From: Joe Carlson <jwcarlson@lbl.gov>
To: pgsql-performance@postgresql.org
Subject: Row level security policy policy versus SQL constraints. Any
 performance difference?
Message-ID: <59E66BA8.6060009@lbl.gov>
Date: Tue, 17 Oct 2017 13:44:24 -0700
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:38.0) Gecko/20100101
 Thunderbird/38.6.0
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Transfer-Encoding: 7bit
Precedence: bulk
Sender: pgsql-performance-owner@postgresql.org

Hello.

I have not used row level security policies in the past but am 
considering using them for a project in which I would like to restrict 
the set returned in a query based on specific fields. This is more as a 
convenience issue (for me) rather than a security issue.

What I was wondering is what is the performance differences between a 
row level security implementation:

CREATE POLICY <policy name> ON <table> TO <role> USING 
(<field>=ANY(<values>));
<series of selects>
DROP POLICY <policy name>

and an implementation where I add on the constraints as part of each 
select statement:

SELECT <whatever> FROM <table> WHERE <constraints> AND <field>=ANY(<values>)

In my (admittedly small) number of EXPLAINs I've looked at, it appears 
that the policy logic is added to the SELECT statement as a constraint. 
So I would not expect any fundamental performance difference in the 2 
different forms.

Is this true? Or is there some extra behind-the-scenes things to be 
aware of? Can there be excessive overhead from the CREATE/DROP POLICY 
statements?

Thanks,

Joe


-- 
Sent via pgsql-performance mailing list (pgsql-performance@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-performance