Horton
Menu

Re: PageIsAllVisible()'s trustworthiness in Hot Standby

Robert Haas <robertmhaas@gmail.com>

View thread
From: Robert Haas <robertmhaas@gmail.com>
To: Pavan Deolasee <pavan.deolasee@gmail.com>
Cc: Andres Freund <andres@2ndquadrant.com>, pgsql-hackers <pgsql-hackers@postgresql.org>
Date: 2012-12-14T20:45:42Z
Lists: pgsql-hackers 
On Wed, Dec 12, 2012 at 12:31 PM, Pavan Deolasee
<pavan.deolasee@gmail.com> wrote:
> On Wed, Dec 12, 2012 at 1:35 AM, Robert Haas <robertmhaas@gmail.com> wrote:
>> On Tue, Dec 4, 2012 at 12:10 PM, Pavan Deolasee
>> <pavan.deolasee@gmail.com> wrote:
>>> Hmm. Yeah, I do not have guts to prove that either. I'll probably write up a
>>> comment for your consideration to explain why we don't trust PD_ALL_VISIBLE
>>> in Hot standby for seq scans, but still trust VM for index-only scans.
>>
>> Sure.
>>
>
> Here is a small patch that adds comments to heap scan code explaining
> why we don't trust the all-visible flag in the page, still continue to
> support index-only scans on hot standby.

Committed, with a few modifications to the last part.

-- 
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company

Commits

Same data as JSON: GET /api/v1/messages/:b64id/commits the thread's linked commits as JSON, with link sources. API reference →

  1. Prevent index-only scans from returning wrong answers under Hot Standby.