Download the presentation

I was aware of the ADR stuff of the server installations, but my client installation seemed to not have those directories in place

CLIENT-> adrci

ADRCI: Release 11.2.0.1.0 – Production on Fri Jun 18 11:49:08 2010

Copyright (c) 1982, 2009, Oracle and/or its affiliates.  All rights reserved.

No ADR base is set
adrci> show homes
No ADR homes are set
adrci> exit
CLIENT->

I googled again and again until i hit with this blog, where he says that we need to mention

DIAG_ADR_ENABLED=off

in the sqlnet.ora to have the traces files written to “TRACE_DIRECTORY_CLIENT”.

$ expdp help=y

:

COMPRESSION

Reduce the size of a dump file.

Valid keyword values are: ALL, DATA_ONLY, [METADATA_ONLY] and NONE.

:

 I did a comparison between  11G’s expdp, 11G’s expdp + compression=ALL, 11G’s expdp + GZIP and exp + pipe node + GZIP on a database of size 2GB(approx) and below table describes the size of the dump file.

 There wasn’t much notable difference in the time taken. Whatever, I would  go for the 11g expdp with compression=all than erstwhile exp.

select sum(bytes)/1024/1024 from dba_segments where segment_name='A' and segment_type='TABLE';

This gives the total segment size. It could be that the table is either full, sparsely populated or even empty, still there would not be any change for the segment size. We could assign 1GB for a table which would actually need only 10MB, but who lose? Assigning more(very much extra) space than necessary has negative impact not only on space, but also on the performance. Each time a full table scan is performed, oracle retrieves all the blocks until the High Water Mark of the table. It could be that most of these blocks are empty. The same for indexes also.  The segment advisor suggests the candidate tables for shrink operation or you could manually select your known tables and check its actual consumed space using dbms_space.unused_space() package.

spool d:\oracle\table.log

DECLARE
 alc_bks     NUMBER;
 alc_bts     NUMBER;
 unsd_bks    NUMBER;
 unsd_bts    NUMBER;
 luefi       NUMBER;
 luebi       NUMBER;
 lub         NUMBER;
BEGIN
 FOR c1_row IN (SELECT table_name
 FROM   dba_tables where owner='JJ') LOOP
 DBMS_SPACE.UNUSED_SPACE (
 segment_owner               => 'JJ'
 ,   segment_name                => c1_row.table_name
 ,   segment_type                => 'TABLE'
 ,   total_blocks                => alc_bks
 ,   total_bytes                 => alc_bts
 ,   unused_blocks               => unsd_bks
 ,   unused_bytes                => unsd_bts
 ,   last_used_extent_file_id    => luefi
 ,   last_used_extent_block_id   => luebi
 ,   last_used_block             => lub
 );
DBMS_OUTPUT.PUT_LINE( c1_row.table_name || ',' || alc_bts ||','|| unsd_bts );
 END LOOP;
EXCEPTION
 WHEN OTHERS THEN
 DBMS_OUTPUT.PUT_LINE(SUBSTR(SQLERRM,1,250));
END;
/

PS:i didn't write the below script, but modified to separate output by a delimit.

To see how much space a character occupies, one can use the oracle vsize(‘<string>’) or dump(‘string’) functions. Both these functions gives the size in bytes.

Let’s see how much space(in bytes) Oracle Db with  AL32UTF8 characterset requires tp store the word ‘tokyo’ in English and Japanese (Hiragana, Full Width Katakana, Half Width Katakana and Kanji).

English
select vsize(‘tokyo’) from dual;
5
select dump(‘tokyo’) from dual;
Typ=96 Len=5: 116,111,107,121,111

Japanese-Hiragana
select vsize(‘とうきょう’) from dual;
15
select dump(‘とうきょう’) from dual;
Typ=96 Len=15: 227,129,168,227,129,134,227,129,141,227,130,135,227,129,134

Japanese-Full Width Katakana
select vsize(‘トウキョウ’) from dual;
15
select dump(‘トウキョウ’) from dual;
Typ=96 Len=15: 227,131,136,227,130,166,227,130,173,227,131,167,227,130,166

Japanese-Half Width Katakana
select vsize(‘ﾄｳｷｮｳ’) from dual;
15
select dump(‘ﾄｳｷｮｳ’) from dual;
Typ=96 Len=15: 239,190,132,239,189,179,239,189,183,239,189,174,239,189,179

Japanese-Kanji
select vsize(‘東京’) from dual;
6
select dump(‘東京’) from dual;
Typ=96 Len=6: 230,157,177,228,186,172

:
Errors in file /u01/app/oracle/admin/SHCL1/bdump/shcl1n02_j004_27414.trc:
ORA-07445: exception encountered: core dump [kglobcl()+412] [SIGSEGV] [Address not mapped to object] [0x49415C002] [] []
Tue Nov  3 00:00:27 2009
Trace dumping is performing id=[cdmp_20091103000027]
Tue Nov  3 00:00:37 2009
Errors in file /u01/app/oracle/admin/SHCL1/bdump/shcl1n02_pmon_3278.trc:
ORA-07445: [kglobcl()+412] [SIGSEGV] [Address not mapped to object] [0x49415C002] [] []
Tue Nov  3 00:00:51 2009
:
:
Tue Nov  3 00:00:54 2009
MMAN: terminating instance due to error 472
Instance terminated by MMAN, pid = 3368
Wed Nov  4 10:44:47 2009
:

I checked

$ crsctl check crs
OK (running successfully)
$ srvctl status nodeapps -n myjpsuolicdbd02
OK (all resources running successfully)
$ srvctl status database -d SHCL1_PRMY
Instance SHCL1N01 is running on node myjpsuolicdbd01
PRKO-2015 : Error in checking condition of instance on node: myjpsuolicdbd02
$

Since only instance 2 was down, i tried to start it and i got the below error.

$ srvctl start instance -d SHCL1_PRMY -i SHCL1N02
PRKP-1001 : Error starting instance SHCL1N02 on node myjpsuolicdbd02
CRS-1028: Dependency analysis failed because of:
CRS-0223: Resource ‘ora.SHCL1_prmy.SHCL1N02.inst’ has placement error.
$

The crsd.log file too didn’t have much(to the point) information.

