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: How do you change a tables tablespace?

Re: How do you change a tables tablespace?

From: Brian Peasland <peasland_at_edcmail.cr.usgs.gov>
Date: Mon, 22 Jan 2001 15:52:09 GMT
Message-ID: <3A6C5729.6ADDD0CD@edcmail.cr.usgs.gov>

Falco,

> The cluster implementation you describe might be right, but a table is still
> a table.

A table is still a table, but a table that is part of a cluster is no longer a "table". It is now one component to a cluster. If the cluster is made of Table A and Table B, then Table A is no longer a distinct entity. True, you can query from Table A only. But when you do this, you in effect query the cluster, not just the table. So that means that you query Table B as well. This is due to the physical nature of a cluster.

> Frankly speeking as a DBA, I want to manage a clustered table just like a
> normal table.

You can't do this. You may want to, but you can't. Once a table becomes part of a cluster, then there are operations that you can not do independently to the table and leave the cluster alone. For instance, you can not move the table and leave the rest of the cluster.

> Partitions are implemented as segments too, yet these I can manage in an
> normal fashion!

Partitions have a completely different physical makeup than a cluster does. A cluster has it's own unique physical characteristics. This is exactly the point that HJR is trying to make. So while there are operations that you can perform on a table, or a partition, you may not be able to perform them on a cluster.

> Further, if it's so difficult for Oracle to find the table rows, how can it
> query correctly then?

It queries correctly, but it does not query only the one table. Going back to my example where the cluster is composed of Table A and Table B... If I issue the following:

   SELECT * FROM Table_A;
then Oracle reads the *entire* cluster. This includes Table B. Before giving the results back to you, Oracle discards the Table B component. But it is still read.

> Also, from a DBA standpoint, everyting in the docs tells me that a table in
> a cluster is
> just a table sitting in a cluster! There are ** NOWHERE ** distictions being
> made!

You might want to go back and re-read the Oracle docs. Specifically refer to
http://technet.oracle.com/docs/products/oracle8i/doc_library/817_doc/server.817/a76965/c08schem.htm#12872 In that document, you will find "multiple tables have data in each block". This sounds like a big distinction from normal tables to me!

HTH,
Brian

-- 
========================================
Brian Peasland
Raytheons Systems at
  USGS EROS Data Center
These opinions are my own and do not
necessarily reflect the opinions of my 
company!
========================================
Received on Mon Jan 22 2001 - 09:52:09 CST

Original text of this message

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