prdhost:oraprd 3> sqlplus '/as sysdba'
SQL*Plus: Release 10.2.0.4.0 - Production on Tue Jun 16 23:07:11 2009
Copyright (c) 1982, 2007, Oracle. All Rights Reserved.
Connected to:
Oracle Database 10g Enterprise Edition Release 10.2.0.4.0 - 64bit Production
With the Partitioning, OLAP, Data Mining and Real Application Testing options
SQL> shutdown normal
ORA-01013: user requested cancel of current operation
SQL> shutdown abort
ORACLE instance shut down.
SQL>
The terminal dump above is probably my last shutdown abort at my current office. I've been doing shutdown abort at Oracle RDBMS all my life for doing parameter change and stuff. Today! It bite me in the ass.
I've just finished with upgrade from version 9.2.6.0 to 10.0.4.0. CRAP CRAP CRAP my database is corrupted!!!! I was getting alot of ORA-01410 and ORA-00600 in SM21 at SAP R/3 . And guess what? SAP blame the shutdown abort by me. CRAPPPPPPPPPPPPPPPPPPP!!!!!
I still stand by my experience and knowledge. shutdown abort is fucking OK! but I will not do it again for SAP systems. Why? Because they ( SAP ) will take it as an easy target for trouble shooting and the person who doing it will be their scape goat! CRAP!!!!
Here are some references on shutdown abort on oracle:-
Misconception: ‘shutdown abort’ is badShutdown AbortOh crap... the Oracle's guide against shutdown abort. Oracle Corp should know better about Oracle RDBMS rite? Hell yeah they ought to but this is yet another big company try to be politically correct. I think the don't have the balls to say "it's ok to use it" but deep in side their lab, their dba probably is doing shutdown abort while toasting champagne.
Oracle 10G Administrator Guide, chapter 3 I stand by my work. I trust what I did was the right thing but since SAP is saying it's not ok and I don't have any solid evidance that the corruption is not due the shutdown abort, I just have to be a sucker. Well everybody in my team is also agreeing to SAP anyway. Craplah this people.
In a tense time like this there will be always one smart ass who is actually know jack about the shit but try to look smart ass. In the post mortem meeting he will be saying..."shutdown abort? you should not use that. It's dangerous"
Below is from my OSS message regarding the db corruption
Hello,
your message was forwared to development support.
I have reviewed the alert_PRD.log starting with the migration path
at Sat Jun 13 12:44:13 2009
I found, that the database was 3 times aborted during the migration
and starts throwing block corrupts immediate after the it becomes
OPEN with the last crash recovery.
Sat Jun 13 14:21:48 2009
Shutting down instance (abort)
Sat Jun 13 14:21:48 2009
Shutting down instance (abort)
Sat Jun 13 14:22:27 2009
Shutting down instance (abort)
Sat Jun 13 15:11:25 2009
Starting ORACLE instance (normal)
Sat Jun 13 16:19:45 2009
Hex dump of (file 41, block 706577) in trace file /oracle/PRD/saptrace/u8.trc
Corrupt block relative dba: 0x0a4ac811 (file 41, block 706577)
For details I have attached the relevant lines from alert_PRD.log
in file alert_PRD.100000.log to this message.
The database has been corrupted by these 3 ABORTS before.
Please restore the last good backup from before the migration
and run the migration again, without to abort the instance during that
phase.
Development Support
Support Solution Center SAP - Oracle Germany
The above message really made the dogs bark at the wrong tree and made me the goat. A pack of dogs vs /me goat? goat lost!!! luckily it's not get lost!!!