Horton

Re: First steps with 8.3 and autovacuum launcher

Simon Riggs <simon@2ndquadrant.com>

View thread View on postgr.es
From: Simon Riggs <simon@2ndquadrant.com>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: Alvaro Herrera <alvherre@commandprompt.com>, Michael Paesold <mpaesold@gmx.at>, Heikki Linnakangas <heikki@enterprisedb.com>, Gregory Stark <stark@enterprisedb.com>, Guillaume Smet <guillaume.smet@gmail.com>, "Matthew T. O'Connor" <matthew@zeut.net>, Stefan Kaltenbrunner <stefan@kaltenbrunner.cc>, PostgreSQL-development <pgsql-hackers@postgresql.org>
Date: 2007-10-12T17:30:30Z
Lists: pgsql-hackers 
On Fri, 2007-10-12 at 12:42 -0400, Tom Lane wrote:
> Simon Riggs <simon@2ndquadrant.com> writes:
> > On Fri, 2007-10-12 at 11:26 -0400, Tom Lane wrote:
> >> Why not SIGINT?
> 
> > I must be missing something. How would I tell the difference between
> > manual and automatic cancels if we use SIGINT for both cases?
> 
> Why do you need to?  I thought the plan was that DeadlockCheck would
> only try to signal autovac workers.

...thinks...

On Fri, 2007-10-12 at 10:19 -0400, Tom Lane wrote:

> In short: put in the automatic lock cancel for regular vacuums, disable
> it for antiwraparound vacuums, but don't disable manual cancels; and
> definitely don't invent a complicated new set of behaviors around that.

So you mean "dont allow automatic cancels of manually submitted
VACUUMs".

I thought you meant don't disable manually-requested cancels of
autovacuums.

Can you explain further what you meant by "don't disable manual
cancels".

(laughs) wish I had a pound for every time we'd misunderstood each
other, or at least, a pound every time I misunderstood you. :-)

-- 
  Simon Riggs
  2ndQuadrant  http://www.2ndQuadrant.com