Re: concepts document part about separating indexes and tablspaces

From: Riyaj Shamsudeen <>
Date: Thu, 15 May 2008 15:44:11 -0500
Message-ID: <>

This is also mentioned in this note *148342.1 with last revision date as ** 18-SEP-2007.

  • Separating Tables and Indexes

  Tables that are joined ( simultaneously accessed during a query) often should   have their data and index separated. The following example shows a table join   and one possible solution for managing the data:

  Select COL1, COL2 ....
  where ...;

  Data management solution:

  Disk1: CUST_HEADER Table
  Disk2: CUST_HEADER Index
  Disk3: CUST_DETAIL Table
  Disk4: CUST_DETAIL Index

  The preceding solution allows the table join to be done while accessing   four different disks and controllers. Separate data and index files onto   different physical disks and controllers; consequently, when tables and   indexes are accessed at the same time, they will not be accessing the   same physical disks.


Riyaj Shamsudeen
The Pythian Group -

On Thu, May 15, 2008 at 3:32 PM, David Aldridge <>

> It's actually in the DBA Guide, and it's still there now:
> "Using different tablespaces (on different disks) for a table and its index
> produces better performance than storing the table and index in the same
> tablespace. Disk contention is reduced."
> ----- Original Message ----
> From: David Sharples <>
> To:
> Cc:
> Sent: Thursday, May 15, 2008 3:18:48 PM
> Subject: Re: concepts document part about separating indexes and tablspaces
> because it doesn't
> 2008/5/15 Rick Ricky <>:
>> i was talking to someone about this today. I cannot remember where in the
>> Concepts document that it says that separating data from indexes improves
>> performances?
Received on Thu May 15 2008 - 15:44:11 CDT

Original text of this message