Re: psql document fix about showing FDW options

From: Shigeru Hanada <shigeru(dot)hanada(at)gmail(dot)com>
To: Robert Haas <robertmhaas(at)gmail(dot)com>
Cc: "pgsql-hackers >> PostgreSQL-development" <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: psql document fix about showing FDW options
Date: 2011-08-12 02:28:33
Message-ID: 4E448FD1.4020409@gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

(2011/08/12 0:48), Robert Haas wrote:
> 2011/8/9 Shigeru Hanada<shigeru(dot)hanada(at)gmail(dot)com>:
>> postgres=# \d pgbench_accounts
>> Foreign table "public.pgbench_accounts"
>> Column | Type | Modifiers | Options
>> ----------+---------------+-----------+---------------
>> aid | integer | not null | {colname=aid}
>> bid | integer | |
>> abalance | integer | |
>> filler | character(84) | |
>> Server: pgbench
>> Options: {nspname=public,"relname=foo bar"}
>
> Looks like you've got postgresql_fdw working there... do you plan to
> submit that for 9.2 soon? Any plans for qual pushdown?

Yeah, I have (hopefully) working FDW for PostgreSQL which is based on
the one which has been proposed for 9.1, and updates done by Heikki.
I've implemented:

* SELECT clause omitting
* WHERE clause pushdown (assuming remote has same functions/oprators)
* private connection caching, and VIEW which shows active connections
* switch simple SELECT and CURSOR based on estimated # of rows,
and FDW options to control this behavior

I'd like to submit this fdw in CF 2011-09 (ASAP) for discussion of
itself and FDW API enhancement. Please see the wiki page below for details.

http://wiki.postgresql.org/wiki/SQL/MED#PostgreSQL

>> I attached WIP patch for these fixes.
>> Please apply show_per_table_options.patch first against head, and then
>> rename_desc_headers.patch can be applied.
>
> Looks good to me. Committed.

Thanks!

Regards,
--
Shigeru Hanada

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Robert Haas 2011-08-12 02:39:02 Re: index-only scans
Previous Message Andrew Dunstan 2011-08-12 01:51:29 Re: index-only scans