Re: VACUUM optimization ideas. - Mailing list pgsql-hackers

From David Lloyd-Jones
Subject Re: VACUUM optimization ideas.
Date
Msg-id 010b01c009d8$00a679f0$f2627bd8@WORKGROUP
Whole thread Raw
In response to VACUUM optimization ideas.  (Alfred Perlstein <bright@wintelcom.net>)
List pgsql-hackers
"Alfred Perlstein" <bright@wintelcom.net>
> Here's two ideas I had for optimizing vacuum, I apologize in advance
> if the ideas presented here are niave and don't take into account
> the actual code that makes up postgresql.
                                                    * * *

This is the fist time I have dared to file in the exalted realm of
[HACKERS]. On the other hand I wrote a memo to Bill Gates a couple of years
ago which apparently resulted in C#, which is really worth a little bit of
attention, given the number of VB writers out there. I'm not quite as stupid
as I look.

Why doesn't `vacuum' happen all the time, instantly?

Like, does everybody feel psychologically more secure if a "commit" is not
really a commit, it's there for some Emergency Refind to find?

(If there are olde hardware reasons, or software -- "Well, uh, back at
BBN..." -- type reasons, I'd be happy to hear them.)

Screw it. "Is that your final answer?" is your final answer. Commit and
rebuild; optimize memory use all the time in the spare milliseconds; no
human is needed to make obvious calls.
                                                    -dlj.









pgsql-hackers by date:

Previous
From: Wim Ceulemans
Date:
Subject: Re: Bug tracking (was Re: +/- Inf for float8's)
Next
From: Tom Lane
Date:
Subject: Re: CREATE/DROP SCHEMA considered harmful