control file and redo log

[RJP]

if I lost "all" redo log and control file, how I recover them?

without redo log, I can not recreate control file
without control file, I can not open database to switch logfile
then recreate redo log.


any help will be appreciated

[Halo]

Restore from the most recent backup and recover the db.

[Kishore]

Restore the latest backup.

Recover it.
U can recover it till the latest Arch is available.

[RJP]

as I understand, redo file never been backup. correct?

after restore the control file I only can startup mount,
drop inactive redo groups (logically) and recreate them.
since db not open I can not switch log it means I can not drop and recreate the current one. ... ???

please help me

[Halo]

Open the db and then drop/recreate the redo log files.

[RJP]

here is the real example:

>startup mount; (can not open becasue control file restored from backup)
> select group#, status from v$log;

GROUP# STATUS
---------- ----------------
1 INACTIVE
2 CURRENT
3 INACTIVE
>alter system drop logfile group 1;
>alter system add logfile group 1 '/ora.......
>alter system drop logfile group 3;
>alter system add logfile group 3 '/ora.......
> select group#, status from v$log;

GROUP# STATUS
---------- ----------------
1 UNUSED
2 CURRENT
3 UNUSED
now how I drop group2?
without system open I can not do:
alter system switch logfile;

I dooooommmmed? any solution?

[pando]

lost all control file, all redo logs

1. shutdown the database
2. Recover latest backup
3. startup nomount
4. create control file
5. recover database until cancel using backup controlfile;
this will recover using the archived logs until you stop it
6. alter database open resetlogs;

database should open successfully

You dont need to drop/recreate redo logs sicne resetlogs will create the log groups

[RJP]

PANDO,

I use this trace file to create control file (step 4) and oracle complain can not locate redo log file ... this is why I can not
move to step 5 and 6...

CREATE CONTROLFILE REUSE DATABASE "DBU2" NORESETLOGS ARCHIVELOG
MAXLOGFILES 3
MAXLOGMEMBERS 3
MAXDATAFILES 62
MAXINSTANCES 1
MAXLOGHISTORY 226
LOGFILE
GROUP 1 (
'/ora/redo1/DBU2/redoDBU201.log',
'/ora/redo2/DBU2/redoDBU201.log',
'/ora/redo3/DBU2/redoDBU201.log'
) SIZE 5M,
GROUP 2 (
'/ora/redo1/DBU2/redoDBU202.log',
'/ora/redo2/DBU2/redoDBU202.log',
'/ora/redo3/DBU2/redoDBU202.log'
) SIZE 5M,
GROUP 3 (
'/ora/redo1/DBU2/redoDBU203.log',
'/ora/redo2/DBU2/redoDBU203.log',
'/ora/redo3/DBU2/redoDBU203.log'
) SIZE 5M
DATAFILE
'/ora/dat1/oradata/DBU2/system01.dbf',
'/ora/dat1/oradata/DBU2/rbs01.dbf',
'/ora/dat2/oradata/DBU2/temp01.dbf',
....

[rcaballe]

You have all the paths created already?
As you don't have ONLINE REDO I think you should put RESETLOGS instead of NORESETLOGS
Don't forget putting the Oracle error so we can help better.

[RJP]

rcaballe, pando,

I use
>create control file ... resetlog ...
>recover database until cancel using backup controlfile;
>alter database open resetlogs

it works!!!

thanks you both are good.