Content-Type: multipart/mixed; boundary="------------VQdy8WI6kVmSqhrrbF0kk40j"
Message-ID: <6b5c342d-45b8-4e01-8ff1-3032b80564ee@gmail.com>
Date: Tue, 8 Oct 2024 12:00:27 +0700
MIME-Version: 1.0
User-Agent: Mozilla Thunderbird
Subject: Re: MergeJoin beats HashJoin in the case of multiple hash clauses
From: Andrei Lepikhov <lepihov@gmail.com>
To: Tomas Vondra <tomas.vondra@enterprisedb.com>,
 Andy Fan <zhihui.fan1213@gmail.com>
Cc: PostgreSQL-development <pgsql-hackers@postgresql.org>,
 Tom Lane <tgl@sss.pgh.pa.us>, "a.rybakina" <a.rybakina@postgrespro.ru>
References: <52257607-57f6-850d-399a-ec33a654457b@postgrespro.ru>
 <CAKU4AWoKiogNir=ESrm6F9pUwYFDfmsoNG9w9k1MG3VUbKM0Dw@mail.gmail.com>
 <84114cba-97e9-49f8-952f-04cf321125c1@app.fastmail.com>
 <1de52b5f-e4a6-5729-eeef-51d524534665@enterprisedb.com>
 <c99dc067-9fec-4649-a0f5-62a771404972@gmail.com>
Content-Language: en-AU
In-Reply-To: <c99dc067-9fec-4649-a0f5-62a771404972@gmail.com>
Archived-At: <https://www.postgresql.org/message-id/6b5c342d-45b8-4e01-8ff1-3032b80564ee%40gmail.com>
Precedence: bulk

This is a multi-part message in MIME format.
--------------VQdy8WI6kVmSqhrrbF0kk40j
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit

On 7/8/24 19:45, Andrei Lepikhov wrote:
> On 3/11/2023 23:43, Tomas Vondra wrote:
>> On 9/11/23 10:04, Lepikhov Andrei wrote:
>>   * Determine bucketsize fraction and MCV frequency for the inner
>>   * relation. We use the smallest bucketsize or MCV frequency estimated
>>   * for any individual hashclause; this is undoubtedly conservative.
>>
>> I'm sure this may lead to inflated cost for "good" cases (where the
>> actual bucket size really is a product), which may push the optimizer to
>> use the less efficient/slower join method.
> Yes, It was contradictory idea, though.
>> IMHO the only principled way forward is to get a better ndistinct
>> estimate (which this implicitly does), perhaps by using extended
>> statistics. I haven't tried, but I guess it'd need to extract the
>> clauses for the inner side, and call estimate_num_groups() on it.
> And I've done it. Sorry for so long response. This patch employs of 
> extended statistics for estimation of the HashJoin bucket_size. In 
> addition, I describe the idea in more convenient form here [1].
> Obviously, it needs the only ndistinct to make a prediction that allows 
> to reduce computational cost of this statistic.
Minor change to make cfbot happier.

-- 
regards, Andrei Lepikhov

--------------VQdy8WI6kVmSqhrrbF0kk40j
Content-Type: text/x-patch; charset=UTF-8;
 name="v1-0001-Use-extended-statistics-for-precise-estimation-of.patch"
Content-Disposition: attachment;
 filename*0="v1-0001-Use-extended-statistics-for-precise-estimation-of.pa";
 filename*1="tch"
Content-Transfer-Encoding: base64
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--------------VQdy8WI6kVmSqhrrbF0kk40j--