Re: Centralizing protective copying of utility statements - Mailing list pgsql-hackers

From Andres Freund
Subject Re: Centralizing protective copying of utility statements
Date
Msg-id 20210617192300.7gkpqim43gk7ernd@alap3.anarazel.de
Whole thread Raw
In response to Centralizing protective copying of utility statements  (Tom Lane <tgl@sss.pgh.pa.us>)
Responses Re: Centralizing protective copying of utility statements  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-hackers
Hi,

On 2021-06-16 21:39:49 -0400, Tom Lane wrote:
> Although this adds some overhead in the form of copying of
> utility node trees that won't actually mutate during execution,
> I think that won't be too bad because those trees tend to be
> small and hence cheap to copy.  The statements that can have
> a lot of substructure usually contain expression trees or the
> like, which do have to be copied for safety.  Moreover, we buy
> back a lot of cost by removing pointless copying when we're
> not executing on a cached plan.

Have you evaluated the cost in some form? I don't think it a relevant
cost for most utility statements, but there's a few exceptions that *do*
worry me. In particular, in some workloads transaction statements are
very frequent.

Greetings,

Andres Freund



pgsql-hackers by date:

Previous
From: Robert Haas
Date:
Subject: Re: [Proposal] Fully WAL logged CREATE DATABASE - No Checkpoints
Next
From: Andres Freund
Date:
Subject: Re: Centralizing protective copying of utility statements