Indices pinned in pageable RAM?

From: Marty Scholes <marty(at)outputservices(dot)com>
To: pgsql-hackers(at)postgresql(dot)org
Cc: Marty Scholes <marty(at)outputservices(dot)com>
Subject: Indices pinned in pageable RAM?
Date: 2004-06-02 19:38:47
Message-ID: 40BE2CC7.9050105@outputservices.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Has anyone seriously looked at how it would impact things to give the
DBA the option of storing certain indices in RAM instead of on disk?

Queries (both select and insert/update) against heavily indexed tables
do most of the reads and writes to the index trees and relatively little
reading and writing to the actual table data.

The index is completely redundant (read: rebuildable during restart) and
its rebuild "should" be very parallelizable. Without the WAL and disk
synchronization munching, even a single-threaded rebuild during restart
should be fairly quick.

This would allow the index to stay in RAM. giving it a higher priority
than all other disk objects in the OS cache. Putting it in pageable RAM
would allow the OS to reclaim core if it absolutely had to.

Am I way off here?

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Jonathan Gardner 2004-06-02 20:28:46 Re: constraint upon view
Previous Message Andrew Dunstan 2004-06-02 18:54:48 Re: Default Locale in initdb