From: | Robert Haas <robertmhaas(at)gmail(dot)com> |
---|---|
To: | Simon Riggs <simon(at)2ndquadrant(dot)com> |
Cc: | pgsql-hackers <pgsql-hackers(at)postgresql(dot)org> |
Subject: | Re: Additional options for Sync Replication |
Date: | 2011-03-28 12:14:10 |
Message-ID: | AANLkTi=oOXOB4EhPFP31aTJ1NWWeMPHTm-gNzZkmypMe@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
On Mon, Mar 28, 2011 at 6:14 AM, Simon Riggs <simon(at)2ndquadrant(dot)com> wrote:
> On Mon, Mar 28, 2011 at 10:52 AM, Simon Riggs <simon(at)2ndquadrant(dot)com> wrote:
>>
>> You have no basis on which to prevent this.
>
> It's also already on the Open Items list, put there by you.
Huh? There is an open item about whether we should merge
synchronous_commit and synchronous_replication into a single GUC,
which might be a better interface since it would typically give the
user one less thing to have to fiddle with, but there is certainly no
open item for adding additional sync rep modes. Merging those two
GUCs is a reasonable thing to consider even at this late date, because
once we cut beta - and certainly once we cut final - we'll be stuck
supporting whatever interface we release for 5+ years. There is no
similar risk for this patch - the only risk of not committing this
feature now is that we won't have this feature in 9.1. But if we
accept that as valid justification for committing it, then everything
is fair game, and that is not going to lead to a timely release.
> Why is this even a discussion point?
Adding more new code is likely to add more new bugs, and we're trying
not to do that right now, so we can make a release.
--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company
From | Date | Subject | |
---|---|---|---|
Next Message | Robert Haas | 2011-03-28 12:18:03 | Re: Problem with streaming replication, backups, and recovery (9.0.x) |
Previous Message | Thom Brown | 2011-03-28 11:14:07 | Comments on system tables and columns |