Horton

Re: recovery modules

Michael Paquier <michael@paquier.xyz>

View thread View on postgr.es
From: Michael Paquier <michael@paquier.xyz>
To: Nathan Bossart <nathandbossart@gmail.com>
Cc: Andres Freund <andres@anarazel.de>, pgsql-hackers@postgresql.org
Date: 2023-01-30T23:13:11Z
Lists: pgsql-hackers 
On Mon, Jan 30, 2023 at 12:04:22PM -0800, Nathan Bossart wrote:
> On Mon, Jan 30, 2023 at 11:48:10AM -0800, Andres Freund wrote:
>> I don't think _PG_archive_module_init() should actually allocate a memory
>> context and do other similar initializations. Instead it should just return
>> 'const ArchiveModuleCallbacks*', typically a single line.
>> 
>> Allocations etc should happen in one of the callbacks. That way we can
>> actually have multiple instances of a module.
> 
> I think we'd need to invent a startup callback for archive modules for this
> to work, but that's easy enough.

If you don't return some (void *) pointing to a private area that
would be stored by the backend, allocated as part of the loading path,
I agree that an extra callback is what makes the most sense,
presumably called around the beginning of PgArchiverMain().  Doing
this kind of one-time action in the file callback woud be weird..
--
Michael

Commits

  1. Redesign archive modules

  2. Remove uses of AssertVariableIsOfType() obsoleted by f2b73c8

  3. Rename contrib module basic_archive to basic_wal_module

  4. Refactor code for restoring files via shell commands

  5. Refactor code in charge of running shell-based recovery commands