The templates are available from the Oracle E-Delivery web site and are available in both OEL 32bit Linux as well as 64bit Linux.

While we are demonstrating a test or development RAC setup where a single Oracle VM server is hosting both the guest nodes, a ‘Production’ type envrionment is also supported using the Oracle VM Templates where we have multiple OVM Servers and the shared disks are configured as ‘phy’ devices which are then passed on to the guest Oracle VM’s. So there are a number of Oracle VM Servers in this case.

In this test environment, however, the shared disk is configured as ‘file’ devices and both guests can run on the same Oracle VM Server.

In other words we have one single DOM-0 and we use files located in the DOM-0 to emulate the shared disks for the Oracle RAC ‘cluster’ nodes running in DOM-U.

The template requires about 40 Gb of disk space and assumes that we have at least 4 Gb of RAM available. In our test case, we have allocated 4 Gb of RAM to each Virtual Machine.

The template assumes the following are already in place.

a) We have already installed Oracle VM Server (we have used OVM 2.2)
b) We have already installed and configured Oracle VM Manager (OVM 2.2)
c) allocated 7 unused IP addresses and registered in our DNS

Assuming a two node RAC cluster, that is

2 IP’s for the public host names
2 IP’s for the VIP
3 IP’s for the SCAN

The high level steps involved would be:

• Download the zip files from Oracle E-Delivery site
  
• Import the Virtual Machine template into OVM
  
• Create the ASM shared storage

• Create the VM’s from the 11gR2 RAC template
• Add the shared disks to the VM’s
• Complete the first boot interview
• run the script buildcluster.sh which will install the 11g R2 Grid Infrastructure and database software and also create the ORCL RAC database
• Customize as required
  

Some resources

• Installing OVM Manager 2.2

• Installing OVM Server 2.2

• RAC OVM Templates Overview

• RAC 11g Release 2 (11.2.0.2.0) Oracle VM templates Test and Development Configuration”
  

We need to unzip the 3 files we have downloaded fron the E-Delivery web site and then also extract the 11gR2 RAC template.

This is all done in the /OVS/seed_pool/ directory on the OVM Server

[root@kens-ovm-002 OVS]# mv p10113572*.zip ./tmp

[root@kens-ovm-002 OVS]# cd tmp

[root@kens-ovm-002 tmp]# ls
p10113572_10_Linux-x86-64_1of3.zip  p10113572_10_Linux-x86-64_2of3.zip  p10113572_10_Linux-x86-64_3of3.zip

[root@kens-ovm-002 tmp]# unzip -q p10113572_10_Linux-x86-64_1of3.zip
[root@kens-ovm-002 tmp]# unzip -q p10113572_10_Linux-x86-64_2of3.zip
[root@kens-ovm-002 tmp]# unzip -q p10113572_10_Linux-x86-64_3of3.zip

[root@kens-ovm-002 tmp]# cd /OVS/seed_pool/

[root@kens-ovm-002 seed_pool]# tar xzf /OVS/tmp/OVM_EL5U4_X86_64_11202RAC_PVM-1of2.tgz

[root@kens-ovm-002 seed_pool]# cat /OVS/tmp/OVM_EL5U4_X86_64_11202RAC_PVM-2of2-parta.tgz /OVS/tmp/OVM_EL5U4_X86_64_11202RAC_PVM-2of2-partb.tgz | tar xzf -

Importing the Virtual Machine Template

We will now import this template using OVM Manager.

Resources >> Virtual Machine Templates and click on Import button

Our template is stored in the /OVS/seed_pool on our OVM Server machine. So we select the option “Select from Server Pool (Discover and Register)”

From the Virtual Machine Template Name drop down list select the 11gR2 RAC PVM.

In the Virtual Machine System Password field, we need to enter the root password. The template assumes the root password to be ‘ovsroot’ and if we set it to anything else as in our case, we got an error.

But there is a configuration file we can edit to change the default root password to whatever we desire and I will show you how that’s done in a bit.

