Re: [HACKERS] Hints proposal

From: "Bucky Jordan" <bjordan(at)lumeta(dot)com>
To: "Jim C(dot) Nasby" <jim(at)nasby(dot)net>, "Josh Berkus" <josh(at)agliodbs(dot)com>
Cc: <pgsql-hackers(at)postgresql(dot)org>, <pgsql-performance(at)postgresql(dot)org>
Subject: Re: [HACKERS] Hints proposal
Date: 2006-10-13 17:33:43
Message-ID: 78ED28FACE63744386D68D8A9D1CF5D4209CBF@MAIL.corp.lumeta.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers pgsql-performance

> I completely agree that it's much better *in the long run* to improve
> the planner and the statistics system so that we don't need hints. But
> there's been no plan put forward for how to do that, which means we
also
> have no idea when some of these problems will be resolved. If someone
> comes up with a plan for that, then we can actually look at which
options
> are better and how soon we can get fixes for these problems in place.
>

Would it be helpful to have a database of EXPLAIN ANALYZE results and
related details that developers could search through? I guess we sort of
have that on the mailing list, but search/reporting features on that are
pretty limited. Something like the "Report Bug" feature that seems to be
growing popular in other software (Windows, OS X, Firefox, etc) might
allow collection of useful data. The goal would be to identify the most
common problems, and some hints at what's causing them.

Maybe have a form based submission so you could ask the user some
required questions, ensure that they aren't just submitting EXPLAIN
results (parse and look for times maybe?), etc?

I guess the general question is, what information could the users
provide developers to help with this, and how can it be made easy for
the users to submit the information, and easy for the developers to
access in a meaningful way?

As a developer/contributor, what questions would you want to ask a user?
>From reading the mailing lists, these seem to be common ones:
- Copy of your postgres.conf
- Basic hardware info
- Explain Analyze Results of poor performing query
- Explain Analyze Results of anything you've gotten to run better
- Comments

If there's interest- web development is something I can actually do
(unlike pg development) so I might actually be able to help with
something like this.

- Bucky

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Tom Lane 2006-10-13 17:35:51 Re: more anti-postgresql FUD
Previous Message Joshua D. Drake 2006-10-13 17:33:24 Re: [PERFORM] Hints proposal

Browse pgsql-performance by date

  From Date Subject
Next Message David Fetter 2006-10-13 17:43:57 Re: [PERFORM] Hints proposal
Previous Message Joshua D. Drake 2006-10-13 17:33:24 Re: [PERFORM] Hints proposal