Re: pgsql: Fix management of pendingOpsTable in auxiliary processes. - Mailing list pgsql-committers

From Simon Riggs
Subject Re: pgsql: Fix management of pendingOpsTable in auxiliary processes.
Date
Msg-id CA+U5nMJ1MBvg8Br=ovUYEeSXGN7xfdXnOc5GgE47VgLfL3+vDA@mail.gmail.com
Whole thread Raw
In response to pgsql: Fix management of pendingOpsTable in auxiliary processes.  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-committers
On 18 July 2012 20:38, Tom Lane <tgl@sss.pgh.pa.us> wrote:

> Creation of the table in the
> bgwriter is fatal, because it absorbs fsync requests that should have gone
> to the checkpointer; instead they just sit in bgwriter local memory and are
> never acted on.

Hmmm, sorry about that.

I would swear I had fixed that already, which is fairly strange.

--
 Simon Riggs                   http://www.2ndQuadrant.com/
 PostgreSQL Development, 24x7 Support, Training & Services

pgsql-committers by date:

Previous
From: Tom Lane
Date:
Subject: pgsql: Fix statistics breakage from bgwriter/checkpointer process split
Next
From: Heikki Linnakangas
Date:
Subject: pgsql: Refactor the way code is shared between some range type function