Subject: Re: [HACKERS] logical decoding of two-phase transactions
To: Andres Freund <andres@anarazel.de>,
 Tomas Vondra <tomas.vondra@2ndquadrant.com>
Cc: Nikhil Sontakke <nikhils@2ndquadrant.com>,
 Craig Ringer <craig@2ndquadrant.com>,
 Peter Eisentraut <peter.eisentraut@2ndquadrant.com>,
 Simon Riggs <simon@2ndquadrant.com>, Sokolov Yura
 <y.sokolov@postgrespro.ru>, Stas Kelvich <s.kelvich@postgrespro.ru>,
 Dmitry Dolgov <9erthalion6@gmail.com>, Tom Lane <tgl@sss.pgh.pa.us>,
 Masahiko Sawada <sawada.mshk@gmail.com>,
 PostgreSQL Hackers <pgsql-hackers@postgresql.org>,
 Robert Haas <robertmhaas@gmail.com>
References: 
 <CAMGcDxc9p7envO8t+29j=NdQXHoba6P1btfU0d3Xeiz7zT-Mvw@mail.gmail.com>
 <20180212162040.rsnlnysmf7n4yo7i@alap3.anarazel.de>
 <CAMGcDxeViP+R-OL7QhzUV9eKCVjURobuY1Zijik4Ay_Ddwo4Cg@mail.gmail.com>
 <20180302005343.zv6nsqimcujzjrcd@alap3.anarazel.de>
 <CAMsr+YEb-f1oG+1wGji_sgswRsVQYtVXuTbVH=nnY6thMfiCqA@mail.gmail.com>
 <CAMGcDxcD2fWqUwiPWGL_d-ie-CoGTVFAsgQvQ91H93q2Jc-mhA@mail.gmail.com>
 <CAMGcDxf9HHmDOti4-DSQTtGo48R4OMki+BktPn2uTOtGr4XomQ@mail.gmail.com>
 <c167340c-0751-d54b-0f32-fa83fd414863@2ndquadrant.com>
 <CAMGcDxcGtRXtF1r3VqddzHba+BTm7q8wbgEsyhmOFuDckMTZFQ@mail.gmail.com>
 <dd5a9cb7-bb10-cb5c-834e-08dc3ffc057f@2ndquadrant.com>
 <20180329215815.36xxakgujx4luo4m@alap3.anarazel.de>
From: Petr Jelinek <petr.jelinek@2ndquadrant.com>
Openpgp: preference=signencrypt
Autocrypt: addr=petr.jelinek@2ndquadrant.com; keydata=
 xsFNBFawwSABEAC45Vua8rSggpM2kkMMmoOHz51a+fi3OXFfnrStVh80pM2/J9nSzjrIKQKf
 H5PtRnoYzSKnoUDoLOkLwDYitK3HzUUWXma+AWAqef2Y0GSZ7A3jJkD+JHOfkpS1CIpkEB/e
 WjA8Wqegwzc/OEE5111/l0Q1e6jxgZI86h8MMnm35Of1Y6Lz4kqWCyxHhaTJKTCkq66ICQ2C
 VJkbvuUoEE+Hr/I38dPHTAqIlCnycVnJPLHBQ/UYGHR/USF8qI/qJJx6tJcNAEqz69QZ/VlY
 31UyJHdYe5NtBVz+rYZCOVWfpe3crbHuq9FjAPT+xcZefdxridmMKuW8eBmmXI9iZ3EG+X8J
 IsdLWL8QoqW0lnO1otYz1QDjdOvbEpq+u5/TAU8pgtaYH2J+6/GaQ6hMLyLBkaIAtnqlLCzq
 pxXcxMsFT14P5/rZA7EbBI8RS1ZIWJBlS2oO+dtVuENLJUC0SpptaRyn/uyOIopt7hSmBekL
 SzFnet7P9qofFJY+fQEz84v0G1gvwonHGHiCGgLpbMdPtnfPzda1aiTURvvOD7w/RmX1/ogk
 TqSGO4ejVNO/28QYouxmxzy8jbbIEiLCbHpLR/M/TvYHHPURxcx+MxVl3G+d/4Ihy4abxOdW
 pe9zLIcVMvqiUFQO28/79Gfhb7G4K+L21ORB82QYkzm3M7ebXQARAQABzSNQZXRyIEplbMOt
 bmVrIDxwam1vZG9zQHBqbW9kb3MubmV0PsLBgAQTAQoAKgIbAwUJBaOagAULCQgHAwUVCgkI
