From: Tom Lane <tgl@sss.pgh.pa.us>
To: Alvaro Herrera <alvherre@alvh.no-ip.org>
cc: "Finnerty, Jim" <jfinnert@amazon.com>, Japin Li <japinli@hotmail.com>,
        Justin Pryzby <pryzby@telsasoft.com>,
        "pgsql-hackers@lists.postgresql.org"
 <pgsql-hackers@lists.postgresql.org>
Subject: Re: parse partition strategy string in gram.y
In-reply-to: <20221024161317.3ywktkv2ghqpp4as@alvherre.pgsql>
References: <20221024161317.3ywktkv2ghqpp4as@alvherre.pgsql>
Comments: In-reply-to Alvaro Herrera <alvherre@alvh.no-ip.org>
	message dated "Mon, 24 Oct 2022 18:13:17 +0200"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <3930525.1666659024.1@sss.pgh.pa.us>
Date: Mon, 24 Oct 2022 20:50:24 -0400
Message-ID: <3930526.1666659024@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/3930526.1666659024%40sss.pgh.pa.us>
Precedence: bulk

Alvaro Herrera <alvherre@alvh.no-ip.org> writes:
> On 2022-Oct-24, Finnerty, Jim wrote:
>> The advantage of hash partition bounds is that they are not
>> domain-specific, as they are for ordinary RANGE partitions, but they
>> are more flexible than MODULUS/REMAINDER partition bounds.

I'm more than a bit skeptical of that claim.  Under what
circumstances (other than a really awful hash function,
perhaps) would it make sense to not use equi-sized hash
partitions?  If you can predict that more stuff is going
to go into one partition than another, then you need to
fix your hash function, not invent more complication for
the core partitioning logic.

			regards, tom lane