Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.server -> Re: LOCAL INDEXES

Re: LOCAL INDEXES

From: Steve Haynes <steve_at_rwx777.demon.co.uk>
Date: Wed, 24 Jun 1998 23:26:45 +0100
Message-ID: <7d+qiFAl0Xk1Ew7Z@rwx777.demon.co.uk>


In article <6mr5j7$lo0$1_at_news.ses.cio.eds.com>, Bert Scalzo <bscalz01_at_7-11.com> writes
>Not true. If you have a partitioned table with 3 partitions (t1, t2, t3),
>then let's examine a sample of your indexing options:
>
>1. local indexing creates 3 index partitions (l1 for t1, l2 for t2, l3 for
>t3)
>
>2. gloabl indexing could have 2 partitions (g1 for t1 and t2, g2 for t3)
>
>3. plain old index can only be 1 segment (i1 for t1 - t3)
>
>There is no way this is internally the same. While I could create a gloabl
>index for with one partition for all three partitions -- it is only
>logically an equivalent ...
>

Accepted.. sort of.
I think The definition of global indexing here is an uneven spread of partitioned indexes. I accept that yould could come up with any spread you so desire in theory, but in my experience it's either one index per partition or one plain index. I can think of some cases where it might be sensible to spread indexes around partitions but my instincts tell me to avoid that.
Steve
--
"The floggings will continue until morale improves." Received on Wed Jun 24 1998 - 17:26:45 CDT

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US