Re: Possible performance regression with pg_dump of a large numberof relations - Mailing list pgsql-hackers

From Stephen Frost
Subject Re: Possible performance regression with pg_dump of a large numberof relations
Date
Msg-id 20180125190908.GT2416@tamriel.snowman.net
Whole thread Raw
In response to Re: Possible performance regression with pg_dump of a large number ofrelations  (Luke Cowell <lcowell@gmail.com>)
List pgsql-hackers
Luke,

* Luke Cowell (lcowell@gmail.com) wrote:
> > On Jan 24, 2018, at 2:56 PM, Stephen Frost <sfrost@snowman.net> wrote:
> >>> ERROR:  relation "pg_init_privs" does not exist
> >>> LINE 139: LEFT JOIN pg_init_privs pip
> >
> > I certainly hope that works on 9.6, since that's when pg_init_privs was
> > added..
>
> My mistake. That error is from my 9.5 server. It does error on 9.6, but I get the following error:

No worries!

> $ psql --version
> psql (PostgreSQL) 9.6.6
> $ psql postgres < query.sql
> ERROR:  function pg_get_partkeydef(oid) does not exist
> LINE 126: pg_get_partkeydef(c.oid) AS partkeydef,
>           ^
> HINT:  No function matches the given name and argument types. You might need to add explicit type casts.

Ah, yes, that makes sense, that function wasn't included in 9.6.  That's
not really a problem though, pg_dump knows how to talk to different
versions of PG, it just happens that the query I sent was the one for
HEAD and not for older versions.

> > Presuming I can make it work, the idea would be to back-port it to 9.6
> > and 10, since pg_init_privs and this code was added in 9.6.
>
> A 9.6 backport would be excellent.

Yup, I'll make sure that the query that's fun for 9.6 server systems
works on, well, 9.6 systems. ;)

Thanks!

Stephen

Attachment

pgsql-hackers by date:

Previous
From: Pavel Stehule
Date:
Subject: Re: plpgsql function startup-time improvements
Next
From: Andres Freund
Date:
Subject: Re: JIT compiling with LLVM v9.0