http://www.idbas.com/blog/index.php http://www.idbas.com/blog/index.php?entry=entry100801-222813 If possible, we can even use this step for ASM database too. but make sure we create the ASM Disk groups. We still need to explore this method in ASM.

select 'dd if='||FILE_NAME||' bs=256k|gzip --fast |ssh urban "gunzip|dd '||' of='||replace(FILE_NAME,('SOURCE_SID'),('TARGET_SID'))||' bs=256k' ||' " ' from dba_data_files order by file_name;



@idbas.com we take every optimize approach to accomplish.We always give 100% database availability using Framework Approach.
Disclaimer : We are not responsible for commands.Information for reference only,not for execution. Before executing strongly recomend to test it]]> http://www.idbas.com/blog/index.php?entry=entry100801-174030
Gather stats information ready sql scripts for DBA : @ iDBA's we thought these are the useful quick statements to find out the status of gather status :


Script to find-out last gather apps stats ran in Oracle applications : It should work fine between R11 -R12 . Assuming that gather schema stats job is scheduled every wk.

For Apps schema :

select count(1)
from apps.fnd_concurrent_programs_vl p , apps.fnd_concurrent_requests r
where r.concurrent_program_id = p.concurrent_program_id
and r.program_application_id = p.application_id
and p.user_concurrent_program_name in (
'OnDemand Gather Schema Statistics',
'Gather Schema Statistics',
'Gather Schema Statistics (IT_ANALYZE)'
)
and (r.phase_code = 'C' and r.status_code= 'C' and r.actual_start_date >= sysdate-7)
/

For Non-Apps schemas :

set pages 0 echo off head off termout off verify off feed off lines 1000
select count(1)
from apps.fnd_concurrent_programs_vl p , apps.fnd_concurrent_requests r
where r.concurrent_program_id = p.concurrent_program_id
and r.program_application_id = p.application_id
and p.user_concurrent_program_name in (
'Gather Statistics for Non-Apps Schema'
)
and (r.phase_code = 'C' and r.status_code= 'C' and r.actual_start_date >= sysdate-7)



Tofind-out list of tables analyzed since last wk.

select count(1) from dba_tables where last_analyzed < sysdate -7

On the same note : Step to find-out for owner sys. Fur standad or small application based databases, some times it is recomend to gather on sys too.

select count(1) from dba_Tables where owner='SYS' and last_analyzed is not null
/

]]> http://www.idbas.com/blog/index.php?entry=entry100722-125756 ./netca /silent \
/responseFile $ORACLE_HOME/network/install/netca_typ.rsp \
/inscomp server \
/nodeinfo <DB Server Name>


RAC :
./netca /silent \
/responseFile $ORACLE_HOME/network/install/netca_typ.rsp \
/inscomp server \
/nodeinfo node1,node2]]> http://www.idbas.com/blog/index.php?entry=entry100722-123616 Sql Used for run away.

#######################################################
1 select a.spid,b.inst_id,b.sid,b.process,b.module,b.status,machine,round(b.last_call_et/60,0) "Minutes Running" from gv$process a, gv$session b where
2 b.last_call_et >= 3600 *10 and a.addr=b.paddr and b.process='1234'
3 and type !='BACKGROUND'
4* order by 8 desc

Top 20 Active processes running more than 10 hours :
#######################################################

1 select a.spid,b.inst_id,b.sid,b.process,b.module,b.status,machine,round(b.last_call_et/60,0) "Minutes Running" from gv$process a, gv$session b where
2 b.last_call_et >= 3600 *10 and a.addr=b.paddr
3 and type !='BACKGROUND' and status='ACTIVE'
4* order by 8 desc
]]> http://www.idbas.com/blog/index.php?entry=entry100722-123414
RMAN> show RETENTION POLICY;
RMAN configuration parameters are:
CONFIGURE RETENTION POLICY TO REDUNDANCY 1; # default
RMAN> CONFIGURE RETENTION POLICY TO RECOVERY WINDOW OF 8 DAYS;
new RMAN configuration parameters:
CONFIGURE RETENTION POLICY TO RECOVERY WINDOW OF 8 DAYS; new RMAN configuration parameters are successfully stored
RMAN> show RETENTION POLICY;
RMAN configuration parameters are:
CONFIGURE RETENTION POLICY TO RECOVERY WINDOW OF 8 DAYS;]]> http://www.idbas.com/blog/index.php?entry=entry100702-164939 srvctl stop asm -n ora-node-rac-2

cd $ORACLE_HOME/rdbms/lib
make -f ins_rdbms.mk ikfed
As ROOT on BOTH NODES
setasm
vi /etc/sysconfig/oracleasm file, make sure to have these values set:

ORACLEASM_ENABLED=false
ORACLEASM_SCANBOOT=false

/etc/init.d/oracleasm stop
/etc/init.d/oracleasm disable

As ROOT ON FIRST NODE ONLY
kfed read /dev/sdg1 > read.out
vi read.out -- find value kfdhdb.grpname change tag from DATA to CLNE and save file
kfed merge /dev/sdh1 text=merge.out

As ROOT on BOTH NODES
vi /etc/sysconfig/oracleasm update values of two tags
ORACLEASM_ENABLED=true
ORACLEASM_SCANBOOT=false
/etc/init.d/oracleasm enable

As ROOT ON FIRST NODE ONLY
/etc/init.d/oracleasm force-renamedisk /dev/sdh1 CLNE

As ROOT on BOTH NODES
/etc/init.d/oracleasm listdisks
chown oracle:oinstall /dev/sdg1 /dev/sdh1
As USER ORACLE ON BOTH NODES
sqlplus "/ as sysdba"
alter system set asm_diskgroups ='DATA', 'CLNE';
alter system set asm_diskstring = '/dev/sdg1','/dev/sdh1'
alter diskgroup data mount;
alter diskgroup clne mount;
exit


asmcmd
ls -l
ASMCMD> ls -l
State Type Rebal Unbal Name
MOUNTED EXTERN N N CLNE/
MOUNTED EXTERN N N DATA/
]]> http://www.idbas.com/blog/index.php?entry=entry100702-105815 -----------------------------------------------------------------
An application running on OC4J appears unresponsive, with simple requests experiencing noticeable delays. The cause is that the JVM has crossed the low memory threshold and is running a full garbage collection to free up memory.
Solution can be varry from setup to setup:
-----------------------------------------------------------------
Consider using the incremental low pause collector, which avoids long major garbage collection pauses by doing portions of the major collection work at each minor collection. This collector (also known as the train collector) collects portions of the tenured generation - a memory pool holding objects that are typically collected in a major collection - at each minor collection. The result is shorter pauses spread over many minor collections.
The incremental collector is even slower than the default tenured generation collector when considering overall throughput.
To use the incremental collector, the -Xincgc option must be passed in on the Java command line at application startup. Set the initial and maximum size of the young generation (object pool) to the same value using the XX:NewSize and -XX:MaxNewSize options. Set the initial and the maximum Java heap sizes to the same value using the -Xms and-Xmx options.
For example, to use this collector with a server with 1 GB of physical memory:
java -server -Xincgc -XX:NewSize=64m -XX:MaxNewSize=64m -Xms512m -Xmx512m
]]> http://www.idbas.com/blog/index.php?entry=entry100702-105347
Download the sun-jdk-6u20-linux-x64 media from Sun Website.

$HOME/sun-jdk-6u20-linux-x64/jdk-6u20-linux-x64.bin

2. Copy the media into the following temporary location

Download required jdk from site : sun-jdk-6u20-linux-x64
cp -p jdk-6u20-linux-x64.bin $HOME

cd $HOME
./jdk-6u20-linux-x64.bin
Follow thru and accept the agreement, and the jdk install will complete and create the following dir under $HOME/jdk1.6.0_20


3. Perform the following steps for any Oracle Homes upgrading from 32 bit sun jdk to 64 bit sun jdk ( in our case it is SOA and WSM Oracle homes)


Steps shown for SOA oracle home:
----------------------------------

.setenv
opmnctl stopall

cd $ORACLE_HOME
mv jdk jdk.32bit.orig
cp -rp $HOME/jdk1.6.0_20 .
mv jdk1.6.0_20 jdk

cd $ORACLE_HOME/opmn/conf
Edit opmn.xml file and remove any OC4J instance java options referencing jdk1.5.0_17.
Remove the following line in opmn.xml file for all OC4J_SOA instances.
<data id="java-bin" value="/oracle/products/jdk1.5.0_17/bin/java"/>

Start the servers
opmnctl startall
---------------------------------
4. Similarly follow the above steps for WSM oracle homes.
5. Valide that the 64 bit Sun JDK is being used
java -version (to check the version is JDK 6 u 20 x_64)

ps -ef | grep java ( to ensure <oracle_home>/jdk/bin/java is being used)