 CwUWAgMBAAIeAQIXgAUCVrDBqwIZAQAKCRBci2l0giH4y/QJEACfiLHyvxlwMvo3Y88Qa9QI
 i7Hu3ZW+xxojojdRd7do0nBnf2a4YstD9u6ZOLJx4f5Ijb/8LxY10NgLugmfQ7sw2W8ui6wT
 NiAeNHpr4wpT3IV+V53psOMRYPdq2zNTLO4iNxqPFPgNufR/m4FqcGe0Wo7U0wd1p0MErXTg
 EXJDVmjPDkYnzE9ejSCOxF57+o3xs6+eW9Al2m3nqP8JCzpauCr0gqpQzCaw0VZ05aLL0t9E
 B0fdm90+ydVe10Wg+I348005smuSSckBCHZqUfjjWvjwE6v/6dZ3/50ycAbi6cTe7Fo4kOSO
 sXBFXkzYGWYrEFPkzK0zDH65D98gL1+l9uokCR2t/OYVwudbUCV3cS3xRd0eFgHbXuzV+aZc
 jRjqkPfpCA2loKtrX2U739mpVSHnpIfSBf48fLg7O5dVn5ZCrydGkEzgvyynPo7IikQrIA02
 91IoSp4RisKsJpaXHwHu6DH2rDCbYFumAMgXk2osLFX6VMPoSWKVRkr87A0VLafCtwaqr6Hy
 sFE6Lmu7jP4/BDr+9A2MXK+v0yYrXqqsZN6V9856ywlZrWK3MB6jc2bKpKS2YshneKxpitNV
 0ITs2Li0OlAXOEHE0eXwI93POgLoyH+8JiNbem3orw3Mnc+7XpIY86AejwUiYl8T7EO/2tIx
 HDaXLK3BPNOdH87BTQRWsMEgARAA6Z3NrDLdcFF/P3TD/F7WQtMALEEKn5TXSF/K2VHt21GD
 8tGY1DZhbxjLuQJeheY/2L4F4Y5WCkAoUdBXQs7nIauN+63EL88/qPdH6f3+/HZ+cmPG1E5o
 dBABCkC42eGcI5O8/9TAaFQcSzg86pdAE8v927tnc6aH4W4/Mj7RJNlQiZaz2ngCC9mr5uu+
 CysOTfKCS0llNUkExZOJzZsQU3zDkcg3+xmgtHkCjglJ006KuytHj55HR6TzaKaAW8coHGoS
 OI/BQ9KAzvlj+32/zS8FSISZSTY3spMpraADsYklejCeJOmqQ3nTrCtwlX3hTSggfr4eelKA
 hdZKs1rIimNBtX90jJc0u2A/GmC959aTSrCUcfaPRweKUuexg7jxuhXyg9br2GscD2G6W7EV
 c+3s4dOHgoJFVMBBIqTtZviaNJ63IsZIFTYJzGKKMHOioZ3naoBVav0w63waDzj82nPYMIPj
 TxnClKiecuBQQPzMaG0jTmK0lis6rO0nIXrCwca8e/3Xzvi7sykOdcz9t4ef8HrDCCkG0yDO
 o8qHy1e7EtmJD7kWffIXZMt4TZVxpiOqgWtv+heJAfmqzr8pkNLUubmgvNdnakXJOJT41Lui
 50C5OhpU9mfFCIDOeZXTQLsHA4KYaL8W/qcjv/YDmvpJtnK24y6tByWuTu7r/SMAEQEAAcLB
 ZQQYAQoADwUCVrDBIAIbDAUJBaOagAAKCRBci2l0giH4yyWpD/91jd4BMN2CDYqLtU7kaoAa
 aXnRVpf0z+R8rJ4LZJX1UNbHjsdClMQLfdBuZeDsdYar4iMfwScz6fiCiRwaRpr3XT6znaM0
 wba6fjzEqVSHr3WB1jrb/TMiNaNgQJaRPhI5C7VsLuPrv5/L8gFmyjl712q7F3busGFBKrjK
 qTESMu+sHNvVeC3QyBFCe2fudbAJdaaRsUoTbAZYiowVoQHOow8eBoWY+RzQoYGuw1fls5Ck
 85N73pYHXzjfk1s49sDKTYI263h9wbGGA234biHRI3NQN6sSYWRr/u5GOwTm/tzyadRj0pnZ
 S6KyLSEL4qx8jiyniD/XLoftlvR4YWQ0BXveEElx2n3TThLL1cVtOI4qE9BZKm32euBqwLta
 k9Jc659hmZB9E9aWLo8KcLjngwqSZupoHRPZ3yAlgZkpDoTIZpMCGxGEuuxgnosRbNpOJ1fE
 nNXenYYfi87tHRUCRKeHhoKkfGZEgC9zv9Y97j75npqV2tQyh4fNQT1tGdBigtvzI4nHM1/1
 PLJtJBdRHucP76vUH9RlP9EE/Sg5DFQqsfSYPqxLW9r+ZfSEeT0APdz7TQC9MzNzUhyXvA5a
 RQQCRMXKbGaYhKkYEYiYOGqtTxN+PmyvNUBmFEcagBX2TXqEkgnI6zOnN13mB4SihJ2lne2r
 9RNiPYs/hmRxYg==
Message-ID: <ff89e6f0-baae-fd4f-6565-13e7c2f23b90@2ndquadrant.com>