Click on the Confirm button

We will see that the status of the Virtual Machine Template is now changed to “Importing”

We will see that the status of the Virtual Machine Template is now changed to “Pending”

We need to click on the Approve button

We see that the template status has now changed to “Active” and we can now use this template to set up our 11g R2 RAC environment.

Create the shared ASM disks

Resources >> Shared Virtual Disks >> Create

Select the appropriate Server Pool Name and Group Name.

We will be creating 5 shared virtual disks – ASM1,ASM2,ASM3, ASM4 and ASM5 each of size 2048 Mb or 2 Gb

Click on Confirm

After the first disk ASM1 has been created, click on the Create button and similarly create the other ASM disks in the same manner

We should now see all the five Shared Virtual Disks with status Active

Create the Virtual Machines from the imported 11gR2 RAC template

Click on the Create Virtual Machine button

Select the option Create virtual machine based on virtual machine template

Select the template OVM_EL5U4_X86_64_11202RAC_PVM

Important

Do not forget to add a SECOND Virtual Network Interface – in this case the name is VIF1 and Bridge name is xenbr1

Keep a note of the console password as we will need to enter this at a later stage

Note the status of the Virtual Machine is now ‘Creating’

After the Virtual Machine is created, it is in a Powered Off state

Click on the Configure tab as we would like to change the Memory Size of this VM from 2 Gb to 4 Gb

Change the Maximum Memory Size and Memory Size to 4096

Change the Memory Size of the second VM as well to 4096 MB and now we will see that both the VM’s have been created and are in the Powered Off state

Attach the Shared Disks to both the VM’s

Select one of the Virtual Machines and click on the Configure tab

Select the Storage tab and then the Attach/Detach Shared Virtual Disk

We see the 5 Shared Virtual Disks which we had earlier created – Click on Move All

We can now see that the 5 ASM Shared Disks have been attached to the VM and the disk status is Active

Repeat the same process for the second VM and attach the same 5 ASM disks to that Virtual Machine as well

Complete the First Boot Interview

Power On both the VM’s

Once the machines are powered on, click on the Console tab and launch both the consoles

This is called the First Boot Interview stage.

On the first screen we enter YES and on the second screen we enter NO.

Only after we have entered NO on the second console does the install continue.

We need to now provide the IP’s for the Public, Private and Virtual hostnames

We need to enter the SCAN name for this cluster. Note – we should have registered the SCAN in our DNS already and have assigned 3 unused IP addresses to the SCAN

Install Grid Infrastructure and Oracle 11g R2 RAC

We now need to run the buildcluster.sh script. This needs to be on only ONE of the VM’s – the one we have designated to be the first node of the cluster.

Note – because we had not used the standad root password ovsroot, we find that the script has failed becuase it tries to establish a passwordless ssh session which fails.

We needed to edit the /u01/racovm/params.ini file and change the entry for ROOTUSERPASSWORD to the one we want.

Have a read through of the params.ini file as it is through this file that we can add a lot of customisation to the RAC cluster build.

While the RAC installation is in progress, we can monitor the same by doing a tail -f of the progress-racovm.out file located on /tmp

]]> http://gavinsoorma.com/2011/11/3366/feed/ 0 http://gavinsoorma.com/2011/11/oracle-11g-fusion-middleware-installation-on-linux/ http://gavinsoorma.com/2011/11/oracle-11g-fusion-middleware-installation-on-linux/#comments Fri, 25 Nov 2011 00:31:59 +0000 Gavin Soorma http://gavinsoorma.com/?p=3310 This note describes the (long) process of installing an 11g Fusion Middleware environment with Forms, Reports, Discoverer and Portal on an Oracle Enterprise Linux 5.6 64 Bit platform.

If you are not using Oracle Portal , then you do not need to configure Identity Management and Single Sign On which is a real big pain as there is no 11g SSO as yet and we have to install, upgrade and configure 10g SSO/DAS as well with a number of patches.

