Horton

Re: Extensible Rmgr for Table AMs

Jeff Davis <pgsql@j-davis.com>

View thread View on postgr.es
From: Jeff Davis <pgsql@j-davis.com>
To: pgsql-hackers@postgresql.org
Cc: Andres Freund <andres@anarazel.de>, Amit Kapila <amit.kapila16@gmail.com>, Tom Lane <tgl@sss.pgh.pa.us>, Simon Riggs <simon@2ndQuadrant.com>
Date: 2021-11-10T01:38:29Z
Lists: pgsql-hackers

Attachments

On Mon, 2021-11-08 at 15:36 -0800, Jeff Davis wrote:
> The attached patch (against v14, so it's easier to test columnar) is
> somewhat like a simplified version of [3] combined with refactoring
> to
> make decoding a part of the rmgr.

I created a wiki page here: 
https://wiki.postgresql.org/wiki/ExtensibleRmgr

To coordinate reservation of RmgrIds, to avoid conflicts. I don't
expect it to be a practical problem given how much work it takes to
create a new table AM that needs full WAL support, but might as well
have some transparency on how to choose a new RmgrId.

I also updated the patch to point to the wiki page in the comments, and
added in a new RM_EXPERIMENTAL_ID that can be used while an extension
is still in development. Hopefully this will prevent people reserving
lots of RmgrIds for extensions that never get released.

Regards,
	Jeff Davis

Commits

  1. Fix off-by-one error in pg_waldump, introduced in 5c279a6d350.

  2. Fix another buildfarm issue from commit 5c279a6d350.

  3. Fix warning introduced in 5c279a6d350.

  4. Custom WAL Resource Managers.

  5. Make logical decoding a part of the rmgr.