Date: Fri, 3 Sep 2021 13:56:16 -0400
From: Alvaro Herrera <alvherre@alvh.no-ip.org>
To: Andres Freund <andres@anarazel.de>
Cc: Kyotaro Horiguchi <horikyota.ntt@gmail.com>,
	masao.fujii@oss.nttdata.com, pgsql-hackers@lists.postgresql.org,
	bossartn@amazon.com, mengjuan.cmj@alibaba-inc.com,
	Jakub.Wartak@tomtom.com
Subject: Re: prevent immature WAL streaming
Message-ID: <202109031756.lrm45nalqy56@alvherre.pgsql>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
In-Reply-To: <20210903174646.dakp7nn6hmoht45z@alap3.anarazel.de>
Archived-At: 
 <https://www.postgresql.org/message-id/202109031756.lrm45nalqy56%40alvherre.pgsql>
Precedence: bulk

On 2021-Sep-03, Andres Freund wrote:

> Hi,
> 
> On 2021-09-03 12:55:23 -0400, Alvaro Herrera wrote:
> > Oh, but of course we can't modify XLogReaderState in backbranches to add
> > the new struct member abortedContrecordPtr (or whatever we end up naming
> > that.)
> 
> Why is that? Afaict it's always allocated via XLogReaderAllocate(), so adding
> a new field at the end should be fine?

Hmm, true, that works.

> That said, I'm worried that this stuff is too complicated to get right in the
> backbranches. I suspect letting it stew in master for a while before
> backpatching would be a good move.

Sure, we can put it in master now and backpatch before the November
minors if everything goes well.

-- 
Álvaro Herrera           39°49'30"S 73°17'W  —  https://www.EnterpriseDB.com/