These are the top level steps:

Continue reading Oracle 11g Fusion Middleware Installation on Linux

A number of Metalink support notes discuss this topic like:

Database Tablespace Full Metric Alerts not Generated In Grid Control [ID 357049.1]
Troubleshooting a Database Tablespace Used(%) Alert problem [ID 403264.1]
Oracle Recommended Patches For Tablespace Monitoring using the Tablespace Space Used % metric [ID 849498.1]

What happens if autoextend is turned on but in some cases we have specified an upper limit for the MAXSIZE parameter or have just left it to UNLIMITED? How will OEM handle those cases?

To clarify things, OEM alert notification mechanism will query the DBA_TABLESPACE_USAGE_METRICS view to determine the Used % in a tablespace in relation to the Warning and Critical thresholds which have been set. This is true for Oracle 10g and above and if we are using Locally Managed tablespaces.

It does not look at DBA_FREE_SPACE or DBA_TABLESPACES for that matter to determine how much of free space is there in a tablespace. In fact many of us had these custom scripts which query these two views and in case of databases which had many tablespaces like an Oracle Applications environment, this query could take a long time to return the results and was quite a ‘costly’ query.

It is important to note that this metric takes into consideration the available disk space into which the tablespace’s datafile can extend if autoextend is turned on.

With a 8k block size, on most Unix systems the maximum a datafile can extend upto with autoextend turned on is 32 Gb. In that case the DBA_TABLESPACE_USAGE_METRICS view will use 32 Gb as the basis for calculating the available free space in the tablespace because if a tablespace consists of one datafile, that is the maximum size the tablespace can grow to.

In the figure below of the OEM Tablespaces screen (Database Instance Home Page –> Administration –> Tablespaces), we see some tablespaces which have used space greater 95%, but we have not received any alert from OEM even though the Critical threshold has been crossed.

Why is that the case?

This is because of the AUTOEXTEND setting. In the figure above, some tablespaces have datafiles which have autoextend turned on and some do not have autoextend enabled.

In this case, the datafile of the SYSTEM tablespace does have autoextend turned on, and even though the SYSTEM tablespace is 99% full, OEM does not consider it a candidate for an alert notification because the DBA_TABLESPACE_USAGE_METRICS view shows it only as 15.71% used.

SQL>  SELECT TABLESPACE_NAME TBSP_NAME
  2  , USED_SPACE
  3  , TABLESPACE_SIZE TBSP_SIZE
  4  , USED_PERCENT
  5  FROM SYS.DBA_TABLESPACE_USAGE_METRICS
  6  order by 4;

TBSP_NAME                      USED_SPACE  TBSP_SIZE USED_PERCENT
—————————— ———- ———- ————
…
…

SYSTEM                              91512     582427   15.7121837

So how did Oracle compute this 15.71% used figure? (91512/582427*100)

Note the USED_SPACE and TBSP_SIZE columns are in blocks - in this case the DB_BLOCK_SIZE of the database is 8k or 8192 bytes.

The maximum size which the file can extend upto via autoextend is 582427 blocks or 4450 Mb

582427 (blocks) *8192 (each block size) / 1048576 (to convert to MB) = 4450 MB

The SYSTEM tablespace is made up of one datafile and that datafile is about 714 MB used. (91512* 8192/1048576)

Let us create a tablespace TESTME of size 50 Mb.

SQL>  create tablespace testme
  2  datafile '/u01/oradata/HDESK11G/testme01.dbf' size 50m;

We now create a table and use the TESTME tablespace to store the table data

SQL> create table sh.test
  2   tablespace testme
  3   as select * from sh.sales;

Table created.

Since we have not specified any AUTOEXTEND for the datafile of the TESTME tablespace, the TBSP_SIZE column for this tablespace shows 6400 blocks which is equal to 50 MB and the tablespace is about 74% full after we have created and populated the TEST table.

