Re: patch: SQL/MED(FDW) DDL

From: Robert Haas <robertmhaas(at)gmail(dot)com>
To: Shigeru HANADA <hanada(at)metrosystems(dot)co(dot)jp>
Cc: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, Heikki Linnakangas <heikki(dot)linnakangas(at)enterprisedb(dot)com>, Itagaki Takahiro <itagaki(dot)takahiro(at)gmail(dot)com>, Alvaro Herrera <alvherre(at)commandprompt(dot)com>, SAKAMOTO Masahiko <sakamoto(dot)masahiko(at)oss(dot)ntt(dot)co(dot)jp>, pgsql-hackers <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: patch: SQL/MED(FDW) DDL
Date: 2010-10-11 17:22:13
Message-ID: AANLkTimzEB1fSMU9YH8Z3WsKdVACnG8LA74RXdTTrrBb@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On Thu, Oct 7, 2010 at 7:20 AM, Shigeru HANADA
<hanada(at)metrosystems(dot)co(dot)jp> wrote:
> [ design for FDW ]

This seems mostly sensible.

> In this step, optimizer calls FdwRoutine->GetStatistics() to get
> estimated costs of the foreign scan.  FDW can calculate costs by
> itself with some statistics, or get costs from remote-side, or leave
> the costs as is.

I think this should be called EstimateCosts rather than GetStatistics.

> Maybe DBAs want to see how much and what kind of connections are
> established now, so we should show list of active connections via a
> view or function.  To treat connections transparently, I propose
> that connection pooling mechanism in the core module, not in each FDW
> module.  Or should we add API to return list of active connections in
> common format?

I think for starters we should make this the responsibility of each
FDW. It seems a bit hard to speculate about what we might want down
the road at this point.

> Should we add API which is called from ANALYZE to allow FDWs to handle
> statistics when user wants.

I would probably put this off also, to a later phase of the project.

--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Tom Lane 2010-10-11 17:30:50 Re: pgxs docdir question
Previous Message Tom Lane 2010-10-11 16:03:19 Re: On the usefulness of hint bits