Startup without redo logs?

[Knut Talman]

Hi,

in our test environment a disk group had bad superblocks.
After fsck all of the redo logs are lost. Now I added new
logfile groups and tried to drop the old (missing) ones.
Of course, I could not drop or even clear all of them, because:

ORA-01624: log 3 needed for crash recovery of thread 1
ORA-00312: online log 3 thread 1: '/opt/oracle/redo/redo31.log'
ORA-00312: online log 3 thread 1: '/opt/oracle/redo/redo32.log'

Question: Is there any possiblity to bring a database online
in this situation (with no backup)?

Regards,

Knut

[Sybrand Bakker]

"Knut Talman" <knut....@mytoys.de> wrote in message
news:3B49EA03...@mytoys.de...

startup mount
alter database open resetlogs

whatever was not there prior to the alter database command will be lost
4ever.

Hth,

Sybrand Bakker, Senior Oracle DBA

[koert54]

if an alter database open resetlogs won't work try setting the following
parameter in
your init.ora :
_allow_resetlogs_corruption=TRUE
startup the db
export the db
recreate the db
import the db

cheers
Kurt

"Knut Talman" <knut....@mytoys.de> wrote in message
news:3B49EA03...@mytoys.de...

[Knut Talman]

> startup mount
> alter database open resetlogs

SVRMGR> alter database open resetlogs;
alter database open resetlogs
*
ORA-01139: RESETLOGS option only valid after an incomplete database
recovery

Regards,

Knut

[Knut Talman]

> if an alter database open resetlogs won't work try setting the following
> parameter in
> your init.ora :
> _allow_resetlogs_corruption=TRUE

Even with setting the above parameter to TRUE I get:

ORA-00313: open failed for members of log group 2 of thread 1
ORA-00312: online log 2 thread 1: '/opt/oracle/redo/redo22.log'
ORA-27037: unable to obtain file status
SVR4 Error: 2: No such file or directory
Additional information: 3
ORA-00312: online log 2 thread 1: '/opt/oracle/redo/redo21.log'
ORA-27037: unable to obtain file status
SVR4 Error: 2: No such file or directory
Additional information: 3

Any ideas?

Regards,

Knut

[Jonathan Lewis]

Try to fake a recovery (there may be
grammar errors in the following, I'm
working from memory).

startup mount
recover database until cancel

> need file XXXXX enter filename or CANCEL

type in a garbage filename

> file 'garbage' not found entry filename or CANCEL

type in
CANCEL

Now you've done an incomplete recovery, so try

alter database open resetlogs

--
Jonathan Lewis

Host to The Co-operative Oracle Users' FAQ
http://www.jlcomp.demon.co.uk/faq/ind_faq.html

Author of:
Practical Oracle 8i: Building Efficient Databases
See http://www.jlcomp.demon.co.uk/book_rev.html

Seminars on getting the best out of Oracle
See http://www.jlcomp.demon.co.uk/seminar.html

Screensaver or Lifesaver: http://www.ud.com
Use spare CPU to assist in cancer research.

Knut Talman wrote in message <3B4AEE98...@mytoys.de>...

[Knut Talman]

> Try to fake a recovery (there may be
> grammar errors in the following, I'm
> working from memory).
>
> startup mount
> recover database until cancel
>
> > need file XXXXX enter filename or CANCEL
>
> type in a garbage filename
>
> > file 'garbage' not found entry filename or CANCEL
>
> type in
> CANCEL
>

Hi Jonathan (I am just reading your book),

> Now you've done an incomplete recovery, so try
>
> alter database open resetlogs

That worked until:

ksedmp: internal or fatal error
ORA-00600: internal error code, arguments: [2662], [0], [445381], [0],
[445871], [8389520], [], []
Current SQL statement for this session:
alter database open resetlogs

Ok, my test database is garbage, I see. But it shows how
necessary is a backup!

