Subject: Re: [HACKERS] AdvanceXLInsertBuffer vs. WAL segment compressibility
To: Tomas Vondra <tomas.vondra@2ndquadrant.com>,
 Michael Paquier <michael@paquier.xyz>, Stephen Frost <sfrost@snowman.net>
References: 
 <CAB7nPqQr1D3gjx8CmHYQREidG2eeGHzx2yQ97B3t2L99GE=+kQ@mail.gmail.com>
 <5A92F0C0.7070706@anastigmatix.net>
 <C0523BD1-485D-48B5-982E-8D63E2D001B7@yesql.se>
 <5AADB130.2020206@anastigmatix.net> <4597.1521338981@sss.pgh.pa.us>
 <46FFB930-B0F3-400A-85ED-D9948C97FADF@yesql.se>
 <5AAEE00E.9050503@anastigmatix.net>
 <21C25ED8-4ED8-4AD4-B551-3E3987ECB18F@yesql.se>
 <5AAEFF11.3070903@anastigmatix.net>
 <20180326032731.GM24540@tamriel.snowman.net>
 <20180326054354.GB2759@paquier.xyz> <5ABAC668.5080203@anastigmatix.net>
 <8c00b26a-d6da-5764-e67e-9c4d8128081a@2ndquadrant.com>
Cc: Daniel Gustafsson <daniel@yesql.se>, Tom Lane <tgl@sss.pgh.pa.us>,
 Michael Paquier <michael.paquier@gmail.com>,
 Heikki Linnakangas <hlinnaka@iki.fi>, Robert Haas <robertmhaas@gmail.com>,
 Bruce Momjian <bruce@momjian.us>,
 PostgreSQL Hackers <pgsql-hackers@postgresql.org>
From: Chapman Flack <chap@anastigmatix.net>
Message-ID: <5ABD9083.5050908@anastigmatix.net>
Date: Thu, 29 Mar 2018 21:18:59 -0400
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:38.0) Gecko/20100101
 Thunderbird/38.7.0
MIME-Version: 1.0
In-Reply-To: <8c00b26a-d6da-5764-e67e-9c4d8128081a@2ndquadrant.com>
Content-Type: text/plain; charset=windows-1252
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 03/27/18 20:09, Tomas Vondra wrote:
> Not sure what's up with gitweb, but git finds it without any issue:
> 
> https://git.postgresql.org/gitweb/?p=postgresql.git;a=commit;h=2dd9322ba6eea76800b38bfea0599fbc459458f2

Thanks, that worked.

On 03/27/18 22:10, Michael Paquier wrote:
> Here you go for one example:
> https://sourceforge.net/projects/pglesslog/

So far, I have been able to study the commit pertaining to
XLP_BKP_REMOVABLE. For again some odd reason, I am striking out on
finding pglesslog code to study. Using the clone URL offered by sourceforge:

$ git clone https://git.code.sf.net/p/pglesslog/code pglesslog-code
Cloning into 'pglesslog-code'...
warning: You appear to have cloned an empty repository.
Checking connectivity... done.

and there's a Files tab, but it tells me This project has no files.

I can find 1.4.2 beta on pgFoundry, but that predates the BKP_REMOVABLE
commit.


In any case, from my study of the commit, it is hard for me to see an issue.
The code comment says: "mark the header to indicate that WAL records
beginning in this page have removable backup blocks."

In the only case where this patch will zero a header--in the unused space
following the switch record in a segment--there are no "WAL records
beginning in this page". There will not be another WAL record of any kind
until the next valid page (with valid xlp_magic xlp_tli xlp_pageaddr),
which will be at the start of the next segment, and that page will have
XLP_BKP_REMOVABLE if it ought to, and that will tell the reader what it
needs to know.

Am I overlooking something?

-Chap