2009-11-04 10:18:30.627: [  CRSRES][2970821] CRS-1028: Dependency analysis failed because of:'Resource in UNKNOWN state: ora.SHCL1_prmy.SHCL1N02.inst'

2009-11-04 10:21:23.269: [  CRSRES][2970843] StopResource: setting CLI values
2009-11-04 10:21:23.340: [  CRSRES][2970843] Attempting to stop `ora.SHCL1_prmy.SHCL1N02.inst` on member `myjpsuolicdbd02`
2009-11-04 10:21:30.478: [  CRSAPP][2970843] StopResource error for ora.SHCL1_prmy.SHCL1N02.inst error code = 1
2009-11-04 10:21:30.502: [  CRSRES][2970843] Stop of `ora.SHCL1_prmy.SHCL1N02.inst` on member `myjpsuolicdbd02` succeeded.
2009-11-04 10:21:49.867: [  CRSRES][2970861] startRunnable: setting CLI values
2009-11-04 10:21:49.895: [  CRSRES][2970861] Attempting to start `ora.SHCL1_prmy.SHCL1N02.inst` on member `myjpsuolicdbd02`
2009-11-04 10:21:55.019: [  CRSAPP][2970861] StartResource error for ora.SHCL1_prmy.SHCL1N02.inst error code = 1
2009-11-04 10:22:00.583: [  CRSAPP][2970861] StopResource error for ora.SHCL1_prmy.SHCL1N02.inst error code = 1
2009-11-04 10:22:00.592: [  CRSRES][2970861] X_OP_StopResourceFailed : Stop Resource failed
(File: rti.cpp, line: 1803

2009-11-04 10:22:00.593: [  CRSRES][2970861][ALERT] `ora.SHCL1_prmy.SHCL1N02.inst` on member `myjpsuolicdbd02` has experienced an unrecoverable failure.
2009-11-04 10:22:00.593: [  CRSRES][2970861] Human intervention required to resume its availability.

Clusterware and nodeapps were up and running but db instance on node 2.

I could startup the instance from sqlplus, but this would not start the db services etc ora.SHCL1_prmy.SHCL1N02.inst was not ONLINE and would be of no use in a RAC.
Didn’t have any clue, until I came across the logs under /u01/app/oracle/product/10.2.0/db_1/log/<hostname>/racg/imonSHCL1_prmy.log and imon_SHCL1_prmy.log.

The files imonSHCL1_prmy.log and imon_SHCL1_prmy.log contains details of the starting up of the DB using SRVCTL command.

SQL*Plus: Release 10.2.0.4.0 - Production on Wed Nov 4 18:59:48 2009
Copyright (c) 1982, 2007, Oracle.  All Rights Reserved.

Enter user-name: ERROR:
ORA-01031: insufficient privileges

Enter user-name: SP2-0306: Invalid option.

2009-11-04 18:59:48.968: [    RACG][176] [28813][176][ora.SHCL1_prmy.SHCL1N02.inst]: Usage: CONN[ECT] [logon] [AS {SYSDBA|SYSOPER}]
where <logon>  ::= <username>[/<password>][@<connect_identifier>] | /
Enter user-name: Enter password:
ERROR:
ORA-01005: null password given; logon denied

2009-11-04 18:59:48.968: [    RACG][176] [28813][176][ora.SHCL1_prmy.SHCL1N02.inst]: SP2-0157: unable to CONNECT to ORACLE after 3 attempts, exiting SQL*Plus
2009-11-04 18:59:48.968: [    RACG][176] [28813][176][ora.SHCL1_prmy.SHCL1N02.inst]: clsrcexecut: env _USR_ORA_PFILE=
2009-11-04 18:59:48.968: [    RACG][176] [28813][176][ora.SHCL1_prmy.SHCL1N02.inst]: clsrcexecut: cmd = /u01/app/oracle/product/10.2.0/db_1/bin/racgeut -e _USR_ORA_DEBUG=0 -e ORACLE_SID=SHCL1N02 520 /u01/app/oracle/product/10.2.0/db_1/bin/racgmdb -d abort

CAUSE:

Due to security concern, recently i added “ SQLNET.AUTHENTICATION_SERVICES=(NONE) ” to sqlnet.ora and this line would insist on providing the password explicitly. It doesn’t allow to login using “sqlplus / as sysdba”. Hence the SRVCTL failed.

So,

After removing this line from sqlnet.ora, i could start the instance.

SOLUTION:

1)Commented the entry in sqlnet.ora
2)crs_stat -u => check for  “ora.SHCL1_prmy.SHCL1N02.inst”, the STATE should be OFFLINE before we ‘start’ again.

NAME=ora.SHCL1_prmy.SHCL1N02.inst
TYPE=application
TARGET=ONLINE
STATE=UNKNOWN on myjpsuolicdbd02

If the STATE is UNKNOWN, ‘start’ing the instance would error out. So stop it by force.

2)crs_stop -f ora.SHCL1_prmy.SHCL1N02.inst
3)crs_stat -u => to confirm that the TARGET is ONLINE and STATE is OFFLINE for ora.SHCL1_prmy.SHCL1N02.inst
4)srvctl start instance -d SHCL1_PRMY -i SHCL1N02
Raised an SR for the ORA 7445 error and let me see what they have in stock for me

Oracle documentation says:

http://download.oracle.com/docs/cd/B19306_01/server.102/b14232/tgvsgc.htm#sthref36

The Heterogeneous Services component in the Oracle database server talks to a Heterogeneous Services agent process which, in turn, talks to the non-Oracle system. We can conceptually divide the code into three parts:

1)The Heterogeneous Services component in the Oracle database server.
Most of the processing related to heterogeneous connectivity is done in this module.
2)Agent generic code.
This is code in the agent that is generic to all Heterogeneous Services products. This consists, for the most part, of code to communicate with the database and multithreading support.
3)The driver.
This is the module that communicates with the non-Oracle system. It is used to map calls from the Heterogeneous Services onto the native API of the non-Oracle system and it is non-Oracle system specific.

