Re: [PATCH] Provide more information to filter_prepare
Amit Kapila <amit.kapila16@gmail.com>
From: Amit Kapila <amit.kapila16@gmail.com>
To: Markus Wanner <markus.wanner@enterprisedb.com>
Cc: "pgsql-hackers@postgresql.org" <pgsql-hackers@postgresql.org>,
Ajin Cherian <itsajin@gmail.com>
Date: 2021-03-29T06:23:52Z
Lists: pgsql-hackers
On Mon, Mar 29, 2021 at 11:42 AM Amit Kapila <amit.kapila16@gmail.com> wrote: > > On Thu, Mar 25, 2021 at 2:07 PM Markus Wanner > <markus.wanner@enterprisedb.com> wrote: > > > > On 22.03.21 09:50, Markus Wanner wrote: > > > thank you for reconsidering this patch. I updated it to include the > > > required adjustments to the documentation. Please review. > > > > I tweaked the wording in the docs a bit, resulting in a v3 of this patch. > > > > One minor comment: > - The callback has to provide the same static answer for a given > - <parameter>gid</parameter> every time it is called. > + The callback may be invoked several times per transaction to decode and > + must provide the same static answer for a given pair of > > Why do you think that this callback can be invoked several times per > transaction? I think it could be called at most two times, once at > prepare time, then at commit or rollback time. So, I think using > 'multiple' instead of 'several' times is better. > + to it not being a unique identifier. Therefore, other systems combine + the <parameter>xid</parameter> with a node identifier to form a + globally unique transaction identifier. What exactly is the node identifier here? Is it a publisher or subscriber node id? We might want to be a bit more explicit here? -- With Regards, Amit Kapila.
Commits
-
Add a xid argument to the filter_prepare callback for output plugins.
- f64ea6dc5c8c 14.0 landed