[Theforum] Re: Oracle Clustering

Dean Mah dmah at shaw.ca
Thu Feb 7 09:58:14 CST 2002


Oracle has cluster tables and indexes but they are not equivalent to
what Microsoft's notion.

Oracle clustering is a way of storing tables that share columns near
each other (in the same data block).  Ordering is *not* implicit in
physical storage.  (In theory, it shouldn't have to be.  If the index
is being used a hash into physical memory, you don't need ordering.)

Dean


Madhu Menon writes:

> (sigh) This will be my last message on the subject.
>
> At 05:46 PM 2/6/2002, Warden, Matt wrote:
> > >Setting a clustered index on a column (often the primary key) in a table
> > >in SQL Server physically rearranges the order of the rows in the
> > >table. The original comment was about whether data could be rearranged in
> > >a table itself, to which I responded that a clustered index did this.
> >
> >I know. Which is why I responded.
> >
> >It doesn't.
> >
> >(see first message)
>
> Then you need to read the SQL Server docs more carefully.
>
> <http://msdn.microsoft.com/library/default.asp?url=/library/en-us/createdb/cm_8_des_05_5h6b.asp>
>
> <quote>
>
> A clustered index determines the physical order of data in a table. A
> clustered index is analogous to a telephone directory, which arranges data
> by last name. Because the clustered index dictates the physical storage
> order of the data in the table, a table can contain only one clustered index.
>
> </quote>
>
> and yet again:
>
> <http://msdn.microsoft.com/library/default.asp?url=/library/en-us/architec/8_ar_da2_8sit.asp>
>
> <quote>
> Clustered tables are tables that have a clustered index.
>
> The data rows are stored in order based on the clustered index key.
> </quote>
>
> The indices you're talking about are *non-clustered* indices.
>
> Other articles that say the *same* thing:
> http://www.4guysfromrolla.com/webtech/sqlguru/q021700-1.shtml
> http://www.swynk.com/friends/achigrik/IndexOptimTips.asp



More information about the theforum mailing list