Re: Reorg Oracle 8 Partition tables

From: Jonathan Lewis <jonathan_at_jlcomp.demon.co.uk>
Date: Sat, 20 Feb 1999 21:20:42 -0000
Message-ID: <919545649.20873.2.nnrp-07.9e984b29@news.demon.co.uk>

Just one little extra to throw in on this discussion:

Thomas brought up the Parallel Query Option - which is exactly when a very large number of 'sub-objects' can cause a problem.

In 7.3, the Query Coordinator will kick off an extent-based checkpoint for every extent in the object being queried.

In 8.0, the Query Coordinator will kick off an object-based checkpoint for every object (e.g. partition) being queried.

If you have a very large db_block_buffer, an extent-based, or object-based checkpoint can take a significant amount of time.

Obviously, as all participants in this thread have commented, the significance of any particular cost is dependent upon what you are trying to achieve - but this is another aspect of space management that can need consideration.

Jonathan Lewis
Yet another Oracle-related web site: www.jlcomp.demon.co.uk

P.S. There is a note about this on my website. Received on Sat Feb 20 1999 - 15:20:42 CST