]]> http://www.idbas.com/blog/index.php?entry=entry100702-103931


<EmailAccounts xmlns="http://xmlns.oracle.com/ias/pcbpel/NotificationService"
EmailMimeCharset=""
NotificationMode="EMAIL">
<EmailAccount>
<Name>Default</Name>
<GeneralSettings>
<FromName>Customer Service</FromName>
<FromAddress>email@xxx.com</FromAddress>
</GeneralSettings>
<OutgoingServerSettings>
<SMTPHost>smtp.company.com</SMTPHost>
<SMTPPort>25</SMTPPort>
</OutgoingServerSettings>
<IncomingServerSettings>
<Server>smtp.company.com</Server>
<Port>110</Port>
<Protocol>pop3</Protocol>
<UserName>accountId</UserName>
<Password ns0:encrypted="false" xmlns:ns0="http://xmlns.oracle.com/ias/pcbpel/NotificationService">password</Password>
<UseSSL>false</UseSSL>
<Folder>Inbox</Folder>
<PollingFrequency>1</PollingFrequency>
<PostReadOperation>
<MarkAsRead/>
</PostReadOperation>
</IncomingServerSettings>
</EmailAccount>

</EmailAccounts>]]> http://www.idbas.com/blog/index.php?entry=entry100630-133716 Use tar -cvzf to create tar file
use tar -xvzf to untar the zip file.
Or gunzip the tar file to rquired location :

cd <Agent Home >
gunzip < agent10g.tar.gz | tar xvf -

cd /oracle/products
rm -rf agent10g.tar.gz

cd <Agent Home >/oui/bin

./runInstaller -clone -forceClone ORACLE_HOME=<Complete Agent Path > ORACLE_HOME_NAME=agent10g -noconfig -silent

exit

Login in another session and run.. root.sh

cd <Agent Home >/bin

./agentca -f

./emctl secure agent

Pass the required autheticated passwd.

./emctl start agent
]]> http://www.idbas.com/blog/index.php?entry=entry100615-170605 We have learned that , set of database parameters which need to addressed to get optimum performance.
Besides adding more number of cpu and more memory these are the additional parameter which need to considered.

Since Virtual and Database (shared memory) fights for memory management and database settings get big stake of it. so back to old school, consider
setting ".db_cache_advice='OFF' and set pga ,sga and shared memory parameters manually.
As a regular practice , also plan to setup huge page to suggested values. Jumbo Frame works just fine in Virtual RAC.
CPU Best practice and Number of Virtual CPUS :
Configure as many cpu as your development strategy is required.. Like UAT (needed less cpus ) and QA needed more CPU and Pre-production requires.. optimum production equal or less cpu.
Unused virtual cpu imposes unwanted extra overhead... which lead to consume more memory, process timer, process migrations in guests leads to extra overhead..
SMP VM's : Bring up all vcpus at once when scheduling smp vm.


Memory Virtualization :
Recent Operating Systems uses Paging. applications use virtual address to access memory. O/S Maintains virtual address to physical address mapping of every process which is called guest page table..
Since Virtual memory maintains Hypervisor guest table to physical address. Each matching address mapping to shadow page tables..



Process (A) ---> Process (A) (Location1) ---> Process (A) (Location10)
Ex : xx90x00 Ex : xx90x00 Ex : xx90x00
Process (B) ---> Process (B) (Location4) ---> Process (B) (Location13)
Ex : xx90x00 Ex : xx90x00 Ex : xx90x00
Process (C) ---> Process (C) (Location6) ---> Process (C) (Location17)
Ex : xx90x00 Ex : xx90x00 Ex : xx90x00
Process (D) ---> Process (D) (Location8) ---> Process (D) (Location20)
Ex : xx90x00 Ex : xx90x00 Ex : xx90x00

Virtual address for Process Guest physical Address (Operating system ) Shadow Page Table..

In case of VM .. Gust page table directly uses shadow page tables.. this is where the performance conflicts...

Do a memory reservations equal to the memory required by the database and Operating system...Please check with syadmin to fix this.
Being virtualized not only saves money.. it also saves a lot time..

I will be covering how to do RAC- to - RAC clone. Including EBS,SOA,Fusin Middleware database and application servers are functioning for optimum performance.
In today’s' world customers who cannot offered to have to physical RAC's or multiple RAC environments for Development environment, we will recommend going for virtual will give 200% productivity and growth.
Here in idbas.com we will help your onsite dba, how to Build and clone a RAC environment in 6 hrs. In general , it takes 30-60 hrs to build another clone environment.
]]> http://www.idbas.com/blog/index.php?entry=entry100317-163709
What is TOGAF : An Architecuture framework that enables you to design, evaluate and build the right architecture for your oraganization.
SOA styles are developed using TOGAF.

