Re: What is the incremental checkpointing?
Date: Sat, 22 Mar 2008 23:25:18 +0800
Message-ID: <804dabb00803220825q5b3ee91esd6d0444d151a0688@mail.gmail.com>
Thank you Jeremy.
On Sat, Mar 22, 2008 at 1:00 PM, Jeremy Schneider
<jeremy.schneider_at_ardentperf.com> wrote:
>
> On Fri, Mar 21, 2008 at 8:26 PM, Peter Teoh <htmldeveloper_at_gmail.com> wrote:
> >
> > And what is no-force-at-commit policy, as mentioned in this paper
> >
>
> I can take a stab at this... "no-force-at-commit policy" just means that
> when you COMMIT a transaction, the change is made to a cached copy of the
> data block in memory - but the data block does not need to be written back
> to disk for the COMMIT to complete. In Oracle when you insert, update or
> delete data the actual data block does not get immediately written to disk.
> The rows you have modified are written to the redo log and the transaction
> is safe because Oracle can reconstruct the data using an old version of the
> block plus the redo log.
>
At the data/bytes level - what are the key differences between complete vs incremental checkpointing.
After some search, I think I have found a possible answer:
http://asktom.oracle.com/pls/asktom/f?p=100:11:0::::P11_QUESTION_ID:19311485023372
Let me analyze further. Thank you for the help!!!
-- Regards, Peter Teoh -- http://www.freelists.org/webpage/oracle-lReceived on Sat Mar 22 2008 - 10:25:18 CDT