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: Another crazy hot backup theory in a book

Re: Another crazy hot backup theory in a book

From: Jeremiah Wilton <jeremiah_at_wolfenet.com>
Date: Thu, 24 Sep 1998 11:37:15 -0700
Message-ID: <Pine.OSF.3.95.980924113212.10689C-100000@gonzo.wolfenet.com>


On 24 Sep 1998 costello_at_iki.fi wrote:
>
> "MotoX" <rat_at_tat.a-tat.com> quotes a book:
> >"When you mark the beginning of the backup, Oracle will divert writes
> >from those datafiles to the SGA until the backup has completed and you
> >have marked the end of the backup. This is all that is necessary to
> >perform the online tablespace backup."
>
> Correct me if I'm wrong but it really is so that the only way to actually
> perform a backup of a running Oracle8 is to run it in ARCHIVELOG-mode
> and just copy the log files the DB generates? I kinda expected a nice
> button labeled "take backup now" or similar when 8 was released but
> it still seems to me that things must be done in the same way as in
> version 7.

I am only taking exception with the "Oracle will divert writes from those datafiles to the SGA" part of these passages. This is a common misonception about Oracle, and should be corrected.

Oracle8's RMAN is much better, in theory, that Oracle7-style hot backup (which is still supported in Oracle8). RMAN uses a server process that guarantees block-level consistency, and builds a recoverable datafile for your storage manager on the fly. Thus, RMAN does not require switching tablespaces in and out of bacup mode.

--
Jeremiah Wilton http://www.wolfenet.com/~jeremiah Received on Thu Sep 24 1998 - 13:37:15 CDT

Original text of this message

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