Re: Patch: add timing of buffer I/O requests

Peter Geoghegan <peter@2ndquadrant.com>

From: Peter Geoghegan <peter@2ndquadrant.com>

To: Tom Lane <tgl@sss.pgh.pa.us>

Cc: Greg Stark <stark@mit.edu>, Robert Haas <robertmhaas@gmail.com>, "ktm@rice.edu" <ktm@rice.edu>, Greg Smith <greg@2ndquadrant.com>, Magnus Hagander <magnus@hagander.net>, Ants Aasma <ants@cybertec.at>, pgsql-hackers@postgresql.org

Date: 2012-04-14T03:21:44Z

Lists: pgsql-hackers

On 14 April 2012 03:01, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> Realistically, I'm more worried about collisions due to inadequacies in
> the jumble calculation logic (Peter already pointed out some risk
> factors in that regard).

It's important to have a sense of proportion about the problem. The
worst thing that a collision can do is lead the DBA on a bit of a wild
goose chase. Importantly, collisions across databases and users are
impossible. I've always taken the view that aggregating query
statistics is a lossy process, and these kinds of problems seem like a
more than acceptable price to pay for low-overhead dynamic query
statistics .

-- 
Peter Geoghegan       http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training and Services