Re: WAL bypass for CTAS

From: Simon Riggs <simon(at)2ndquadrant(dot)com>
To: Alvaro Herrera <alvherre(at)surnet(dot)cl>
Cc: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, pgsql-patches(at)postgresql(dot)org, Alon Goldshuv <agoldshuv(at)greenplum(dot)com>
Subject: Re: WAL bypass for CTAS
Date: 2005-06-20 21:19:09
Message-ID: 1119302349.3645.336.camel@localhost.localdomain
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers pgsql-patches

On Mon, 2005-06-20 at 17:09 -0400, Alvaro Herrera wrote:
> On Mon, Jun 20, 2005 at 09:55:12PM +0100, Simon Riggs wrote:
>
> > I put those changes in mainly for COPY. If you don't make any request at
> > all to FSM then a relation never gets to the MRU relation FSM list. I
> > agree that it is not strictly necessary, but leaving it off would be a
> > change in behaviour, since COPY did previously cause the relation to get
> > to the MRU. That could be a problem, since a relation might not then be
> > allocated any FSM pages following a vacuum.
>
> Is that a problem?

Not for me, but I wanted to explain the change in behaviour that
implies.

> If the pages don't fit in FSM, then maybe the system
> is misconfigured anyway. The person running the DW should just increase
> the FSM settings, which is hardly a costly thing because it uses so
> little memory.

If you aren't on the relation list you don't get any more pages than the
minimum. No matter how many fsm_pages you allocate. If fsm_pages covers
everything, then you are right, there is no problem.

Best Regards, Simon Riggs

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Andreas Pflug 2005-06-20 22:43:51 Re: [PATCHES] default database creation with initdb
Previous Message Alvaro Herrera 2005-06-20 21:09:39 Re: WAL bypass for CTAS

Browse pgsql-patches by date

  From Date Subject
Next Message Neil Conway 2005-06-20 23:37:05 Re: code cleanup for tz
Previous Message Alvaro Herrera 2005-06-20 21:09:39 Re: WAL bypass for CTAS