Date: Fri, 30 Mar 2018 00:30:40 +0200
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:52.0) Gecko/20100101
 Thunderbird/52.6.0
MIME-Version: 1.0
In-Reply-To: <20180329215815.36xxakgujx4luo4m@alap3.anarazel.de>
Content-Type: text/plain; charset=utf-8
Content-Language: en-US
Content-Transfer-Encoding: 8bit
Precedence: bulk

On 29/03/18 23:58, Andres Freund wrote:
> On 2018-03-29 23:52:18 +0200, Tomas Vondra wrote:
>>> I have added details about this in src/backend/storage/lmgr/README as
>>> suggested by you.
>>>
>>
>> Thanks. I think the README is a good start, but I think we also need to
>> improve the comments, which is usually more detailed than the README.
>> For example, it's not quite acceptable that LogicalLockTransaction and
>> LogicalUnlockTransaction have about no comments, especially when it's
>> meant to be public API for decoding plugins.
> 
> FWIW, for me that's ground to not accept the feature. Burdening output
> plugins with this will make their development painful (because they'll
> have to adapt regularly) and correctness doubful (there's nothing
> checking for the lock being skipped).  Another way needs to be found.
> 

I have to agree with Andres here. It's also visible in the latter
patches. The pgoutput patch forgets to call these new APIs completely.
The test_decoding calls them, but it does so even when it's processing
changes for committed transaction.. I think that should be avoided as it
means potentially doing SLRU lookup for every change. So doing it right
is indeed not easy.

I as wondering how to hide this. Best idea I had so far would be to put
it in heap_beginscan (and index_beginscan given that catalog scans use
is as well) behind some condition. That would also improve performance
because locking would not need to happen for syscache hits. The problem
is however how to inform the heap_beginscan about the fact that we are
in 2PC decoding. We definitely don't want to change all the scan apis
for this. I wonder if we could add some kind of property to Snapshot
which would indicate this fact - logical decoding is using it's own
snapshots it could inject the information about being inside the 2PC
decoding.

-- 
  Petr Jelinek                  http://www.2ndQuadrant.com/
  PostgreSQL Development, 24x7 Support, Training & Services