Thanks to all for your help,

Knut

[Knut Talman]

Hi Jonathan ( I am just reading your book),

> Try to fake a recovery (there may be
> grammar errors in the following, I'm
> working from memory).
>
> startup mount
> recover database until cancel
>
> > need file XXXXX enter filename or CANCEL
>
> type in a garbage filename
>
> > file 'garbage' not found entry filename or CANCEL
>
> type in
> CANCEL
>
> Now you've done an incomplete recovery, so try
>
> alter database open resetlogs

ksedmp: internal or fatal error

ORA-00600: internal error code, arguments: [2662], [0], [445381], [0],
[445871], [8389520], [], []
Current SQL statement for this session:
alter database open resetlogs

Ok, my test database is garbage, I see. But anyway, this shows,

how necessary is a backup!

Regards,

Knut

[koert54]

did you do a startup mount & then alter database open resetlogs using this
parameter ?

if that won't work then you'll have to contact Oracle support and ask for a
data recovery
using DUL.
this tool will create an export dmp of your DB even when it's down and
corrupted

"Knut Talman" <knut....@mytoys.de> wrote in message

news:3B4AEF8B...@mytoys.de...

[Paul Drake]

ouch.
that was the last option.
what is it, $10K for someone to arrive with the data unloader tool?

be sure that you have exhausted all options before running this drill.

I hope that this works out for you,

Paul

[Knut Talman]

> ouch.
> that was the last option.
> what is it, $10K for someone to arrive with the data unloader tool?
>
> be sure that you have exhausted all options before running this drill.
>
> I hope that this works out for you,

Thanks for your compassion, but as I mentioned in my first post,
it was only our test database. But now I know, how it would be, if
I had no backup for our production database...

Regards,

Knut

[Niall Litchfield]

make sure your business knows as well.

--
Niall Litchfield
Oracle DBA
Audit Commission UK

"Knut Talman" <knut....@mytoys.de> wrote in message

news:3B4BF8F0...@mytoys.de...

[dream_spinner]

The one time I used it, about 2 and a half years ago, Data Unloader cost
$5k for every 8 hours you had the field rep onsite, with a minimum of 16
hours.

Data Unloader is the LAST desperate step. And yes, I used it because we
had lost the online redo logs. You need FIELD support, not Tech Support
and it will create either export dump files (one for each table) or files
to be fed into sqlldr. We used export files as we had LONG columns.

Data Unloader reads the system tablespace data file(s) and creates its own
mappings to find the table data. It can work if the system tablespace is
corrupted but you'll probably lose more data.

Rachel

Paul Drake wrote:

> Paul

> koert54 wrote:

Posted via www.orafocus.com - Focusing on the World of Oracle

[JK Yao]

Before you recreate the database, should you drop or delete the
database first? Otherwise how to recreate the database while it
already exists?

Please clarify, thanks a lot.

Cheers,
JK

"koert54" <k...@k.com> wrote in message news:<yfm27.4729$z7.11...@afrodite.telenet-ops.be>...

[Norman Dunbar]

-----Original Message-----
From: jkyf...@hotmail.com (JK Yao) [mailto:jkyf...@hotmail.com]
Posted At: Friday, July 13, 2001 2:39 AM
Posted To: server
Conversation: Startup without redo logs?
Subject: Re: Startup without redo logs?

Before you recreate the database, should you drop or delete the
database first? Otherwise how to recreate the database while it
already exists?

Please clarify, thanks a lot.

Cheers,
JK

Because there is no 'drop database' command, this might be difficult !
However, to rebuild a broken one, simply build yourself a control file
trace
script and massage it to use the datafile names etc from the broken
database then create a new controlfile and open the database resetlogs.
This will (!) fix the broken db, and create new redo logs in the
process.

If the db is so broken that you can't build a controlfile trace, do it
on another database to get the meat and change the file names, etc as
appropriate. Have a look at this for details, it discusses moving a
database, but the principles are the same :
http://www.jlcomp.demon.co.uk/faq/db_move.html

Norman.

------------------------------------------------------------------------
--------
Norman Dunbar EMail: NDu...@LynxFinancialSystems.co.uk
Database/Unix administrator Phone: 0113 289 6265
Lynx Financial Systems Ltd. Fax: 0113 201 7265
URL: http://www.LynxFinancialSystems.com
------------------------------------------------------------------------
--------