Email: service@parnassusdata.com 7 x 24 online support!

Recover Oracle database after disk loss

Recover Oracle database after disk loss

 
PURPOSE
-------
 
This article aims at walking you through some of the common
recovery techniques after a disk failure
 
SCOPE & APPLICATION
-------------------
 
All Oracle support Analysts, DBAs and Consultants who have a role
to play in recovering an Oracle database
 
Loss due to Disk Failure
------------------------
What can we lose due to disk failure:
A) Control files
B) Redo log files
C) Archivelog files
D) Datafiles
E) Parameter file or SPFILE
F) Oracle software installation
 
Detecting disk failure
-----------------------
1) Run copy utilities like "dd" on unix
2) If using RAID mechanisms like RAID 5, parity information may mask 
    the disk failure and more vigorous check would be needed
3) As always, check the Operating system log files
4) Another obvious case would be when the disk could not be seen
    or mounted by the OS.
5) On the Oracle side, run dbv if the file affected is a datafile
6) The best way to detect disk failure is by running Hardware 
diagnostic tools and OS specific disk utilities.
 
Next Action
------------
Once the type of failure is identified, the next step is to rectify them.
 
Options could be:
(1) Replace the corrupted disk with a new one and mount them with 
     the same name (say /oracle or D:\)
(2) Replace the corrupted disk with a new one and mount them with 
     a different name (say /oracle1 as the new mount point)
(3) Decide to use another existing disk mounted with a different name
     (say /oracle_new)
 
The most common methods are (1) AND (3).
 
Oracle Recovery
---------------
Once the disk problem is sorted, the next step is to perform recovery
at the Oracle level. This would depend on the type of files that is lost (see
"Loss due to Disk Failure" section) and also on the type of disk recovery done
as mentioned in the "Next Action" section above.
 
(A) Control Files
------------------
Normally, we have multiplexing of controlfiles and they are expected to be
placed in different disks.
 
If one or more controlfile is/are lost,mount will fail as shown below:
SQL> startup
Oracle Instance started
....
ORA-00205: error in identifying controlfile, check alert log for more info
 
You can verify the controlfile copies using:
SQL> select * from v$controlfile;
 
   **If atleast one copy of the controlfile is not affected by the disk failure, 
   When the database is shutdown cleanly:
   (a) Copy a good copy of the controlfile to the missing location
   (b) Start the database 
 
   Alternatively, remove the lost control file location specified in the
   init parameter control_files and start the database.
 
   **If all copies of the controlfile are lost due to the disk failure, then:
   Check for a backup controlfile. Backup controlfile is normally taken using 
   either of the following commands:
   (a) SQL> alter database backup controlfile to '/backup/control.ctl';
    -- This would have created a binary backup of the current controlfile --
 
    -->If the backup was done in binary format as mentioned above, restore the 
       file to the lost controlfile locations using OS copying utilities.
    --> SQL> startup mount;
    --> SQL> recover database using backup controlfile;
    --> SQL> alter database open;
 
   (b) SQL> alter database backup controlfile to trace;
    -- This would have created a readable trace file containing create controlfile
    script --
 
    --> Edit the trace file created (check user_dump_dest for the location) and
        retain the SQL commands alone. Save this to a file say cr_ctrl.sql
    --> Run the script
    
    SQL> @cr_ctrl
 
    This would create the controlfile, recover database and open the database.
 
    ** If no copy of the controlfile or backup is available, then create a controlfile
    creation script using the datafile and redo log file information. Ensure that the
    file names are listed in the correct order as in FILE$.
    Then the steps would be similar to the one followed with cr_ctrl.sql script.
 
 
Note that all controlfile related SQL maintenance operations are done in the 
database nomount state
 
 
(B) Redo logs
    ---------
In normal cases, we would not have backups of online redo log files. But the 
inactive logfile changes could already have been checkpointed on the datafiles
and even archive log files may be available.
 
SQL> startup mount
     Oracle Instance Started
     Database mounted
     ORA-00313: open failed for members of log group 1 of thread 1
     ORA-00312: online log 1 thread 1: '/(path)/REDO01.LOG'
     ORA-27041: unable to open file
     OSD-04002: unable to open file
     O/S-Error: (OS 2) The system cannot find the file specified.
 
** Verify if the lost redolog file is Current or not.
     SQL> select * from v$log;
     SQL> select * from v$logfile; 
 
     --> If the lost redo log is an Inactive logfile, you can clear the logfile:
 
     SQL> alter database clear logfile '/(path)/REDO01.LOG';
 
     Alternatively, you can drop the logfile if you have atleast two other   
     logfiles:
     SQL> alter database drop logfile group 1;
 
     
     --> If the logfile is the Current logfile, then do the following:
     SQL> recover database until cancel;
         
     Type Cancel when prompted
 
     SQL>alter database open resetlogs;
 
     
     The 'recover database until cancel' command can fail with the following 
     errors:
     ORA-01547: warning: RECOVER succeeded but OPEN RESETLOGS would get error 
     below
     ORA-01194: file 1 needs more recovery to be consistent
     ORA-01110: data file 1: '/(Path)/SYSTEM01.DBF'
 
     In this case , restore an old backup of the database files and apply the
     archive logs to perform incomplete recovery.
     --> restore old backup
     SQL> startup mount
     SQL> recover database until cancel using backup controlfile;
     SQL> alter database open resetlogs;
 
 
If the database is in noarchivelog mode and if ORA-1547, ORA-1194 and ORA-1110 errors occur, then you would have restore from an old backup and start the database.
 