TBSP_NAME                      USED_SPACE  TBSP_SIZE USED_PERCENT
------------------------------ ---------- ---------- ------------
...
...
SYSTEM                              91512     552558   16.5615193
TESTME                               4736       6400           74

Let us now force a space used alert to be fired from OEM

SQL> insert into  sh.test
  2  select * from sh.sales;
insert into  sh.test
                *
ERROR at line 1:
ORA-01653: unable to extend table SH.TEST by 128 in tablespace TESTME

We can see that the tablespace TESTME is now 100% full and an alert has been raised as well.

TBSP_NAME                      USED_SPACE  TBSP_SIZE USED_PERCENT
------------------------------ ---------- ---------- ------------
...
...
TESTME                               6400       6400          100

SELECT REASON
, METRIC_VALUE
, MESSAGE_TYPE
, TO_CHAR(CREATION_TIME,'DD-MON-YYYY HH24:MI:SS')
, HOST_ID
FROM SYS.DBA_OUTSTANDING_ALERTS;

SQL> /

REASON
----------------------------------------------
METRIC_VALUE MESSAGE_TYPE TO_CHAR(CREATION_TIM
------------ ------------ --------------------
HOST_ID
----------------------------------------------
Tablespace [TESTME] is [100 percent] full
         100 Warning      15-NOV-2011 08:38:39
kens-orasql-001.corporateict.domain

Now we turn on AUTOEXTEND for the datafile of the TESTME tablespace - note in this case we specify the MAXIZE as UNLIMITED

SQL> alter database datafile '/u01/oradata/HDESK11G/testme01.dbf' autoextend on next 100M maxsize unlimited;

Database altered.

Now we see that the DBA_TABLESPACE_USAGE_METRICS view shows the tablespace as only 1.37% used.

Currently 50 MB is being used (the file size initially was 50 Mb and all the space in the datafile was used when we inserted rows into the table), but the TBSP_SIZE is showing 466770 (blocks) which is about 3646 Mb ( 466770*8192/1048576).

So why has Oracle chosen this figure?

If we run a df -k command, we see that the /u01 file system where this datafile of the TESTME tablespace is located has about 3.7 GB of free space and with AUTOEXTEND turned on, Oracle estimates that this is the maximum that the datafile can extend upto.

TBSP_NAME                      USED_SPACE  TBSP_SIZE USED_PERCENT
------------------------------ ---------- ---------- ------------
...
....
TESTME                               6400     466770   1.37112497

SQL> !df -k |grep u01

Filesystem           1K-blocks      Used Available Use% Mounted on

/dev/xvdb1            20635700  16007140   3780324  82% /u01

Now what happens if we instead specify an upper file size limit to which the datafile can extend upto with autoextend turned on.

SQL> alter database datafile '/u01/oradata/HDESK11G/testme01.dbf' autoextend on next 100M maxsize 500M;

Database altered.

We now see that the TESTME tablespace is 10% full because the maximum size that this tablespace can now grow to is 64000 blocks or 500 Mb and currently it has 50 Mb of used space.

The tablespaces screen from OEM however shows the tablepace as 33% full because it has grown the datafile to 150 MB from 50 Mb as we had specified the NEXT increment as 100 MB when we had enabled autoextend for the datafile of the TESTME tablesapace.

TBSP_NAME                      USED_SPACE  TBSP_SIZE USED_PERCENT
------------------------------ ---------- ---------- ------------
...
...
TESTME                               6400      64000           10

But what we see from OEM Tablespaces screen (Database Instance Home Page --> Administration --> Tablespaces), the tablespace is 33% full and not 10%.

Remember, this is not what OEM will use to determine the tablespace free space available and when it needs to raise an alert if the space usage threshold is crossed.

The Metalink note Data Guard Support for Heterogeneous Primary and Physical Standbys in Same Data Guard Configuration [ID 413484.1] describes the supported cross platform combinations between a primary and standby database.

In Oracle 10g, we could have a 32 bit and 64 bit Primary/Standby combination on some supported platforms and in 11g this has been extended to cover heterogenous platforms for the Primary and Standby database.

