Postgres Pro
Downloads
Home   >   mailing lists

Re: Materializing the relation - Mailing list pgsql-general

From Shoaib Mir
Subject Re: Materializing the relation
Date
Msg-id bf54be870706110844p9180055n64ff1a3dd7776c28@mail.gmail.com
Whole thread Raw
In response to Materializing the relation  ("rupesh bajaj" <rupesh.bajaj@gmail.com>)
Responses Re: Materializing the relation  (Gregory Stark <stark@enterprisedb.com>)
List pgsql-general
Tree view
From /src/include/utils/tuplestore.h

"The materialize shields the sort from the need to do mark/restore and thereby allows it to perform its final merge pass on-the-fly; while the materialize itself is normally cheap since it won't spill to disk unless the number of tuples with equal key values exceeds work_mem"

--
Shoaib Mir
EnterpriseDB (www.enterprisedb.com)

On 6/11/07, rupesh bajaj < rupesh.bajaj@gmail.com> wrote:
Hi,

What is the meaning of 'materializing' a relation after sequential scan?



explain select * from tb1, tb2  where tb1.c1 = tb2.c2;

                           QUERY PLAN
----------------------------------------------------------------
 Nested Loop  (cost=1.03..2.34 rows=3 width=24)
   Join Filter: (tb1.c1 = tb2.c1)
   ->  Seq Scan on tb2  (cost=0.00..1.04 rows=4 width=12)
   ->  Materialize  (cost=1.03..1.06 rows=3 width=12)
         ->  Seq Scan on tb1  (cost= 0.00..1.03 rows=3 width=12)


In this above plan, what does 'Material' mean?

Thanks,
Rupesh

pgsql-general by date:

Previous
From: "rupesh bajaj"
Date:
Subject: Materializing the relation
Next
From: "Filip Rembiałkowski"
Date:
Subject: Re: When should I worry?