Re: What is the purpose of segment level checkpoint before DROP/TRUNCATE of a table?
From: Tanel Poder <tanel_at_tanelpoder.com>
Date: Thu, 14 Jul 2011 00:59:22 +0300
Message-ID: <CAMHX9JJ+BaOf7xc9=fbMXKg9dih6Sd32pEpt9TOApKdFXXaS2A_at_mail.gmail.com>
I don't think overwriting files of an open database with old copies (with old data) should be used to prove how Oracle works... there's much more going on in SGA & buffer cache for example, other than just writes/reads to/from the file....
Date: Thu, 14 Jul 2011 00:59:22 +0300
Message-ID: <CAMHX9JJ+BaOf7xc9=fbMXKg9dih6Sd32pEpt9TOApKdFXXaS2A_at_mail.gmail.com>
I don't think overwriting files of an open database with old copies (with old data) should be used to prove how Oracle works... there's much more going on in SGA & buffer cache for example, other than just writes/reads to/from the file....
-- Tanel Poder http://blog.tanelpoder.com On Mon, Jul 11, 2011 at 9:46 AM, Saibabu Devabhaktuni <saibabu_d_at_yahoo.com>wrote:Received on Wed Jul 13 2011 - 16:59:22 CDT
> Here is a test case to simulate the behavior of Oracle not doing object
> level checkpoint during truncate table operation (and hence not guaranteeing
> dirty buffers write to disk) and Oracle reading stale data (logical
> corruption).
>
>
> host cp $ORACLE_HOME/dbs/tr1.dbf.org $ORACLE_HOME/dbs/tr1.dbf
>
>
>
-- http://www.freelists.org/webpage/oracle-l