Re: hashed subplan 5000x slower than two sequential operations - Mailing list pgsql-performance

From Tom Lane
Subject Re: hashed subplan 5000x slower than two sequential operations
Date
Msg-id 20170.1291839146@sss.pgh.pa.us
Whole thread Raw
In response to Re: hashed subplan 5000x slower than two sequential operations  (Shrirang Chitnis <Shrirang.Chitnis@hovservices.com>)
Responses Re: hashed subplan 5000x slower than two sequential operations
List pgsql-performance
Shrirang Chitnis <Shrirang.Chitnis@hovservices.com> writes:
> Bryce,
> The two queries are different:

I suspect the second one is a typo and not what he really wanted.

> WHERE (contexts.parent_key = 392210
>       OR contexts.context_key IN
>          (SELECT collection_data.context_key
>          FROM collection_data
>           WHERE collection_data.collection_context_key = 392210)

The only really effective way the planner knows to optimize an
"IN (sub-SELECT)" is to turn it into a semi-join, which is not possible
here because of the unrelated OR clause.  You might consider replacing
this with a UNION of two scans of "contexts".  (And yes, I know it'd be
nicer if the planner did that for you.)

            regards, tom lane

pgsql-performance by date:

Previous
From: "Marc Mamin"
Date:
Subject: Re: hashed subplan 5000x slower than two sequential operations
Next
From: "Marc Mamin"
Date:
Subject: Re: hashed subplan 5000x slower than two sequential operations