II.1 Primary Database
Primary database is the operasional database accessed by users or aplications.

II.2 Standby Database
Standby database is copy from primary database. Production and standby database connected through oracle net. Primary database send archive log to standby database and then archived log will be apply in the standby database.

III. Installation steps Oracle Data Guard
III.1 Oracle software install
For this post i’m using O/S Red Hat Enterprise Linux AS release 4 (Nahant Update 7) 64 bit and Oracle 10.2.0.1. When we install the Oracle 10.2.0.1 we have to choose Installation type Enterprise Edition, because it have oracle data guard feature.
Do Oracle installation on production database and standby database with installation type Enterprise Edition.
In production database we install Oracle with the database (the database instance named borneodb)
In standby database install just install the oracle instance , because the database for standby database is copy from primary database (cold backup primary and the restore tostandby).
III.2 Preparation in production database (oraDBServer1) for data guard installation
There are many things that we hav e to prepare before configuring database for oracle database and create the standby database using oracle data guard configuration :
1.Enable Force Logging
After oracle installation and database installation had completed we have to alter database to force logging mode with this sql below :
SQL> ALTER DATABASE FORCE LOGGING;
2.Make a Password file
$ orapwd file=orapw password=top_secret
3.Make standby redo log
alter database add standby logfile group 6
(‘/u01/app/oracle/oradata/borneodb/DRCREDO_01.log’) size 5242880;
To see the redo log file in primary we can use this query : select * from v$logfile;
From Oracle documentation ‘A standby redo log is required for the maximum protection and maximum availability modes and the LGWR ASYNC transport mode is recommended for all databases. Data Guard can recover and apply more redo data from a standby redo log than from archived redo log files alone. ‘

4.Init file for data guard configuration
Make init file from spfile with command :
SQL> create pfile=’$ORACLE_HOME/dbs/initborneodb.ora’ from spfile
Then edit pfile for data guard configuration :
borneodb.__db_cache_size=1207959552
borneodb.__java_pool_size=16777216
borneodb.__large_pool_size=16777216
borneodb.__shared_pool_size=352321536
borneodb.__streams_pool_size=0
db_name=’borneodb’
db_unique_name=’borneodb’
service_names=’borneodb’
fal_client=’borneodb’
fal_server=’borneodrc’
log_archive_config=’dg_config=(borneodb,borneodrc)’
log_archive_dest_1=’LOCATION=/data/oracle/borneodb/archived’
log_archive_dest_2=’service=BORNEODBDRC valid_for=(ALL_LOGFILES,PRIMARY_ROLES) db_unique_name=borneodrc lgwr async=20480′
log_archive_dest_state_1=’ENABLE’
log_archive_dest_state_2=’DEFER’
log_archive_format=’%t_%s_%r.dbf’
standby_archive_dest=’/data/oracle/borneodb/archived’
standby_file_management=’auto’
*.audit_file_dest=’/u01/app/oracle/admin/borneodb/adump’
*.background_dump_dest=’/u01/app/oracle/admin/borneodb/bdump’
*.compatible=’10.2.0.1.0′
5.Enable mode archive log
To activate archivelog mode :
SQL > Shutdown immediate;
SQL > Startup mount pfile=initborneodb.ora;
SQL > alter database archivelog;
6.Start data guard broker and check the status (untuk switchover)
SQL> ALTER SYSTEM SET DG_BROKER_START=TRUE;
SQL> SHOW PARAMETER DG_BROKER_START
7.Listener configuration for file LISTENER.ORA
Example :
LISTENER = (DESCRIPTION =
(ADDRESS_LIST=(ADDRESS=(PROTOCOL=tcp)(HOST=host_name)
(PORT=port_num))))
SID_LIST_LISTENER=(SID_LIST=(SID_DESC=(SID_NAME=sid_name)
(GLOBAL_DBNAME=db_unique_name_DGMGRL.db_domain)
(ORACLE_HOME=oracle_home)))
LISTENER.ORA pada BORNEODDB :
SID_LIST_LISTENER =
(SID_LIST =
(SID_DESC =
(GLOBAL_DBNAME = borneodb_DGMGRL)
(ORACLE_HOME = /u01/app/oracle/product/10.2.0/db_1)
(SID_NAME = borneodb)
)
)

LISTENER =
(DESCRIPTION_LIST =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = TCP)(HOST = oraDBServer1)(PORT = 1521))
(ADDRESS = (PROTOCOL = IPC)(KEY = EXTPROC0))
)
)
8.Make tnsnames.ora for connection to drc (borneodrc)
BORNEODB =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = TCP)(HOST = oraDBServer1)(PORT = 1521))
(CONNECT_DATA =
(SERVER = SHARED)
(SERVICE_NAME = borneodb)
)
)

