 AM
Subject: RE: [GENERAL]
i think performance is better in postgres if you rewrite this sort of query as:
select * from purchasemaster where exists (select 1 from itemmaster where itempriority = 2 and transactionid = purchasemaster.transactionid)
hope this helps,
tamsin
Hi,
I have two tables purchasemaster (Master) and ItemMaster (child) linked through TransactionID
I have a query like
select * from purchasemaster where transactionid in (select transactionid from itemmaster where itempriority = 2)
****I need this query for MS-Access, MS-SQL and Postgres. Because the application is supposed to be such that for upto 2-3 users, the application would be running on MS-Access or MSDE, while for heavier databases i.e., greater than 4 , the ideal database would be Postgres
Only the problem is ... (correct me if i am wrong, PLEEEEEEEEEEEASE) that for each tuple in purchasemaster the sub-query is re-evaluated (True or False????). What I am looking at is whether it is possible to execute the query in the sub-query just once, get its value as in "(1,2,5,23,7,90)", etc. in place of the sub-query and accordingly, execute the main query.
Thank you,
Kapil