Connecting to PostgreSQL from Oracle using an ODBC data source

To connect to a PostgreSQL from Oracle, we will have to use the Oracle Generic Connectivity Agent(hsodbc).
A simple call would be like: Oracle DB -> Oracle Heterogeneous Service -> Oracle Generic Connectivity Agent -> ODBC Driver for PostgreSQL -> PostgreSQL DB

http://download.oracle.com/docs/cd/B19306_01/server.102/b14232/gencon.htm#i1005976

Oracle does not provide the ODBC driver for PostgreSQL, it should be acquired from third party. I downloaded the psqlOBDC driver from http://www.postgresql.org/download/products/2 .

1.PostgreSQL Server [PostgreSQL 8.3.5 on Windows]
1.1)Create the DB and user, which would be used by oracle to make connection.

2.Oracle Server [Oracle 10.2.0.1 on Windows]
2.1)Install Oracle(standard edition also will do)
2.2)Create a database
2.3)Go to $ORACLE_HOME/hs/admin/ and confirm the files
2.4)Download and install the psqlOBDC driver from postgresql.org
2.5)Create the ODBC data source
2.6)Configure initHSODBC.ora ($ORACLE_HOME/hs/admin/) and listener.ora and tnsnames.ora ($ORACLE_HOME/network/admin/)
2.7)Create the database link

This post will cover only the details to be done at the Oracle side.

2.1 – 2.3:
I installed oracle 10.2.0.1 standard edition and created a generic database.
The $ORACLE_HOME/hs/admin/ would contain the files inithsodbc.ora,inithsoledb.ora,listener.ora.sample and tnsnames.ora.sample.
The inithsodbc.ora is the one of our interest. We will configure it later.

2.4:
I downloaded the “psqlodbc_08_04_0100.zip”(the latest when i installed) from http://www.postgresql.org/ftp/odbc/versions/msi/ .
(I could not access this site using IE and hence used Firefox.)
The installation is nothing big the typical “click next button”

Click Next -> Accept the terms -> Next -> Install -> Finish

2.5:
Go to Control Panel -> Administrative Tools -> Data Sources -> System DSN -> Add

Select the “PostgreSQL ANSI” driver -> Finish.

Enter the ODBC Datasource name ( i gave PG), Server Name (put the IP address of PostgreSQL DB server), Database Name (PostgreSQL DB name) and User Name (PostgreSQL userid) and password.
Click “Test” to test the connection to PostgreSQL DB and it would show “Connection Succesful” and Save. Now we have an ODBC Data Source called “PG”.

2.6:
Now we have the datasource PG which would use the ODBC driver psqlODBC.
We need to configure 3 files to complete the connectivity
a)initHSODBC.ora
b)listener.ora
c)tnsnames.ora

a.1)
Go to $ORACLE_HOME/hs/admin/ and make a copy of the initHSODBC.ora and name it init<SID>.ora, where sid is the system identifier you want to use for the instance of the non-Oracle system to which the agent connects and i renamed the file to “initPG.ora“. Edit the file and provide the ODBC Data Source Name which we created in step 2.5.

# This is a sample agent init file that contains the HS parameters that are
# needed for an ODBC Agent.
#
# HS init parameters
#
HS_FDS_CONNECT_INFO = PG                        <<- ODBC Data Source Name
HS_FDS_TRACE_LEVEL = ON
#
# Environment variables required for the non-Oracle system
#
#set <envvar>=<value>

b.1)

Listener.ora ($ORACLE_HOME/network/admin/)

 SID_LIST_LISTENER =
 (SID_LIST =
 (SID_DESC =
 (SID_NAME = PG)          <<- Should be same as the SID in $ORACLE_HOME/hs/admin/initSID.ora 
 (ORACLE_HOME = D:\oracle\product\10.2.0\db_1)
 (PROGRAM = hsodbc)       <<- Set the program as "hsodbc"
 )
 (SID_DESC =
 (SID_NAME = PLSExtProc)
 (ORACLE_HOME = D:\oracle\product\10.2.0\db_1)
 (PROGRAM = extproc)
 )
 )

LISTENER =
 (DESCRIPTION_LIST =
 (DESCRIPTION =
 (ADDRESS = (PROTOCOL = IPC)(KEY = PNPKEY))        <<- Set key as PNPKEY
 (ADDRESS = (PROTOCOL = TCP)(HOST = 192.168.20.20)(PORT = 1521))
 )
)

Restart the listener.

The sid that you add to the listener.ora file must match the sid in an initsid.ora file, because the agent spawned by the listener searches for a matching initsid.ora file. When you copy and rename the initsid.ora file, ensure that it remains in the $ORACLE_HOME/hs/admin directory

c.1)

Add an entry for the PostgreSQL DB into the tnsnames.ora for the DB link to work.

 PG =
 (DESCRIPTION=
 (ADDRESS=(PROTOCOL=tcp)(HOST=192.168.20.20)(PORT=1521))
 (CONNECT_DATA=(SID=PG))
 (HS=OK)
 )

Confirm the “tnsping PG” works.

2.7:
Create the database link in oracle

 create public database link topg connect to "postgreuser" identified by "password" using 'PG';

Confirm the DBlink to PostgreSQL using this query.

 select * from all_tables@topg

Unlike Oracle, PostgreSQL is case sensitive. So for user tables(other than all_tables) use the double quotes like

 select * from "tbl_user_data"@topg;

References:

http://www.databasejournal.com/features/oracle/article.php/10893_3442661_2/Making-a-Connection-from-Oracle-to-SQL-Server.htm

http://birijan.com.np/?q=Oracle+Heterogeneous+Services+(Accessing+ODBC+datasource)

On Windows

make_conn.bat
——————

FOR %%A in (1 2 3) DO start sqlplus userid/passwd@tns_entry @conn.sql

conn.sql
———–

select sysdate from dual;
exec dbms_lock.sleep(10);
exit;

Save both the files in the same folder with respective names.
Now double click the make_conn.bat and this would create 3 connections to the db and will exceute the conn.sql in each. The contents of the conn.sql are self explanatory. To make more connections all you need to do is to add more numerics to the FOR statement. Changing the make_conn.bat to (1 2 3 1 2 3 1 2 3 4 5 6) would create 12 connections and the more numbers you add, more threads(connections) are created.

