From: Tom Lane <tgl@sss.pgh.pa.us>
To: "Heinemann, Manfred (IMS)" <HeinemannM@imsweb.com>
cc: Fabio Pardi <f.pardi@portavita.eu>,
        "pgsql-admin@lists.postgresql.org" <pgsql-admin@lists.postgresql.org>
Subject: Re: Function search_path
In-reply-to: <b0078bdae3e04de1ba3ddaa3527c8922@THALASSA.omni.imsweb.com>
References: <db194d98f61d4aa398c1399a24326e7c@THALASSA.omni.imsweb.com>
 <5f698c34-d01d-3196-1db4-75ba8f6a37d3@portavita.eu>
 <b0078bdae3e04de1ba3ddaa3527c8922@THALASSA.omni.imsweb.com>
Comments: In-reply-to "Heinemann, Manfred (IMS)" <HeinemannM@imsweb.com>
	message dated "Fri, 16 Mar 2018 21:09:52 -0000"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <2327.1521236730.1@sss.pgh.pa.us>
Content-Transfer-Encoding: quoted-printable
Date: Fri, 16 Mar 2018 17:45:30 -0400
Message-ID: <2328.1521236730@sss.pgh.pa.us>
Precedence: bulk

"Heinemann, Manfred (IMS)" <HeinemannM@imsweb.com> writes:
> I have played around with the postgres memory settings and setting searc=
h_path on a function causes a lot more memory to be used than if no search=
_path was set.

That's a pretty broad claim with a pretty small amount of evidence
offered.

I can certainly believe that attaching a SET clause (whether for
search_path or any other GUC variable) would have an efficiency impact;
one non-obvious example is that it prevents inlining if the function is
a SQL function.  But I don't immediately see a reason for major memory
consumption from that.  I suspect what you're seeing is specific to a
particular use-case.  If you were to provide a concrete example, we
could look into what's happening.

			regards, tom lane