The environment used is as follows:

Primary

Windows 7 64 bit
11g Release 2
DB_UNIQUE_NAME=orcl

Active Standby

Oracle Enterprise Linux 5.7 64 bit
11g Release 2
DB_UNIUE_NAME=orcl_dr

Read the rest of the note Oracle 11g Cross platform Active Standby – Windows Primary database and Linux Active Standby

I tried out the procedure listed by the author and thought I’d share this with those who may not be aware of this note and must say thanks to the author Ginny Henningsen.

When I have tried to install Oracle 11g on Linux, in most cases I have found that the OUI either reports that some packages are missing (typically unixODBC or libaio) or some kernel parameters are not appropriately set. In this case however, we download and install an RPM package called oracle-validated and the oracle-validated RPM does a number of things for us.

Note that the oracle-validated RPM package is accessible through the Oracle Unbreakable Linux Network (ULN) which requires a support contract or from the Oracle public yum repository.

First we needed to configure the Linux machine to be able to connect to the internet using the corporate proxy and we did this by editing the /etc/profile.d/proxy.sh file with some environment variables like FTP_PROXY and HTTP_PROXY

export http_proxy=http://test-apache-001-dev.corporateict.domain:8123/
export ftp_proxy=http://test-apache-001-dev.corporateict.domain:8123/
export no_proxy=.corporateict.domain
export HTTP_PROXY=http://test-apache-001-dev.corporateict.domain:8123/
export FTP_PROXY=http://test-apache-001-dev.corporateict.domain:8123/

Login once again as root to invoke the changes made to the environment.

We then use wget to download the file which has the repository locations

[root@kens-orasql-001 yum.repos.d]# wget http://public-yum.oracle.com/public-yum-el5.repo

We have to update this file and change the line enabled=0 to enabled=1 for the tag which is appropriate to our Linux version – in our case we were using OEL 5.7 so we have to update these lines:

Change

[ol5_u7_base]
name=Oracle Linux $releasever – U7 – $basearch – base
baseurl=http://public-yum.oracle.com/repo/OracleLinux/OL5/7/base/$basearch/
gpgkey=http://public-yum.oracle.com/RPM-GPG-KEY-oracle-el5
gpgcheck=1
enabled=0

to

[ol5_u7_base]
name=Oracle Linux $releasever – U7 – $basearch – base
baseurl=http://public-yum.oracle.com/repo/OracleLinux/OL5/7/base/$basearch/
gpgkey=http://public-yum.oracle.com/RPM-GPG-KEY-oracle-el5
gpgcheck=1
enabled=1

We then ran a yum update followed by the yum install oracle-validated

[root@kens-orasql-002 yum.repos.d]# yum update
Loaded plugins: rhnplugin, security
This system is not registered with ULN.
ULN support will be disabled.
ol5_u7_base                                                                                                                                    | 1.1 kB     00:00
ol5_u7_base/primary                                                                                                                            | 1.5 MB     00:07
ol5_u7_base                                                                                                                                                 4605/4605
Skipping security plugin, no data
Setting up Update Process
No Packages marked for Update