If you want to trigger the connections from cmd.exe then you will need to modify the FOR statement like :

FOR %A in (1 2 3) DO start sqlplus userid/passwd@burp @conn.sql

Note that there is only one % symbol.

On Unix/Linux

make_conn.sh
—————–

#!/bin/ksh
#
#Make n number of connections to db
#
#
echo "Enter the number of connections to make :\c"
read NOF;
i=0
while [[ $i -lt $NOF ]];
do
 sqlplus -s userid/passwd @conn.sql >/dev/null &
 i=$i+1
done
exit;

conn.sql
——–

select sysdate from dual;
exec dbms_lock.sleep(10);
exit;

$chmod 755 make_conn.sh
$make_conn.sh

The “&” in the while loop will run the process in background and will direct the o/p to /dev/null, so you will not see any message in your terminal unless there is an error.

In both cases, you can test your number of connections by

select count(8) from v$session where username=’USERID’;

The destination will be a single instnace.
Destination :
Unique Name    : SHCL1REC
Db_Name           : SHCL1
SID                        : SHCL1REC

1] Prepare SHCL1DR for taking RMAN backup

• 1.a] Create directories to store the RMAN backup

 mkdir /u04/RMAN_BACKUPS

• 1.b] Configure the RMAN persistant parameters to point to the backup directory and enable controlfile autobackup.

export ORACLE_SID=SHCL1DR2

 rman target /

 CONFIGURE CONTROLFILE AUTOBACKUP ON;
 CONFIGURE CHANNEL 1 DEVICE TYPE DISK FORMAT '/u04/RMAN_BACKUPS/%d_%I_%U' connect 'sys/pasd@SHCL1';
 CONFIGURE CHANNEL 2 DEVICE TYPE DISK FORMAT '/u04/RMAN_BACKUPS/%d_%I_%U' connect 'sys/pasd@SHCL1';
 CONFIGURE CHANNEL 3 DEVICE TYPE DISK FORMAT '/u04/RMAN_BACKUPS/%d_%I_%U' connect 'sys/pasd@SHCL1';
 CONFIGURE CHANNEL 4 DEVICE TYPE DISK FORMAT '/u04/RMAN_BACKUPS/%d_%I_%U' connect 'sys/pasd@SHCL1';
 CONFIGURE CHANNEL 5 DEVICE TYPE DISK FORMAT '/u04/RMAN_BACKUPS/%d_%I_%U' connect 'sys/pasd@SHCL1';
 CONFIGURE CHANNEL 6 DEVICE TYPE DISK FORMAT '/u04/RMAN_BACKUPS/%d_%I_%U' connect 'sys/pasd@SHCL1';

The backupsets will be created under /u04/RMAN_BACKUPS/ and the controlfile would be backed up at the default location.

 RMAN> show all;
 using target database control file instead of recovery catalog
 RMAN configuration parameters are:

 CONFIGURE RETENTION POLICY TO RECOVERY WINDOW OF 90 DAYS;
 CONFIGURE BACKUP OPTIMIZATION ON;
 CONFIGURE DEFAULT DEVICE TYPE TO DISK;
 CONFIGURE CONTROLFILE AUTOBACKUP ON;
 CONFIGURE CONTROLFILE AUTOBACKUP FORMAT FOR DEVICE TYPE DISK TO '%F'; # default
 CONFIGURE DEVICE TYPE DISK BACKUP TYPE TO COMPRESSED BACKUPSET PARALLELISM 6;
 CONFIGURE DATAFILE BACKUP COPIES FOR DEVICE TYPE DISK TO 1;
 CONFIGURE ARCHIVELOG BACKUP COPIES FOR DEVICE TYPE DISK TO 1;
 CONFIGURE CHANNEL 1 DEVICE TYPE DISK FORMAT   '/u04/RMAN_BACKUPS/%d_%I_%U' CONNECT '*';
 CONFIGURE CHANNEL 2 DEVICE TYPE DISK FORMAT   '/u04/RMAN_BACKUPS/%d_%I_%U' CONNECT '*';
 CONFIGURE CHANNEL 3 DEVICE TYPE DISK FORMAT   '/u04/RMAN_BACKUPS/%d_%I_%U' CONNECT '*';
 CONFIGURE CHANNEL 4 DEVICE TYPE DISK FORMAT   '/u04/RMAN_BACKUPS/%d_%I_%U' CONNECT '*';
 CONFIGURE CHANNEL 5 DEVICE TYPE DISK FORMAT   '/u04/RMAN_BACKUPS/%d_%I_%U' CONNECT '*';
 CONFIGURE CHANNEL 6 DEVICE TYPE DISK FORMAT   '/u04/RMAN_BACKUPS/%d_%I_%U' CONNECT '*';
 CONFIGURE MAXSETSIZE TO UNLIMITED;
 CONFIGURE ENCRYPTION FOR DATABASE OFF; # default
 CONFIGURE ENCRYPTION ALGORITHM 'AES128'; # default
 CONFIGURE ARCHIVELOG DELETION POLICY TO APPLIED ON STANDBY;
 CONFIGURE SNAPSHOT CONTROLFILE NAME TO '/u01/app/oracle/product/10.2.0/db_1/dbs/snapcf_SHCL1DR2.f'; # default
RMAN>

• 1.c] Take the backup from on of the node

rman target /
 crosscheck archivelog all;
 backup database plus archivelog;

I used the below script to run RMAN in background mode.

