From: | Michael Paquier <michael(at)paquier(dot)xyz> |
---|---|
To: | Masahiko Sawada <sawada(dot)mshk(at)gmail(dot)com> |
Cc: | Peter Eisentraut <peter(dot)eisentraut(at)2ndquadrant(dot)com>, PostgreSQL-development <pgsql-hackers(at)postgresql(dot)org> |
Subject: | Re: Copy function for logical replication slots |
Date: | 2018-07-06 00:21:41 |
Message-ID: | 20180706002141.GB2366@paquier.xyz |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
On Thu, Jul 05, 2018 at 05:24:48PM +0900, Masahiko Sawada wrote:
> Yes, you're right. To guarantee that restart LSN of copied slot is
> available, it seems to me that it's better to copy new slot while
> holding the origin slot as you mentioned before. Since the replication
> slot creation code assumes that a process creating a new slot doesn't
> have any slots we should save origin slot temporary and create new
> one, and then restore it.
This will require some refactoring first I think as most of the slot
routines assume that the process owning it is the one doing the calls,
so this has a string smell of a patch set being splitted.
> It might be a bit tricky but would work fine.
Sawada-san, will you be able to rewrite this patch soon or should it be
moved to the next commit fest? I would suggest to do the latter as this
is no small work, and this needs careful thoughts.
--
Michael
From | Date | Subject | |
---|---|---|---|
Next Message | Peter Geoghegan | 2018-07-06 00:26:11 | Re: Why B-Tree suffix truncation matters |
Previous Message | Thomas Munro | 2018-07-06 00:18:02 | Re: [HACKERS] Optional message to user when terminating/cancelling backend |