#ListenerForDRC
BORNEODBDRC =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = TCP)(HOST = 10.128.16.126)(PORT = 1521))
(CONNECT_DATA =
(SERVER = DEDICATED)
(SERVICE_NAME = borneodrc)
)
)

9.Create standby control file from primary to be copied to standby database
SQL> SHUTDOWN IMMEDIATE;
SQL> STARTUP MOUNT;
SQL> ALTER DATABASE CREATE STANDBY CONTROLFILE AS ‘/data/standby_control01.ctl’;
SQL> SHUTDOWN IMMEDIATE;
10.Copy all database file from primary database to standby database
Copy datafile, standby control file, initialization file to standby database.

III.3 Preparation in standby database (oraDBServer2)
After Copy all database file from primary database to standby database have been completed :
1.Edit initialization parameter file from primary database for standby database (file $ORACLE_HOME/dbs/initborneodb.ora)
borneodb.__db_cache_size=1207959552
borneodb.__java_pool_size=16777216
borneodb.__large_pool_size=16777216
borneodb.__shared_pool_size=352321536
borneodb.__streams_pool_size=0
db_name=’borneodb’
db_unique_name=’borneodrc’
service_names=’borneodb’
fal_client=’borneodrc’
fal_server=’borneodb’
log_archive_config=’dg_config=(borneodb,borneodrc)’
log_archive_dest_1=’LOCATION=/data/oracle/borneodb/archived’
log_archive_dest_2=’service=borneodb valid_for=(ALL_LOGFILES,PRIMARY_ROLES) db_unique_name=borneodrc lgwr async=20480′
log_archive_dest_state_1=’ENABLE’
log_archive_dest_state_2=’DEFER’
log_archive_format=’%t_%s_%r.dbf’
standby_archive_dest=’/data/oracle/borneodb/archived’
standby_file_management=’auto’
*.audit_file_dest=’/u01/app/oracle/admin/borneodb/adump’
*.background_dump_dest=’/u01/app/oracle/admin/borneodb/bdump’
*.compatible=’10.2.0.1.0′
@
2.Edit file listener.ora dan start listener
SID_LIST_LISTENER =
(SID_LIST =
(SID_DESC =
(SID_NAME = PLSExtProc)
(ORACLE_HOME = /u01/app/oracle/product/10.2.0/db_1)
(PROGRAM = extproc)
)
(SID_DESC =
(GLOBAL_DNAME = borneodrc_DGMGRL)
(ORACLE_HOME = /u01/app/oracle/product/10.2.0/db_1)
(SID_NAME = borneodb)
)

)

LISTENER =
(DESCRIPTION_LIST =
(DESCRIPTION =
(ADDRESS = (PROTOCOL = IPC)(KEY = EXTPROC1))
(ADDRESS = (PROTOCOL = TCP)(HOST = oraDBServer2)(PORT = 1521))
)
)
Start listener : #lsnrctl start
3.Startup mount and change database mode to manage standby
#sqlplus / as sysdba
SQL> startup mount pfile=’initborneodb.ora’;
SQL> ALTER DATABASE RECOVER MANAGED STANDBY DATABASE DISCONNECT FROM SESSION;
IV. START THE ORACLE DATA GUARD
1.in primary startup oracle using pfile that have been modified
#sqlplus / as sysdba
SQL> startup open pfile=’initborneodb.ora’
2.Di primary ubah log_archive_dest_state_2 menjadi ENABLE
SQL> Alter system set log_archive_dest_state_2=’ENABLE’;
3.Di primary lakukan log switch untuk pembuatan archive log
SQL> alter system switch logfile
4.Check status of redo log send from primary database to standby database :
SQL>select dest_id, destination, status, error from v$archive_dest;
See destination 2, destination to standby database. If archive from primary not standby has not sent, status column from above query will give error message. If archive log send from primary to standby database succes then the value of status column is VALID.

