Date: Tue, 6 Jun 2006 12:20:19 -0400
From: Andrew Sullivan <ajs@crankycanuck.ca>
To: pgsql-performance@postgresql.org
Subject: Re: Some queries starting to hang
Message-ID: <20060606162019.GC10018@phlogiston.dyndns.org>
References: <1149534308.3583.56.camel@bg002441.pro-unlimited.com>
	<22706.1149538039@sss.pgh.pa.us>
	<20060605203851.GB7071@phlogiston.dyndns.org>
	<23915.1149541591@sss.pgh.pa.us>
	<1149604741.2621.487.camel@localhost.localdomain>
	<17148.1149605005@sss.pgh.pa.us>
	<1149605576.2621.502.camel@localhost.localdomain>
	<87ejy29tx1.fsf@stark.xeocode.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <87ejy29tx1.fsf@stark.xeocode.com>
User-Agent: Mutt/1.5.9i

On Tue, Jun 06, 2006 at 11:37:46AM -0400, Greg Stark wrote:

> An alternate approach would be to implement a SIGINFO handler that
> prints out the explain analyze output for the data built up so far.
> You would be able to keep hitting C-t and keep getting updates
> until the query completes or you decided to hit C-c.

This is even better, and pretty much along the lines I was thinking
in my other mail.  If you can see the _first_ spot you break, you can
start working.  We all know (or I hope so, anyway) that it would be
better to get the full result, and know everything that needs
attention before starting.  As nearly as I can tell, however, they
don't teach Mill's methods to MBAs of a certain stripe, so changes
start getting made without all the data being available.  It'd be
nice to be able to bump the set of available data to something
higher than "none".

(That said, I appreciate that there's precious little reason to spend
a lot of work optimising a feature that is mostly there to counteract
bad management practices.)

A

-- 
Andrew Sullivan  | ajs@crankycanuck.ca
When my information changes, I alter my conclusions.  What do you do sir?
		--attr. John Maynard Keynes