 
Note that all redo log maintenance operations are done in the database mount state
 
 
(C) Archive logs
-----------------
If the previous archive log files alone have been lost, then there is not much
to panic.
** Backup the current database files using hot or cold backup which would ensure
that you would not need the missing archive logs
 
(D) Datafiles
--------------
This obviously is the biggest loss.
 
(1) If only a few sectors are damaged, then you would get ora-1578 when 
accessing those blocks.
 --> Identify the object name and type whose block is corrupted by querying dba_extents
 --> Based on the object type, perform appropriate recovery
 --> Check metalink Note:28814.1 for resolving this error
 
(2) If the entire disk is lost, then one or more datafiles may need to be 
recovered . 
  SQL> startup
  ORACLE instance started.
  ...
  Database mounted.
  ORA-01157: cannot identify/lock data file 3 - see DBWR trace file
  ORA-01110: data file 3: '/(path)/USERS01.DBF'
 
Other possible errors are ORA-00376 and ORA-1113
 
The views and queries to identify the datafiles would be:
   SQL> select file#,name,status from v$datafile;
   SQL> select file#,online,error from v$recover_file;
 
 
** If restoring to a replaced disk mounted with the same name, then :
  (1) Restore the affected datafile(s) using OS copy/restore commands from the 
      previous backup
  (2) Perform recovery based on the type of datafile affected namely SYSTEM, 
      ROLLBACK or UNDO, TEMP , DATA or INDEX.
  (3) The recover commands could be 'recover database', 'recover tablespace'
      or 'recover datafile' based on the loss and the database state
 
** If restoring to a different mount point, then :
  (1) Restore the files to the new location from a previous backup
  (2) SQL> STARTUP MOUNT
  (3) SQL> alter database rename file '/old path_name' to 'new path_name';     
      -- Do this renaming for all datafiles affected. --
  (4) Perform recovery based on the type of datafile affected namely SYSTEM, 
      ROLLBACK or UNDO, TEMP , DATA or INDEX.
  (5) The recover commands could be 'recover database', 'recover tablespace'
      or 'recover datafile' based on the loss and the database state
 
The detailed steps of recovery based on the datafile lost and the Oracle error 
are outlined in the articles referenced at the end of this note.
 
 
  NOARCHIVELOG DATABASE
  =====================
  The loss mentioned in (A),(B) and (D) would be different in this case
  wherever archive logs are involved. 
 
  We will discuss the datafile loss scenarios here:
 
  (a) If the datafile lost is a SYSTEM datafile, restore the complete
      database from the previous backup and start the database.
  (b) If the datafile lost is Rollback related datafile with active transactions,
      restore from the previous backup and start the database.
  (c) If the datafile contains rollback with no active rollback segments, you can
      offline the datafile (after commenting the rollback_segments parameter 
      assuming that they are private rollback segments) and open the database. 
  (d) If the datafile is temporary, offline the datafile and open the database. 
      Drop the tablespace and recreate the tablespace.
  (e) If the datafile is DATA or INDEX, 
      **Offline the tablespace and start the database.
      **If you have a previous backup, restore it to a separate location.
      **Then export the objects in the affected tablespace ( using User or 
        table level export).
      **Create the tablespace in the original database.
      **Import the objects exported above.
 
      If the database is 8i or above, you can also use Transportable tablespace
      feature.
 
 
(E) Parameter file
    ---------------
This is not a major loss and can be easily restored. Options are:
  (1) If there is a backup, restore the file
  (2) If there is no backup, copy sample file or create a new file and add the 
      required parameters. Ensure that the parameters db_name, control_files,
       db_block_size, compatible are set correctly
  (3) If the spfile is lost, you can create it from the init parameter file if it is available
 
 
(F) Oracle Software Installation
    ----------------------------
There are two ways to recover from this scenario:
  (1) If there is a backup of the Oracle home and Oracle Inventory, restore
      them to the respective directories. Note if you change the Oracle Home, 
      the inventory would not be aware of thid new path and you would not be
      able to apply patchsets. Also restore to the same OS user and group.
 
  (2) Perform a fresh Install, bringing it to the same patchset level
 
 
PRACTICAL SCENARIO
==================
 
In most cases, when a disk is lost, more than one type of file could be lost.
The recovery in this scenario would be:
  (1) A combination of each of these data loss recovery scenarios
  (2) Perform entire database restore from the more recent backup and apply
      archive logs to perform recovery. This is a highly preferred method 
      but could be time consuming.