TOGAF consists of 3 main parts :
1.The TOGAF Architecture Development Method.
2.The Enterprise continuum.
3.The TOGAF Resourece Base.


ABCDEFGH.. of TOGAF :
Preliminary Phase :Define Priniciples, Adapt Framework.
A.Architecure Vision. : Define Scope,create vision, obtain apporovals.
B.Business Architecture : Develop Business Architecture.
C.Information systems Architecture : Develop Data and Application Architecture.
D.Technology Architecture : Develop a technology architecture.
E.Opportunities and solutions : Check point suitability for implementation.
F.Migration Planning. : Prioritize, select major work packages,develop migration plan.
G.Implementation Governence. : Provide architectural Oversight of the implementation.
H.Architecture change Management :Establish for procedures for managing change to new architecture.


To start with , have your objectives ready prior to building a Framework.

Objectives
 To ensure everyone is committed to architecture process.
 To define the architecture principles.
 To define scope and assumptions.
 To assess SOA readiness, maturity and define SOA adoption plan.
 To define framework and detail methodologies (i.e. ADM)
 To setup and monitor a process.
 To define criteria for architecture tools.


@idbas.com We will deliver end-2-end SOA ,IDM,B2B,BI Design,Install and setup/configure. for more information, please reach us @ info@idbas.com.
]]> http://www.idbas.com/blog/index.php?entry=entry100315-140402 Detach from oracle inventory

Run the following steps to remove the source nodes reference in inventory. This will detach and attach your new oracle home to the Oracle Inventory.

o Execute this as user "oracle".
o $ cd $ORACLE_HOME/oui/bin

Run this and it will detach oracle home of source server from inventory

o ./runInstaller -silent -detachHome -local -noClusterEnabled -invPtrLoc $ORACLE_HOME/oraInst.loc -ignorePreReq ORACLE_HOME=$ORACLE_HOME ORACLE_HOME_NAME=Ora10gR2_DBNAME << change DBNAME

Attach to oracle inventory
Run this and it will attach oracle home of target server from inventory

o ./runInstaller -silent -attachHome -local -noClusterEnabled -invPtrLoc $ORACLE_HOME/oraInst.loc -ignorePreReq ORACLE_HOME=$ORACLE_HOME ORACLE_HOME_NAME=Ora10gR2_DBNAME CLUSTER_NODES=node1 << change this to actual hostname ( target server ) & DBNAME
]]> http://www.idbas.com/blog/index.php?entry=entry100315-135258 info@idbas.com for Database/application Hardning report.


Sample Hardning report includes :

Account Hardening :
1. OS users
2. Database Users
3. Application Users

Check for Installed/Non-Installed modules.

Process Hardening.
File system Hardening.
Application servers Hardening.
Networking Hardening.

]]> http://www.idbas.com/blog/index.php?entry=entry100315-133222 In order to clone ASM Database using rman from OMF. Rman will not convert OMF (Oracle manged files ) to ASM. Apart from using db_file_name_convert and log_file_name_convert, also need to use db_create_file_dest with OMF and ASM RAC.


Set ths parameter in init.ora file :
db_create_file_dest='+DATA' or What ever the diskstring for ASM to be redirected.


]]> http://www.idbas.com/blog/index.php?entry=entry100308-174235
One of the most important components of an Application Server install is the OPMN. OPMN is essential to the running the Application Server.
It is the one component that allows managing various components within the Application Server.
The opmn consists of three components.
a) The Oracle Notification Server
b) The Process Manager
c) Process Manager Modules

ONS is a pub/sub based transport mechanism that allows for notifications to be sent among components within the Application Server. The notifications are related to the startup/failure/recvery of the components on the server.

The Process Manager is the centralized process management mechanism in Oracle Application Server, and is used to manage App Server processes. It is responsible for starting/stopping/restarting/monitoring every process it manages.Any requests sent to OPMN about controlling a process/obtaining status about a process are redirected to the Process Manager. Even the Enterprise manager uses the PM to manage the processes on the Application Server. The Process manager uses the ONS to detect any process initialization, port usages, and obtain runtime information about any component on the application server.
In case there are multiple application server nodes , the PM running on each node uses ONS to communicate with the PM on the other node and collects information about components in either containers. Note that ONS uses dynamic discovery to announce new servers and join them into the ONS topology dynamically. There are three ways to configure dynamic discovery.
a) Multi-cast
b) Discovery Server
c) Gateway
]]> http://www.idbas.com/blog/index.php?entry=entry100308-171156 alter system set "_hash_join_enabled"=false scope=spfile;

And turn off for your bulk insert/generate.
Once Bulkdelete has finished please run the following
alter system set "_hash_join_enabled"=true scope=spfile;
Bounce the database.