5.Monitor archived in standby database sent from primary database
SQL>select name,to_char(FIRST_TIME,’dd-mon-yy hh24:mi:ss’) TIME ,sequence#,applied from v$archived_log order by sequence#
6.Cek archive gap in standby database
SQL> select * from v$archive_gap;
7.Cek sequence in standby database is it already same with primary database
SQL>select max(sequence#) from v$log_history;

V. Cek standby database
To make sure the data guard configuration have been success and standby database identic with primary database we can alter standby database to open read only with these steps :
1.Login to sqlplus
#sqlplus / as sysdba
2.Do these sql below
SQL>alter database recover managed standby database cancel;
SQL>alter database open read only;
3.Query to check table :
Contoh : SQL>select * from tab; →check this on primary too.
4.to alter back to standby mode
alter database recover managed standby database disconnect from session;

3. Create shell script to automate execute sql script using cronjob (for unix user)
sqlplus -s $schema/$pass@$dbLink as sysdba<<!
set timing on
@ORA_KILL

1. Query for Status Monitoring
Examples of Status Monitoring are, check tablespace where it’s about to fills up, check segment near their maximum extents.
-Tablespace related
This query will show tablespace with percentage free less than 20 percent or equal 20 percent
select ddf.TABLESPACE_NAME,
ddf.BYTES,
ddf.BYTES-DFS.BYTES “BYTES_USED”,
round(((ddf.BYTES-dfs.BYTES)/ddf.BYTES)*100,2) “PERCENT_USED”,
dfs.BYTES “BYTES_FREE”,
round((1-((ddf.BYTES-dfs.BYTES)/ddf.BYTES))*100,2) “PERCENT_FREE”
from (select TABLESPACE_NAME,
sum(BYTES) bytes
from   dba_data_files
where autoextensible=’NO’
group  by TABLESPACE_NAME) ddf,
(select TABLESPACE_NAME,
sum(BYTES) bytes
from dba_free_space
group by TABLESPACE_NAME) dfs
where ddf.TABLESPACE_NAME=dfs.TABLESPACE_NAME and round((1-((ddf.BYTES-dfs.BYTES)/ddf.BYTES))*100,2) < 20 or              round((1-((ddf.BYTES-dfs.BYTES)/ddf.BYTES))*100,2) = 20
order by ((ddf.BYTES-dfs.BYTES)/ddf.BYTES) desc

Query to show tablespace and the datafile
SELECT dfs.TABLESPACE_NAME, SUBSTR (df.NAME, 1, 70) file_name, df.bytes / 1024 / 1024 allocated_mb,
((df.bytes / 1024 / 1024) – NVL (SUM (dfs.bytes) / 1024 / 1024, 0))
used_mb,
NVL (SUM (dfs.bytes) / 1024 / 1024, 0) free_space_mb
FROM v$datafile df, dba_free_space dfs
WHERE df.file# = dfs.file_id(+)
GROUP BY dfs.file_id, df.NAME, df.file#, df.bytes, dfs.TABLESPACE_NAME
ORDER BY file_name

-Oracle Objects Extents related
Show objects more than 50% of Max Extents
select   OWNER,
TABLESPACE_NAME,
SEGMENT_NAME,
SEGMENT_TYPE,
BYTES,
EXTENTS,
MAX_EXTENTS,
(EXTENTS/MAX_EXTENTS)*100 \”PERCENTAGE\”
from     dba_segments
where    SEGMENT_TYPE in (‘TABLE’,'INDEX’)
and      EXTENTS > MAX_EXTENTS/2
order by (EXTENTS/MAX_EXTENTS) desc

Show Segment near their maximum extents
SELECT e.owner,
e.segment_type,
Substr(e.segment_name, 1, 30) “SEGMENT_NAME”,
Trunc(s.initial_extent/1024) “INITIAL K”,
Trunc(s.next_extent/1024) “NEXT K”,
s.max_extents,
Count(*) as extents
FROM   dba_extents e,
dba_segments s
WHERE  e.owner        = s.owner
AND    e.segment_name = s.segment_name
AND    e.owner        NOT IN (‘SYS’, ‘SYSTEM’)
GROUP BY e.owner, e.segment_type, e.segment_name, s.initial_extent, s.next_extent, s.max_extents
HAVING Count(*) > s.max_extents – 10
ORDER BY e.segment_type, e.owner, e.segment_type, Count(*) DESC

-Show Oracle session related and memory used
SELECT NVL(a.username,’(oracle)’) AS username,
a.sid,
a.serial#,
a.machine,
a.module,
a.program,
a.LOGON_TIME,
a.status,
a.process,
Trunc(b.value/1024) AS memory_kb
FROM   v$session a,
v$sesstat b,
v$statname c
WHERE  a.sid = b.sid
AND    b.statistic# = c.statistic#
AND    c.name = ‘session pga memory’
AND    a.program IS NOT NULL
ORDER BY  MEMORY_KB desc, a.MODULE asc,username desc

