Re: Index Unqiueness - Mailing list pgsql-hackers

From Bruce Momjian
Subject Re: Index Unqiueness
Date
Msg-id 20130309004104.GF3005@momjian.us
Whole thread Raw
In response to Re: Index Unqiueness  (Robert Haas <robertmhaas@gmail.com>)
List pgsql-hackers
On Fri, Mar  8, 2013 at 10:26:21AM -0500, Robert Haas wrote:
> overhead seems badly overpriced for insert-only tables.  These are not
> fundamental truths of the universe, or even of PostgreSQL; they are
> specific consequences of the representation we've chosen for heaps.
> Many of them are things that we've grown into, rather than designed
> with malice aforethought: for example, freezing is a consequence of
> the after-the-fact desire to be able to support more than 4bn
> transactions over the lifetime of the database.  So it's way better
> than what we had before, and yet, if we all sat down and designed a
> new on-disk storage format for a new product today, I'm sure none of
> us would pick one that expires after 2bn transactions.

One thing to remember is that our freeze level is 200M transactions
because of clog lookups/size, not wraparound concerns.

--  Bruce Momjian  <bruce@momjian.us>        http://momjian.us EnterpriseDB
http://enterprisedb.com
 + It's impossible for everything to be true. +



pgsql-hackers by date:

Previous
From: Andrew Dunstan
Date:
Subject: Re: Duplicate JSON Object Keys
Next
From: Greg Stark
Date:
Subject: Re: Why do we still perform a check for pre-sorted input within qsort variants?