 T02[SHCL1DR2]$ cat /export/home/oraprd/bkp.sh
 #!/bin/sh
 rman target / <<EOF
 backup database plus archivelog;
 exit;
 EOF
 exit
 T02[SHCL1DR2]$

 T02[SHCL1DR2]$ nohup /export/home/orapd/bkp.sh > /export/home/orapd/bkp. 2>&1 &    

2] Prepare environmet for SHCL1REC

• 2.a] Create the directories to hold the controlfile,redo logs,datafiles and archivelogs

(I used the same machine, hence restored to a different location)
mkdir /u02/oracle/shcl1rec/

This directory would hold all files pertaining to this new instance.
(If you are going to use this restored instance for Prod or Dev activities, use separate directories.
All i wanted was to test the backups, hence restoring them to same dir for simplicity)

 mkdir /u01/app/oracle/admin/SHCL1REC/adump        
 mkdir /u01/app/oracle/admin/SHCL1REC/bdump
 mkdir /u01/app/oracle/admin/SHCL1REC/cdump
 mkdir /u01/app/oracle/admin/SHCL1REC/udump
 mkdir /u02/oracle/shcl1rec/archivelogs

• 2.b] Restore the spfile from the controlfile backup

 export ORACLE_SID=SHCL1REC
 rman target /
 startup nomount
 restore spfile from '/u02/oracle/shcl1/flashbacklog/SHCL1DR/autobackup/2009_07_23/o1_mf_s_692991347_56j874gd_.bkp';

[Make sure that you have enough memory for the SGA as that of the Source DB, else you cannot mount using this
spfile as its a clone of the source spfile. If not enough memory, then copy the initSHCL1DR2.ora to initSHCL1REC.ora]

• 2.c] Modify the initSHCL1REC.ora file

Change the path for controlfiles,*dumps,archive_logs.
Comment the cluster parameters
Change the db_unique_name,service_names and domain
Change the SGA and PGA
Comment/change those parameters that refer the RAC instances.

• 2.d] Using this new init.ora file, start the instance in nomount mode.

• 2.e] Restore the controlfile from the backup

export ORACLE_SID=SHCL1REC

 rman target /    
 restore controlfile from '/u02/oracle/shcl1/flashbacklog/SHCL1DR/autobackup/2009_07_23/o1_mf_s_692991347_56j874gd_.bkp';
 alter database mount;

RMAN> restore controlfile from '/u02/oracle/shcl1/flashbacklog/SHCL1DR/autobackup/2009_07_22/o1_mf_s_692897986_56ff1mk7_.bkp';
Starting restore at 22-JUL-09
allocated channel: ORA_DISK_1
channel ORA_DISK_1: sid=647 devtype=DISK
channel ORA_DISK_1: restoring control file
channel ORA_DISK_1: restore complete, elapsed time: 00:00:04
output filename=/u02/oracle/shcl1rec/control1
output filename=/u02/oracle/shcl1rec/control2
output filename=/u02/oracle/shcl1rec/control3
Finished restore at 22-JUL-09
RMAN>
RMAN> alter database mount;
database mounted
released channel: ORA_DISK_1
RMAN>
 

Confirm that the control files are created under /u02/oracle/shcl1rec/.

• 2.f] Create the password file for the new instance.

RMAN need to connect to the new instance as sysdba, so create the password file.
For simplicity copy the password file of the source.

 cp $ORACLE_HOME/dbs/orapwSHCL1DR2 $ORACLE_HOME/dbs/orapwSHCL1REC

• 2.g] Create the tns entry

RMAN would need dedicated connection, so if dispatchers are configured, then make sure you have SERVER=DEDICATED
embedded in the tnsenrty.

SHCL1REC =
 (DESCRIPTION =
 (ADDRESS = (PROTOCOL = TCP)(HOST=10.28.19.6)(PORT= 1521))
 (CONNECT_DATA = (SERVICE_NAME = SHCLREC.s.com) (SERVER = DEDICATED))
)   

• 2.h] Configure the persistant parameters for the new instance

Configure the rman persistant parameter (to change the connection string)

 CONFIGURE CHANNEL 1 DEVICE TYPE DISK FORMAT '/u04/RMAN_BACKUPS/%d_%I_%U' connect 'sys/pasd@SHCL1REC';
 CONFIGURE CHANNEL 2 DEVICE TYPE DISK FORMAT '/u04/RMAN_BACKUPS/%d_%I_%U' connect 'sys/pasd@SHCL1REC';
 CONFIGURE CHANNEL 3 DEVICE TYPE DISK FORMAT '/u04/RMAN_BACKUPS/%d_%I_%U' connect 'sys/pasd@SHCL1REC';
 CONFIGURE CHANNEL 4 DEVICE TYPE DISK FORMAT '/u04/RMAN_BACKUPS/%d_%I_%U' connect 'sys/pasd@SHCL1REC';
 CONFIGURE CHANNEL 5 DEVICE TYPE DISK FORMAT '/u04/RMAN_BACKUPS/%d_%I_%U' connect 'sys/pasd@SHCL1REC';
 CONFIGURE CHANNEL 6 DEVICE TYPE DISK FORMAT '/u04/RMAN_BACKUPS/%d_%I_%U' connect 'sys/pasd@SHCL1REC';    

• 2.i] Prepare the RMAN restore script and restore it.

Since we are using a different path for datafiles, we need to specify it using the ‘SET NEWNAME’ and rename the
redo log files.

