Oracle FAQ | Your Portal to the Oracle Knowledge Grid |
![]() |
![]() |
Home -> Community -> Mailing Lists -> Oracle-L -> RE: Self-Healing ORA-01578's
Perhaps the change activity that resulted in creating the corruption was
uncommitted, and therefore rolled back with the shutdown immediate. The
other possibility is that the corruption occurred only in the buffer and the
block was not marked as dirty so it was not written to disk, or the
corruption occurred after the block was written to disk. In either of these
last two cases the rereading of the block from disk got you clean data.
Just a guess -- Mark D Powell --
-----Original Message-----
From: oracle-l-bounce_at_freelists.org
[mailto:oracle-l-bounce_at_freelists.org]On Behalf Of Singer, Phillip
(P.W.)
Sent: Monday, April 19, 2004 12:49 PM
To: oracle-l_at_freelists.org
Subject: Self-Healing ORA-01578's
This actually happened to me Saturday Night:
Running my 4-way Linux box (4.21 Kernel) with 9.2.0 (and using a SAN for my disks) I discovered a ORA-01578. After several hours of work with Oracle support, running dbverify, and creating test queries, we determined that there really was block corruption.
We used dbms_repair to get everything but the bad block, and were getting ready to see what would happen on a tablespace recovery, when the backup group decided to stop the instance to do a cold backup (I work for a large organization, with every task subdivided, with little communication).
When the instance came back up the block corruption was gone. This is the one part of the whole affair that I find incredible. Does someone have an idea as to how bouncing an instance can fix block corruption which was proven by dbverivy?
FAQ is at http://www.freelists.org/help/fom-serve/cache/1.htmlput 'unsubscribe' in the subject line.
-----------------------------------------------------------------
----------------------------------------------------------------
Please see the official ORACLE-L FAQ: http://www.orafaq.com
----------------------------------------------------------------
To unsubscribe send email to: oracle-l-request_at_freelists.org
![]() |
![]() |