Re: consistent reads and unique key updates

From: Ls Cheng <exriscer_at_gmail.com>
Date: Wed, 25 Apr 2012 21:58:04 +0200
Message-ID: <CAJ2-Qb_8Gu-iHAeANOiUjsvUEcoxSr2QwRqMa4-RTEQ_B8e6Gw_at_mail.gmail.com>

ok, just rerun the test but same output :-S end of the test and final query session statistics: ### CR blocks

created                                                         0
### RowCR -
resume                                                            0
### RowCR - row
contention                                                    0
### RowCR
attempts                                                            0
### RowCR
hits                                                                0
### consistent
gets                                                          27
### data blocks consistent reads - undo records
applied                       0

I rerun again and checking only the statistics which has changed between queries and updates but none seems interesting, puzzled

On Wed, Apr 25, 2012 at 2:51 PM, Sidney Chen <huanshengchen_at_gmail.com>wrote:

> Quote from Jonathan Lewis's new book "Oracle Core", you can check the
> stats "RowCR attempts", "RowCR hits" in session 1 to confirm this.
>
> --quote--
>
> When Oracle accesses a row through an indexed access path, it may be able
> to use the RowCR (row consistent read) mechanism when it gets to the table
> block. This means that Oracle has gone through the complex processing I’ve
> just described to get the index leaf block into a read-consistent state and
> has then been able to check that the row itself has not changed since the
> reconstructed commit SCN for the index entry.
>
> --end of quote--
>
> --
> Regards
> Sidney Chen
>
>
>

--
http://www.freelists.org/webpage/oracle-l

Received on Wed Apr 25 2012 - 14:58:04 CDT