I used the below pl/sql to generate the script.

sqlplus / as sysdba
set head off
set feedback off
set pages 1000
set lines 160
set serveroutput on
--The below plsql is for creating script for renaming the datafile
declare
 dfile varchar2(100);
 fileno varchar2(100);
 begin
 for d_path in (select file#,name from v\$datafile)
 loop
 dfile :=SUBSTR(d_path.name, INSTR(d_path.name,'/', 1, 5)+1);
 fileno := d_path.file#;
 dbms_output.put_line('set newname for datafile '||fileno||' to ''/u02/oracle/shcl1rec/'||dfile||''';');
 end loop;
 end;
/
--The below plsql is for creating script for renaming the log files
declare
 rlog varchar2(100);
 begin
 for l_path in (select member from v\$logfile)
 loop
 rlog :=SUBSTR(l_path.member, INSTR(l_path.member,'/', 1, 5)+1);
 dbms_output.put_line('sql "alter database rename file '''''||l_path.member||''''' to ''''/u02/oracle/shcl1rec/'||rlog||''''' ";');
 --dbms_output.put_line(rlog);
 end loop;
 end;
/

Below is the script that i used

T02[SHCL1DR2]$ cat /export/home/oraprd/restore.sh        
#!/bin/ksh
rman target / <<EOF    
set newname for datafile 1 to '/u02/oracle/shcl1rec/system01.dbf';
set newname for datafile 2 to '/u02/oracle/shcl1rec/undo01.dbf';
set newname for datafile 3 to '/u02/oracle/shcl1rec/sysaux01.dbf';
set newname for datafile 4 to '/u02/oracle/shcl1rec/undo02.dbf';
sql "alter database rename file ''/u04/oracle/shcl1/onlinelog/redo1_1.log'' to ''/u02/oracle/shcl1rec/redo1_1.log'' ";
sql "alter database rename file ''/u03/oracle/shcl1/onlinelog/redo1_2.log'' to ''/u02/oracle/shcl1rec/redo1_2.log'' ";
sql "alter database rename file ''/u04/oracle/shcl1/onlinelog/redo2_1.log'' to ''/u02/oracle/shcl1rec/redo2_1.log'' ";
sql "alter database rename file ''/u03/oracle/shcl1/onlinelog/redo2_2.log'' to ''/u02/oracle/shcl1rec/redo2_2.log'' ";
sql "alter database rename file ''/u04/oracle/shcl1/onlinelog/redo3_1.log'' to ''/u02/oracle/shcl1rec/redo3_1.log'' ";
sql "alter database rename file ''/u03/oracle/shcl1/onlinelog/redo3_2.log'' to ''/u02/oracle/shcl1rec/redo3_2.log'' ";
sql "alter database rename file ''/u04/oracle/shcl1/onlinelog/redo4_1.log'' to ''/u02/oracle/shcl1rec/redo4_1.log'' ";
sql "alter database rename file ''/u03/oracle/shcl1/onlinelog/redo4_2.log'' to ''/u02/oracle/shcl1rec/redo4_2.log'' ";
sql "alter database rename file ''/u04/oracle/shcl1/onlinelog/redo5_1.log'' to ''/u02/oracle/shcl1rec/redo5_1.log'' ";
sql "alter database rename file ''/u03/oracle/shcl1/onlinelog/redo5_2.log'' to ''/u02/oracle/shcl1rec/redo5_2.log'' ";
sql "alter database rename file ''/u04/oracle/shcl1/onlinelog/redo6_1.log'' to ''/u02/oracle/shcl1rec/redo6_1.log'' ";
sql "alter database rename file ''/u03/oracle/shcl1/onlinelog/redo6_2.log'' to ''/u02/oracle/shcl1rec/redo6_2.log'' ";
sql "alter database rename file ''/u04/oracle/shcl1/onlinelog/redo7_1.log'' to ''/u02/oracle/shcl1rec/redo7_1.log'' ";
sql "alter database rename file ''/u03/oracle/shcl1/onlinelog/redo7_2.log'' to ''/u02/oracle/shcl1rec/redo7_2.log'' ";
sql "alter database rename file ''/u04/oracle/shcl1/onlinelog/redo8_1.log'' to ''/u02/oracle/shcl1rec/redo8_1.log'' ";
sql "alter database rename file ''/u03/oracle/shcl1/onlinelog/redo8_2.log'' to ''/u02/oracle/shcl1rec/redo8_2.log'' ";
sql "alter database rename file ''/u04/oracle/shcl1/dataguard/standby_redo01.log'' to ''/u02/oracle/shcl1rec/standby_redo01.log'' ";
sql "alter database rename file ''/u04/oracle/shcl1/dataguard/standby_redo02.log'' to ''/u02/oracle/shcl1rec/standby_redo02.log'' ";
sql "alter database rename file ''/u04/oracle/shcl1/dataguard/standby_redo03.log'' to ''/u02/oracle/shcl1rec/standby_redo03.log'' ";
sql "alter database rename file ''/u04/oracle/shcl1/dataguard/standby_redo04.log'' to ''/u02/oracle/shcl1rec/standby_redo04.log'' ";
sql "alter database rename file ''/u04/oracle/shcl1/dataguard/standby_redo05.log'' to ''/u02/oracle/shcl1rec/standby_redo05.log'' ";
sql "alter database rename file ''/u04/oracle/shcl1/dataguard/standby_redo06.log'' to ''/u02/oracle/shcl1rec/standby_redo06.log'' ";
sql "alter database rename file ''/u04/oracle/shcl1/dataguard/standby_redo07.log'' to ''/u02/oracle/shcl1rec/standby_redo07.log'' ";
sql "alter database rename file ''/u04/oracle/shcl1/dataguard/standby_redo08.log'' to ''/u02/oracle/shcl1rec/standby_redo08.log'' ";
sql "alter database rename file ''/u04/oracle/shcl1/dataguard/standby_redo09.log'' to ''/u02/oracle/shcl1rec/standby_redo09.log'' ";
sql "alter database rename file ''/u04/oracle/shcl1/dataguard/standby_redo10.log'' to ''/u02/oracle/shcl1rec/standby_redo10.log'' ";
sql "alter database rename file ''/u04/oracle/shcl1/dataguard/SHCL1_2srl0.f'' to ''/u02/oracle/shcl1rec/SHCL1_2srl0.f'' ";
sql "alter database rename file ''/u04/oracle/shcl1/dataguard/SHCL1_2srl1.f'' to ''/u02/oracle/shcl1rec/SHCL1_2srl1.f'' ";
sql "alter database rename file ''/u04/oracle/shcl1/dataguard/SHCL1_2srl2.f'' to ''/u02/oracle/shcl1rec/SHCL1_2srl2.f'' ";
sql "alter database rename file ''/u04/oracle/shcl1/dataguard/SHCL1_2srl3.f'' to ''/u02/oracle/shcl1rec/SHCL1_2srl3.f'' ";
sql "alter database rename file ''/u04/oracle/shcl1/dataguard/SHCL1_2srl4.f'' to ''/u02/oracle/shcl1rec/SHCL1_2srl4.f'' ";
sql "alter database rename file ''/u04/oracle/shcl1/dataguard/SHCL1_1srl0.f'' to ''/u02/oracle/shcl1rec/SHCL1_1srl0.f'' ";
sql "alter database rename file ''/u04/oracle/shcl1/dataguard/SHCL1_1srl1.f'' to ''/u02/oracle/shcl1rec/SHCL1_1srl1.f'' ";
sql "alter database rename file ''/u04/oracle/shcl1/dataguard/SHCL1_1srl2.f'' to ''/u02/oracle/shcl1rec/SHCL1_1srl2.f'' ";
restore database;switch datafile all;
recover database;
exit;
EOF
exit    
T02[SHCL1DR2]$                
T02[SHCL1DR2]$ nohup /export/home/oraprd/restore.sh > /export/home/oraprd/restore.log 2>&1 &

