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: Dilip Kumar <dilipbalaut@gmail.com>
Cc: 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: 2022-03-19T18:33:10Z
Lists: pgsql-hackers 
On Fri, Mar 18, 2022 at 12:39 AM Dilip Kumar <dilipbalaut@gmail.com> wrote:
> > One question that occurred to me when looking this over is whether, or
> > why, it's safe against concurrent smgr invalidations.
>
> We are only accessing the smgr of the source database and the
> destination database.  And there is no one else that can be connected
> to the source db and the destination db is not visible to anyone.  So
> do we really need to worry about the concurrent smgr invalidation?
> What am I missing?

A sinval reset can occur at any moment due to an overflow of the
queue. That acts as a universal reset of everything. So you can't
reason on the basis of what somebody might be sending.

-- 
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.