2. Query for Performance Monitoring
-Show Oracle Hit Ratio
Dictionary Hit Ratio
The ratio of cache hits to cache lookup attempts in the data dictionary cache. In other words, the rate the database goes to the dictionary   instead of the hard disk to retrieve data. A low ratio suggests more RAM should be added
The Dictionary Hit Ratio value have to be more than 90 percent
SELECT sum(gets) “GETS”,sum(getmisses) “CACHE MISSES”,round((1 – (sum(getmisses) / sum(gets))) * 100,3) “DICTIONARY”
FROM v$rowcache

Buffer Hit Ratio
The rate the database goes to the buffer instead of the hard disk to retrieve data. A low ratio suggests more RAM should be added to the system.
Buffer hit ratio have to be more than 85%
SELECT cur.value “block”,con.value “con”,phy.value “phy”, ROUND((1-(phy.value / (cur.value + con.value)))*100,2) “BUFFER”
FROM v$sysstat@”.$dbName.” cur, v$sysstat@”.$dbName.” con, v$sysstat@”.$dbName.” phy
WHERE cur.name = ‘db block gets’
AND con.name = ‘consistent gets’
AND phy.name = ‘physical reads’

Library Hit Ratio
Monitors the percentage of entries in the library cache that were parsed more than once (reloads) over the lifetime of the instance.
Library Hit Ratio have to be more than 95%
SELECT sum(pins) “EXEC”, sum(pinhits) “EXEC HITS”, round((sum(pinhits) / sum(pins)) * 100,3) “HIT RATIO”, sum(reloads) “MISSES”,round((sum(pins) / (sum(pins) + sum(reloads))) * 100,3) “HITRATIO”
FROM v$librarycache

-Show memory usage by user connected
SELECT NVL(a.username,’(oracle)’) AS username,
a.machine,
a.module,
a.program,
Trunc(b.value/1024) AS memory_kb
FROM   v$session a,
v$sesstat b,
v$statname c
WHERE  a.sid = b.sid
AND    b.statistic# = c.statistic#
AND    c.name = ‘session pga memory’
AND    a.program IS NOT NULL
ORDER BY  MEMORY_KB desc, a.MODULE asc,username desc

-Show Oracle Sort Information
select name,to_char(value,’999,999,999,999,999′) “VALUE”
from sys.v_$sysstat
where name like ‘sort%’
Show hit ratio for sort
Sorting percentage on disk have less than 5 %. If the value more than 5 % then change parameter SORT_AREA_SIZE.
select to_char(100*a.value/decode((a.value+b.value),0,1,(A.value+b.value)),’999.90′) “VALUE”
from sys.v_$sysstat a, sys.v_$sysstat b
where a.name = ‘sorts (disk)’
and b.name = ‘sorts (memory)’

-Show Oracle SGA Info
select name, to_char(value,’999,999,999,999,999′) “VALUE” from sys.v_$sga

-Show Oracle Invalid Objects
This query is to show invalid objects in Oracle
select OWNER,
OBJECT_TYPE,
OBJECT_NAME,
STATUS
from dba_objects
where STATUS = ‘INVALID’
order by OWNER, OBJECT_TYPE, OBJECT_NAME

Solution :
if the object invalid you can recompile it or you can delete it if it’s not used.
sql : SQL>ALTER PROCEDURE PROC_NAME COMPILE ;
SQL>ALTER PACKAGE PACKAGE_NAME COMPILE ;
SQL>ALTER VIEW VIEW_NAME COMPILE ;
Or if you have many invalid objects run this script reside in :  $ORACLE_HOME/rdbms/admin/utlrp.sql
Login to sqlplus
sqlplus > @utlrp

-Show Query running in Oracle
This query will show currently queries running in Oracle.
SELECT *
FROM (SELECT Substr(a.sql_text,1,50) sql_text,Trunc(a.disk_reads/Decode(a.executions,0,1,a.executions))     reads_per_execution,a.buffer_gets,a.disk_reads,a.executions,a.sorts,a.address FROM v$sqlarea a) WHERE  rownum <= 10

To be continued ……………………………

Unfortunately i’m accidentally delete one datafile (not the system datafile) and when i want to shutdown the database i get this error code 

ORA-01157: cannot identify/lock data file 121 – see DBWR trace file

ORA-01110: data file 121: ‘/ora_data/ORADATA/USAGE_INDEX23_01.dbf’

to solve just do this simple step:

1.Shutdown abort your database to force shutdown

    sqlplus > shutdown abort

2.Startup mount database

 sqlplus > startup mount

3.After that issue this following command

sqlplus > ALTER DATABASE DATAFILE ‘<datafile name with complete path>’ OFFLINE DROP;

4.Then open the database

sqlplus > alter database open

5.Drop the tablespace by issue this following command

DROP TABLESPACE <tablespace name> INCLUDING CONTENTS;

6.After that if you have backup your database you can recover your tablespace.

