Horton

Re: [Proposal] Fully WAL logged CREATE DATABASE - No Checkpoints

Robert Haas <robertmhaas@gmail.com>

View thread View on postgr.es
From: Robert Haas <robertmhaas@gmail.com>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: Andres Freund <andres@anarazel.de>, Dilip Kumar <dilipbalaut@gmail.com>, Ashutosh Sharma <ashu.coek88@gmail.com>, Maciek Sakrejda <m.sakrejda@gmail.com>, Bruce Momjian <bruce@momjian.us>, Alvaro Herrera <alvherre@alvh.no-ip.org>, Andrew Dunstan <andrew@dunslane.net>, Heikki Linnakangas <hlinnaka@iki.fi>, PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>
Date: 2022-03-29T18:24:40Z
Lists: pgsql-hackers 
On Tue, Mar 29, 2022 at 2:17 PM Tom Lane <tgl@sss.pgh.pa.us> wrote:
> Robert Haas <robertmhaas@gmail.com> writes:
> > On Tue, Mar 29, 2022 at 1:53 PM Tom Lane <tgl@sss.pgh.pa.us> wrote:
> >> That test script is expecting glibc-like laxness of switch
> >> parsing.  Put the switches before the non-switch arguments.
>
> > I just did that. :-)
>
> Yup, you pushed while I was typing.
>
> FWIW, I don't think it's "Windows" enforcing this, it's our own
> src/port/getopt[_long].c.  If there were a well-defined spec
> for what glibc does with such cases, it might be interesting to
> try to make our version bug-compatible with theirs.  But AFAIK
> it's some random algorithm that they probably feel at liberty
> to change.

I guess that characterization surprises me. The man page for
getopt_long() says this, and has for a long time at least on systems
I've used:

ENVIRONMENT
     POSIXLY_CORRECT  If set, option processing stops when the first non-
                      option is found and a leading `-' or `+' in the
                      optstring is ignored.

And also this:

BUGS
     The argv argument is not really const as its elements may be permuted
     (unless POSIXLY_CORRECT is set).

Doesn't that make it pretty clear what the GNU version is doing?

-- 
Robert Haas
EDB: http://www.enterprisedb.com

Commits

  1. When using the WAL-logged CREATE DATABASE strategy, bulk extend.

  2. Avoid using a fake relcache entry to own an SmgrRelation.

  3. Fix data-corruption hazard in WAL-logged CREATE DATABASE.

  4. initdb: When running CREATE DATABASE, use STRATEGY = WAL_COPY.

  5. Simplify a needlessly-complicated regular expression.

  6. In 020_createdb.pl, change order of command-line arguments.

  7. Add new block-by-block strategy for CREATE DATABASE.

  8. Fix replay of create database records on standby

  9. Refactor code for reading and writing relation map files.

  10. Replace RelationOpenSmgr() with RelationGetSmgr().

  11. Refactor the fsync queue for wider use.