If you are applying the archive logs to recover to a point in time, you might want to disable ‘Block Change Tracking’ before ‘recover’ if it is enabled.

SQL> ALTER DATABASE DISABLE BLOCK CHANGE TRACKING

• 2.j] Open the database using RESETLOGS

 rman target /
 alter database open resetlogs;
 Opening the database would create the redologs, but not the standby logs.

• 2.k] Remove the standby redologs

Since the source db contained standby redo logs, either we should create it or should drop them, else we would
get ORA-00313 errors as the files are not created along with the redologs.

 Errors in file /u01/app/oracle/admin/SHCL1REC/bdump/shcl1rec_arc2_12529.trc:
 ORA-00313:

 *** SERVICE NAME:(SYS$BACKGROUND) 2009-07-24 17:59:53.687
 *** SESSION ID:(631.5) 2009-07-24 17:59:53.687
 *** 2009-07-24 17:59:53.687 2561 kcrf.c
 tkcrf_clear_srl: Started clearing Standby Redo Logs   
 ORA-00312: /u02/oracle/shcl1rec/standby_redo01.log'
 SVR4 Error: 2: No such file or directory
 Additional information: 3                              
 ORA-00312: /u02/oracle/shcl1rec/standby_redo02.log'
 SVR4 Error: 2: No such file or directory

 SQL> select group#,status,type from v$logfile;

 GROUP# STATUS                TYPE
---------- --------------------- ---------------------
 1                       ONLINE
 1                       ONLINE
 2                       ONLINE
 2                       ONLINE
 3                       ONLINE
 3                       ONLINE
 4                       ONLINE
 4                       ONLINE
 5                       ONLINE
 5                       ONLINE
 6                       ONLINE
 6                       ONLINE
 7                       ONLINE
 7                       ONLINE
 8                       ONLINE
 8                       ONLINE
 9                       STANDBY
 10                       STANDBY
 11                       STANDBY
 12                       STANDBY
 13                       STANDBY
 14                       STANDBY
 15                       STANDBY
 16                       STANDBY
 17                       STANDBY
 18                       STANDBY
 19                       STANDBY
 20                       STANDBY
 21                       STANDBY
 22                       STANDBY
 23                       STANDBY
 24                       STANDBY
 25                       STANDBY
 26                       STANDBY

34 rows selected.

 alter database drop standby logfile group 9
 alter database drop standby logfile group 10
 alter database drop standby logfile group 11
 alter database drop standby logfile group 12
 alter database drop standby logfile group 13
 alter database drop standby logfile group 14;
 alter database drop standby logfile group 15;
 alter database drop standby logfile group 16;
 alter database drop standby logfile group 17;
 alter database drop standby logfile group 18;
 alter database drop standby logfile group 19;
 alter database drop standby logfile group 20;
 alter database drop standby logfile group 21;
 alter database drop standby logfile group 22;
 alter database drop standby logfile group 23;
 alter database drop standby logfile group 24;
 alter database drop standby logfile group 25;
 alter database drop standby logfile group 26;

• 2.l] Disable the redo thread(unused)

Typically, an oracle instance would have only 1 redo thread. An n node RAC would have n redo threads.
Since we restored the backup of a 2 node RAC database to a single instance, whilst startup of the single instnace
Oracle would consider one of the thread as Private to the instance and the other as Public.
A peep into the alert.log says, this instance has mounted redo thread 1.

 :
 Mon Jul 27 13:45:56 2009
 Successful mount of redo thread 1, with mount id 1863696975
 Mon Jul 27 13:45:56 2009
 Database mounted in Exclusive Mode
 Completed: ALTER DATABASE   MOUNT
 :

 SQL> select group#,thread#,sequence#,status from v$log;

 GROUP#    THREAD#  SEQUENCE# STATUS
 ---------- ---------- ---------- ------------------------------------------------
 1          1          5 INACTIVE
 2          1          6 ACTIVE
 3          1          7 CURRENT
 4          1          4 INACTIVE
 5          2          1 INACTIVE
 6          2          2 INACTIVE
 7          2          3 INACTIVE
 8          2          4 CURRENT

 8 rows selected.

Groups 1-4 belong to the Thread 1, which is private to this instance and 5-8 belong to Thread 2, which is public.
Means, this instance would write the commit/rollback changes only to thread 1, but in case of recovery it would use
thread 2 also.

A look into the archivelog destination says,

 -rw-r-----   1 oraprd   oinstall    1024 Jul 27 14:03 SHCL1REC_2_2_693078959.arch
 -rw-r-----   1 oraprd   oinstall 4768768 Jul 27 14:03 SHCL1REC_1_5_693078959.arch
 -rw-r-----   1 oraprd   oinstall    1024 Jul 27  2009 SHCL1REC_2_3_693078959.arch
 -rw-r-----   1 oraprd   oinstall 1798144 Jul 27  2009 SHCL1REC_1_6_693078959.arch

The size of the archive logs created by Thread 2 are much small when compared to those created by Thread 1.
I used the logminer utility to confirm that the Thread 2 does not contain any Undo/Rollback data.