@idbas.com we take every optimize approach to accomplish.We always give 100% database availability using Framework Approach.
Disclaimer : We are not responsible for commands.Information for reference only,not for execution. Before executing strongly recomend to test it.]]> http://www.idbas.com/blog/index.php?entry=entry091231-165557

What is OTM?
OTM Stands for Oracle Transportation Management,earlier known as GLOG. Widely used in Transportation and Fleet world.Scalability allows for high availability of OTM by providing automatic failover to another application server.


Components of OTM :

WebServer :Oracle Transportation Management Web server uses the Apache Web server and the Tomcat Java servlet engine to manage the user interface and communications with the Application server.

Application Server
The Application Server controls the application logic and communicates that logic between the Web server and the Database. The Application server runs Oracle Application Server software from Oracle (all supported platforms), WebLogic Server software from BEA Systems (all supported platforms) or WebSphere Application Server from IBM (AIX and Linux only) to manage system-level details and operates at the center of the Oracle Transportation Management multi-tier architecture. In this architecture, business logic is executed in the application server, rather than in client applications. The resulting "thin" client, allows the client to manage the presentation layer, the application server to Manage the business logic, and the back-end data services to manage the data.

And Database Server.

High Scalability : Following are types configurations for application servers usage.

1. Active -Active Servers, based on weighted Routing we can distribute the load to each servers.
2. Active - Passive. (1Failover cluster with 1 Passive server).
3.Domain Separation.
1Default cluster with n active servers.(M Domain clusters, each dedicated to a specified domain with 1 active server).
Example : Company has two web portals,it can used to distribute application servers based on weight routing on each server.
4.Functional Separation.
1 default cluster with n active servers.This will help to differentiate the load between schedule work and other related work.


Advance Topologies :
Partition Bulk plan Scalability : These are the most advance high availability, it would be good idea to build a strong testing strategy before implementing.
This type to clusters scalability helps to further drill down on weighted routing based on Bulk plan Partitions. Parent job spawn a child process to use scalability.
And other advance scalability options are :
OAQ Scalability : Oracle Advance queuing Scalability.
Integration and staging Scalability.

Few Tips about installation :

1.Please clearly follow the installation document.
2.Need root access to start and stop web tier and application tiers.
3.changeme is the default paswd for dba.admin username.
4.Scalabilty is the only option for high availability.
5.Any change to the application and web server propettyfiles would require services bounce.
6.It is very is to diagnose the issue with Oracle structured logging mechanism.
7.Jar files are key files,please make sure you have good backup during patching.
8.Following are sample domains in OTM.
GUEST – This is a sample domain that you can use to get started using Oracle Transportation Management and experiment with all of the software functionality and data.
SERVPROV – This is a special domain for use with service providers. Special logic is built into Oracle Transportation Management to provide security for service providers that access Oracle, Transportation Management for web tendering. This is the domain that must be used for all Service providers.
PUBLIC – This domains stores public data that can be accessed by all other domains.

@ idbas.com with our domain expertise, we are ready to setup and support your Deployment strategy on Oracle application server or Oracle Weblogic servers for OTM. Please reach us for more information. info@idbas.com


Source :Oracle Installation ,Administration and scalability document from Oracle.
Disclaimer : Above details are for information only .We strongly recommend to test before using production system.
]]> http://www.idbas.com/blog/index.php?entry=entry091230-175931
]]> http://www.idbas.com/blog/index.php?entry=entry091230-174204
Problem : A particular query was taking more than 30 min and was still executing on warehouse database before it had be killed .
This  earlier used to take less than a min .
 
Solution :
1.  Found the missing Statistics for salesorderplaneitem table.
2.  Right Mouse Clicked on the Result pane and created the statistics for salesorderplaneitem.
3.  Ran sp_updatestats for the database and the query started executing in 8 min .
 
The problem again re-occurred.
1. Checked the Database properties and found both Auto_Create_Statistics & Torn_Page_Detection was not configured on warehouse object.
2. Checked the Display Estimated Execution Plan & found in the  Statistics, the cost to be 92% for salesorderplaneitem.
3. Checked when the statistics was updated using DBCC SHOW_STATISTICS.
4.Ran update stats for the particular table using update statistics salesorderplaneitem and brought down the cost factor to 5% .
 
The query executed in 2sec .

@idbas.com we take every optimize approach to accomplish.We always give 100% database availability using Framework Approach.
Disclaimer : We are not responsible for commands.Information for reference only,not for execution. Before executing strongly recomend to test it.]]>