Postgres Pro
Downloads
Home   >   mailing lists

Re: Make deparsing of column defaults faster - Mailing list pgsql-hackers

From Peter Eisentraut
Subject Re: Make deparsing of column defaults faster
Date
Msg-id 53e9c6cc-460c-756e-e619-a4dcedd6f3d7@2ndquadrant.com
Whole thread Raw
In response to Re: Make deparsing of column defaults faster  (Justin Pryzby <pryzby@telsasoft.com>)
List pgsql-hackers
Tree view 
On 05.07.18 18:58, Justin Pryzby wrote:
> I checked on one customer running PG10.4, for which pg_dump takes 8 minutes to
> pg_dump -s.
> 
> I imported existing schema to PG12dev (which itself took 25min) and compared:
> patched: 2m33.616s
> unpatched: 7m19.578s

Which patch is that, Jeff's or mine?

> Note, that affects pg_upgrade, which is how this issue originally came up [0].
> (But I believe pg_upgrade likes to call pg_dump from the old server version, so
> pg_upgrade to v11 couldn't benefit unless this was included in PG10.5).

pg_upgrade calls the new pg_dump.

-- 
Peter Eisentraut              http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services

pgsql-hackers by date:

Previous
From: Tom Lane
Date:
Subject: Re: "interesting" issue with restore from a pg_dump with a database-wide search_path
Next
From: Peter Eisentraut
Date:
Subject: Re: [PATCH] Use access() to check file existence inGetNewRelFileNode().