–Mining archvielog from Thread 1

 EXECUTE DBMS_LOGMNR.ADD_LOGFILE( LOGFILENAME => '/u02/oracle/shcl1rec/archivelogs/SHCL1REC_1_6_693078959.arch', OPTIONS => DBMS_LOGMNR.NEW);
 EXECUTE DBMS_LOGMNR.START_LOGMNR( OPTIONS => DBMS_LOGMNR.DICT_FROM_ONLINE_CATALOG);
 SQL> select count(8) from V$LOGMNR_CONTENTS;
 COUNT(8)
 ----------
 5449
 EXECUTE DBMS_LOGMNR.END_LOGMNR();

–Mining archvielog from Thread 2

 EXECUTE DBMS_LOGMNR.ADD_LOGFILE( LOGFILENAME => '/u02/oracle/shcl1rec/archivelogs/SHCL1REC_2_3_693078959.arch', OPTIONS => DBMS_LOGMNR.NEW);
 EXECUTE DBMS_LOGMNR.START_LOGMNR( OPTIONS => DBMS_LOGMNR.DICT_FROM_ONLINE_CATALOG);
 SQL>  select count(8) from V$LOGMNR_CONTENTS;
 COUNT(8)
 ----------
 0
 EXECUTE DBMS_LOGMNR.END_LOGMNR();

No data is being written to the thread 2 ( log groups# 5,6,7 and 8 ), so we can disable it using:

 ALTER DATABASE DISABLE THREAD 2;

 SQL>  select group#,thread#,sequence#,status from v$log;

 GROUP#    THREAD#  SEQUENCE# STATUS
 ---------- ---------- ---------- ------------------------------------------------
 1          1          5 INACTIVE
 2          1          6 INACTIVE
 3          1          7 CURRENT
 4          1          4 INACTIVE
 5          2          1 INACTIVE
 6          2          2 INACTIVE
 7          2          3 INACTIVE
 8          2          4 INACTIVE

 8 rows selected.

Since Group# 8 was the CURRENT(of Thread 2) before disabling, archive it.

 SQL> alter system archive log group 8;

 System altered.

Now we can drop the groups that belonged to Thread 2.

 ALTER DATABASE DROP LOGFILE GROUP 5;
 ALTER DATABASE DROP LOGFILE GROUP 6;
 ALTER DATABASE DROP LOGFILE GROUP 7;
 ALTER DATABASE DROP LOGFILE GROUP 8;

SQL>  select group#,thread#,sequence#,status from v$log;
 
 GROUP#    THREAD#  SEQUENCE# STATUS
 ---------- ---------- ---------- ------------------------------------------------
 1          1          5 INACTIVE
 2          1          6 INACTIVE
 3          1          7 CURRENT
 4          1          4 INACTIVE

1)Using exp and imp utilities
.Export the table/schema with rows=n (This will not export the data but only the table structure)
.Import it with indexfile=table_metadata.log (indexfile will make imp to write the contens of .dmp file to the mentioned log file. Nothing will be done inside the schema)

To get the metdata of the table emp in schema john:

 $ exp tables=john.emp file=john_emp.dmp rows=n
 $ imp full=y file=john_emp.dmp indexfile=john_emp_metadata.log
 $ cat john_emp_metadata.log
REM  CREATE TABLE "JOHN"."EMP" ("EMPNO" NUMBER, "ENAME" VARCHAR2(20),
REM  "SAL" NUMBER, "DEPTNO" NUMBER) PCTFREE 10 PCTUSED 40 INITRANS 1
REM  MAXTRANS 255 STORAGE(INITIAL 131072 FREELISTS 1 FREELIST GROUPS 1
REM  BUFFER_POOL DEFAULT) TABLESPACE "USERS" LOGGING NOCOMPRESS ;

2)Using DBMS_METADATA.GET_DDL()

To get the metadata of individual objects:

select dbms_metadata.get_ddl(‘<OBJ_TYPE>’,'<OBJ_NAME>’,'<SCHEMA>’) from dual;

SQL> SET PAGES 1000
SQL> SET LONG 1000
SQL> select dbms_metadata.get_ddl('TABLE','EMP','JOHN') from dual;
DBMS_METADATA.GET_DDL('TABLE',
--------------------------------------------------------------------------------
 CREATE TABLE "JOHN"."EMP"
 ("EMPNO" NUMBER,
 "ENAME" VARCHAR2(20),
 "SAL" NUMBER,
 "DEPTNO" NUMBER )
PCTFREE 10 PCTUSED 40 INITRANS 1 MAXTRANS 255 NOCOMPRESS LOGGING
 STORAGE(INITIAL 65536 NEXT 1048576 MINEXTENTS 1 MAXEXTENTS 2147483645 PCTINCREASE 0 FREELISTS 1 FREELIST GROUPS 1 BUFFER_POOL DEFAULT)
TABLESPACE "USERS"

To get the metadata of the tables or indexes or materialized views with a ” ; ” at the end of each script use

SET LINE 240
SET ECHO OFF
SET PAGES 0
SET LONG 90000
SET HEADING OFF
SET TRIMSPOOL OFF
SET FEEDBACK OFF
SET LONGCHUNKSIZE 600
SPOOL USER_METADATA.LOG
EXEC DBMS_METADATA.SET_TRANSFORM_PARAM(DBMS_METADATA.SESSION_TRANSFORM,'SQLTERMINATOR',TRUE);  
SELECT DBMS_METADATA.GET_DDL('TABLE',UT.TABLE_NAME) FROM USER_TABLES UT;
SELECT DBMS_METADATA.GET_DDL('INDEX',IDX.INDEX_NAME) FROM USER_INDEXES IDX;
SELECT DBMS_METADATA.GET_DDL(REPLACE(OBJECT_TYPE,' ','_'),OBJECT_NAME)
FROM USER_OBJECTS
WHERE OBJECT_TYPE IN ('MATERIALIZED VIEW');

SPOOL OFF