WE WILL NOT GO DOWN (Song for Gaza)

(Composed by Michael Heart)

Copyright 2009

A blinding flash of white light

Lit up the sky over Gaza tonight

People running for cover

Not knowing whether they’re dead or alive

They came with their tanks and their planes

With ravaging fiery flames

And nothing remains

Just a voice rising up in the smoky haze

We will not go down

In the night, without a fight

You can burn up our mosques and our homes and our schools

But our spirit will never die

We will not go down

In Gaza tonight Women and children alike

Murdered and massacred night after night

While the so-called leaders of countries afar

Debated on who’s wrong or right

But their powerless words were in vain

And the bombs fell down like acid rain

But through the tears and the blood and the pain

You can still hear that voice through the smoky haze

We will not go down

In the night, without a fight

You can burn up our mosques and our homes and our schools

But our spirit will never die

We will not go down

In Gaza tonight

Hope we all will be better people in this very new year….. Amin….

Alhamdulillah for the things we got in 1429 Hijriah and 2008, bad things and good things is not coming for no reasons, but it gives us experiences, to strengthen us, to taught us to be the best person,  give our best for everything goods  and do not ever we lost tracks ( Follow Alquran and the prophet Muhammad ways)

And to my fellow Muslim Palestinian ..I am praying for you… Your victory will come…

Database need to be monitored to check the performance of the monitored database, so we can change the database parameter based on result from monitoring and keep the database in it’s best performance.

Database monitoring also needed so we always know the condition of database, and know the symptom of our database and we can take action right before more problem happens to database.

Database Monitoring can be grouped into three types :

This monitoring type monitors current status of an events and give reports or alert when it exceeds a defined threshold

Examples of Status Monitoring are, check tablespace where it’s about to fills up, check segment near their maximum extents.

Monitors oracle instance memory (sga, buffer, etc), sql query running in database, session which consumes a lot of memory, etc.

Examples of Performance monitoring are, query from v_$sqlarea to check what SQL has been running in our Oracle Machine, query from v$sgastat and query to get hit ratio for cache , library and buffer cache.

Collects historical data on specific event that occur in database like SGA status, segment that cannot extends, filled up tablespace and make analysis from that historical data so that we can predict the status of SGA in any given time or we can now when the tablespace will filled up.

Examples of Trend Analysis are, make a histroy table for events we monitored. We can make history table for tablespace so that from that history table we could gain information and make prediction when will the time the tablespace will be filled up. Or we can make history table from v_$session, so from that table we can know what time our database accessed by so many user.

Next is the query to monitored database … soon…

There are two ways to write error-free programs. Only the third one works.

I love that you get cold when it’s 71 degrees out. I love that it takes you an hour and a half to order a sandwich. I love that you get a little crinkle above your nose when you’re looking at me like I’m nuts. I love that after I spend the day with you, I can still smell your perfume on my clothes. And I love that you are the last person I want to talk to before I go to sleep at night. And it’s not because I’m lonely, and it’s not because it’s New Year’s Eve. I came here tonight because when you realize you want to spend the rest of your life with somebody, you want the rest of your life to start as soon as possible.

Harry Burns: There are two kinds of women: high maintenance and low maintenance.
Sally Albright: Which one am I?
Harry Burns: You’re the worst kind. You’re high maintenance but you think you’re low maintenance.

Just love the quote……

To remote your Linux into your workstation you need software to remote your Linux, for now i choose Xmanager (Quite simple but “LICENSED” huh). There is other tool like VNC or XMing(I think i choose XManager because i’ve use it in my Solaris Box long time ago and always got good connection using it)

So to enable your Linux for remote desktop just follow this simple  instruction :

1. Open gdm.conf configuration file, usually located in /etc/X11/gdm/gdm.conf or /etc/gdm/custom.conf

2. Edit gdm.conf file, in section [gdcmp] change enable =  false  to  enable  =  true

3. Change run level to 5 in file /etc/inittab (example : x:5:once:/etc/X11/prefdm -nodaemon

4. Restart gdm service by simply run command : gdm-restart

5. Find your Xmanager source, installed it but not crack it.

PS : To enable X graphic open your terminal using root and type xhost +

Now let see my Linux Box in my workstation