[root@kens-orasql-002 yum.repos.d]# yum install oracle-validated
Loaded plugins: rhnplugin, security
This system is not registered with ULN.
ULN support will be disabled.
Setting up Install Process
Resolving Dependencies
--> Running transaction check
---> Package oracle-validated.x86_64 0:1.1.0-14.el5 set to be updated
--> Processing Dependency: kernel-uek >= 2.6.32 for package: oracle-validated
--> Processing Dependency: libXp.so.6 for package: oracle-validated
--> Processing Dependency: /usr/lib/libodbccr.so for package: oracle-validated
--> Processing Dependency: /usr/lib64/libodbccr.so for package: oracle-validated
--> Processing Dependency: unixODBC-devel for package: oracle-validated
--> Processing Dependency: libaio-devel for package: oracle-validated
--> Processing Dependency: sysstat for package: oracle-validated
--> Processing Dependency: /usr/lib64/libaio.so for package: oracle-validated
--> Processing Dependency: /usr/lib/libodbc.so.1 for package: oracle-validated
--> Processing Dependency: /usr/lib/libaio.so for package: oracle-validated
--> Processing Dependency: compat-gcc-34-c++ for package: oracle-validated
--> Processing Dependency: compat-gcc-34 for package: oracle-validated
--> Processing Dependency: libdb-4.2.so()(64bit) for package: oracle-validated
--> Processing Dependency: libodbc.so.1()(64bit) for package: oracle-validated
--> Running transaction check
---> Package compat-db.x86_64 0:4.2.52-5.1 set to be updated
---> Package compat-gcc-34.x86_64 0:3.4.6-4.1 set to be updated
---> Package compat-gcc-34-c++.x86_64 0:3.4.6-4.1 set to be updated
---> Package kernel-uek.x86_64 0:2.6.32-200.13.1.el5uek set to be installed
--> Processing Dependency: kernel-firmware >= 2.6.32-200.13.1.el5uek for package: kernel-uek
---> Package libXp.i386 0:1.0.0-8.1.el5 set to be updated
---> Package libaio-devel.i386 0:0.3.106-5 set to be updated
---> Package libaio-devel.x86_64 0:0.3.106-5 set to be updated
---> Package sysstat.x86_64 0:7.0.2-11.el5 set to be updated
---> Package unixODBC.i386 0:2.2.11-7.1 set to be updated
---> Package unixODBC.x86_64 0:2.2.11-7.1 set to be updated
---> Package unixODBC-devel.i386 0:2.2.11-7.1 set to be updated
---> Package unixODBC-devel.x86_64 0:2.2.11-7.1 set to be updated
--> Running transaction check
---> Package kernel-uek-firmware.noarch 0:2.6.32-200.13.1.el5uek set to be updated
--> Finished Dependency Resolution

Dependencies Resolved

======================================================================================================================================================================
Package                                     Arch                           Version                                         Repository                           Size
======================================================================================================================================================================
Installing:
oracle-validated                            x86_64                         1.1.0-14.el5                                    ol5_u7_base                          24 k
Installing for dependencies:
compat-db                                   x86_64                         4.2.52-5.1                                      ol5_u7_base                         1.6 M
compat-gcc-34                               x86_64                         3.4.6-4.1                                       ol5_u7_base                         4.3 M
compat-gcc-34-c++                           x86_64                         3.4.6-4.1                                       ol5_u7_base                          13 M
kernel-uek                                  x86_64                         2.6.32-200.13.1.el5uek                          ol5_u7_base                          25 M
kernel-uek-firmware                         noarch                         2.6.32-200.13.1.el5uek                          ol5_u7_base                         2.5 M
libXp                                       i386                           1.0.0-8.1.el5                                   ol5_u7_base                          22 k
libaio-devel                                i386                           0.3.106-5                                       ol5_u7_base                          12 k
libaio-devel                                x86_64                         0.3.106-5                                       ol5_u7_base                          11 k
sysstat                                     x86_64                         7.0.2-11.el5                                    ol5_u7_base                         187 k
unixODBC                                    i386                           2.2.11-7.1                                      ol5_u7_base                         830 k
unixODBC                                    x86_64                         2.2.11-7.1                                      ol5_u7_base                         835 k
unixODBC-devel                              i386                           2.2.11-7.1                                      ol5_u7_base                         743 k
unixODBC-devel                              x86_64                         2.2.11-7.1                                      ol5_u7_base                         799 k

Transaction Summary
======================================================================================================================================================================
Install      14 Package(s)
Upgrade       0 Package(s)

