Sorry Mladen, you are right. Thanks for correcting me. Now I see my error.
On Wed, Dec 9, 2015 at 8:53 AM, Mladen Gogala <gogala.mladen@xxxxxxxxx>
wrote:
Hi Michael,
No, it shouldn't be that. I am looking for the extent which contains the
block# 3650527. That means that the beginning block of the extent is less
or equal 3650527 and that the end of the extent is greater or equal to
3650527.
Regards
On 12/09/2015 09:44 AM, Michael Cunningham wrote:
Not sure I'm correct or if this will help but it looks to me like the
operators are backwards in this statement.
block_id<=3650527 and
5 block_id+blocks>=3650527;
Shouldn't it be
block_id>=3650527 and
5 block_id+blocks<=3650527;
On Dec 8, 2015 3:46 PM, "Mladen Gogala" <gogala.mladen@xxxxxxxxx> wrote:
I ran into a corrupt block which makes rman backup fail:
released channel: backup_ch1
RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-03002: failure of backup plus archivelog command at 12/08/2015
01:58:24
RMAN-03009: failure of backup command on backup_ch1 channel at 12/08/2015
01:58:01
ORA-19566: exceeded limit of 0 corrupt blocks for file
+DATA/prod/datafile/indexes.1030.873578613
OK, let's see what is it:
SQL> select file#,block# from v$database_block_corruption;
FILE# BLOCK#
---------- ----------
36 3650527
SQL> select owner,segment_name,segment_type
2 from dba_extents
3 where file_id=36 and
4 block_id<=3650527 and
5 block_id+blocks>=3650527;
no rows selected
Great! No segments are affected! Let's fix it:
Copyright (c) 1982, 2011, Oracle and/or its affiliates. All rights
reserved.
connected to target database: PROD (DBID=204240718)
RMAN> recover corruption list;
Starting recover at 08-DEC-15
RMAN-00571: ===========================================================
RMAN-00569: =============== ERROR MESSAGE STACK FOLLOWS ===============
RMAN-00571: ===========================================================
RMAN-03002: failure of recover command at 12/08/2015 15:23:40
RMAN-05009: Block Media Recovery requires Enterprise Edition
RMAN>
@#$%! and standard edition! Does anyone have an idea how to fix it, short
of taking the entire tablespace offline and recovering it? The tablespace
contains indexes for the central application schema and such a procedure
would amount to downtime, which would sit well with the customer. I cannot
use DBMS_REPAIR because this repairs objects and the corrupt block is not
in any segment. Backup validate datafile 36 still finds corrupt blocks,
which means that the subsequent backup will fail.
--
Mladen Gogala
Oracle DBAhttp://mgogala.freehostia.com
--
Mladen Gogala
Oracle DBAhttp://mgogala.freehostia.com
