| From: | Chris Travers <chris(at)travelamericas(dot)com> |
|---|---|
| To: | Akshay Mathur <akshay(dot)mathur(at)airtightnetworks(dot)net> |
| Cc: | 'Ron' <rjpeace(at)earthlink(dot)net>, pgsql-performance(at)postgresql(dot)org |
| Subject: | Re: Massive performance issues |
| Date: | 2005-09-02 18:00:46 |
| Message-ID: | 4318934E.7070404@travelamericas.com |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-performance |
Akshay Mathur wrote:
>Ron,
>
>Can you give me some pointers to make the tables RAM resident. If one
>does so, is the program accessing the data need to change. Does pgsql
>take care to write the data to disk?
>
>
>
PostgreSQL tried to intelligently cache information and then will also
use the OS disk cache as a secondary cache. So a sufficiently small and
frequently accessed table will be resident in RAM.
The simplest way to affect this calculus is to put more RAM in the
machine. There are hacks I can think of to create RAM caches of
specific tables, but I don't want to take responsibility for anyone
trying these and running into trouble.
Best Wishes,
Chris Travers
Metatron Technology Consulting
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Matteo Beccati | 2005-09-02 18:33:51 | Re: ORDER BY and LIMIT not propagated on inherited |
| Previous Message | Simon Riggs | 2005-09-02 17:54:56 | Re: ORDER BY and LIMIT not propagated on inherited |