Total size: 50 M
Is this ok [y/N]: y
Downloading Packages:
warning: rpmts_HdrFromFdno: Header V3 DSA signature: NOKEY, key ID 1e5e0159
ol5_u7_base/gpgkey                                                                                                                             | 1.4 kB     00:00
Importing GPG key 0x1E5E0159 "Oracle OSS group (Open Source Software group) " from http://public-yum.oracle.com/RPM-GPG-KEY-oracle-el5
Is this ok [y/N]: y
Running rpm_check_debug
Running Transaction Test
Finished Transaction Test
Transaction Test Succeeded
Running Transaction
  Installing     : unixODBC                                                                                                                                      1/14
  Installing     : compat-gcc-34                                                                                                                                 2/14
  Installing     : sysstat                                                                                                                                       3/14
  Installing     : compat-db                                                                                                                                     4/14
  Installing     : unixODBC                                                                                                                                      5/14
  Installing     : unixODBC-devel                                                                                                                                6/14
  Installing     : unixODBC-devel                                                                                                                                7/14
  Installing     : libaio-devel                                                                                                                                  8/14
  Installing     : libaio-devel                                                                                                                                  9/14
  Installing     : kernel-uek-firmware                                                                                                                          10/14
  Installing     : libXp                                                                                                                                        11/14
  Installing     : compat-gcc-34-c++                                                                                                                            12/14
Change references of /dev/hd in /etc/fstab to disk label
  Installing     : kernel-uek                                                                                                                                   13/14
WARNING: No module ehci-hcd found for kernel 2.6.32-200.13.1.el5uek, continuing anyway
WARNING: No module ohci-hcd found for kernel 2.6.32-200.13.1.el5uek, continuing anyway
WARNING: No module uhci-hcd found for kernel 2.6.32-200.13.1.el5uek, continuing anyway
  Installing     : oracle-validated                                                                                                                             14/14

Installed:
  oracle-validated.x86_64 0:1.1.0-14.el5

Dependency Installed:
  compat-db.x86_64 0:4.2.52-5.1                         compat-gcc-34.x86_64 0:3.4.6-4.1                               compat-gcc-34-c++.x86_64 0:3.4.6-4.1
  kernel-uek.x86_64 0:2.6.32-200.13.1.el5uek            kernel-uek-firmware.noarch 0:2.6.32-200.13.1.el5uek            libXp.i386 0:1.0.0-8.1.el5
  libaio-devel.i386 0:0.3.106-5                         libaio-devel.x86_64 0:0.3.106-5                                sysstat.x86_64 0:7.0.2-11.el5
  unixODBC.i386 0:2.2.11-7.1                            unixODBC.x86_64 0:2.2.11-7.1                                   unixODBC-devel.i386 0:2.2.11-7.1
  unixODBC-devel.x86_64 0:2.2.11-7.1

Complete!
[root@kens-orasql-002 yum.repos.d]#

We can see that the oracle-validated RPM has been installed.

[root@kens-orasql-001 yum.repos.d]# rpm -qa |grep oracle-valid
oracle-validated-1.1.0-14.el5

We can also check and see that the user and groups have been created as well.

[root@kens-orasql-001 yum.repos.d]# cat /etc/passwd |grep oracle

oracle:x:54321:54321::/home/oracle:/bin/bash

[root@kens-orasql-001 yum.repos.d]# cat /etc/group |grep dba 
dba:x:54322:oracle

It has also updated the /etc/security/limits.conf file

[root@kens-orasql-001 yum.repos.d]# cat /etc/security/limits.conf

# Oracle-Validated setting for nofile soft limit is 131072
oracle   soft   nofile    131072

# Oracle-Validated setting for nofile hard limit is 131072
oracle   hard   nofile    131072

# Oracle-Validated setting for nproc soft limit is 131072
oracle   soft   nproc    131072

# Oracle-Validated setting for nproc hard limit is 131072
oracle   hard   nproc    131072

# Oracle-Validated setting for core soft limit is unlimited
oracle   soft   core    unlimited

# Oracle-Validated setting for core hard limit is unlimited
oracle   hard   core    unlimited

# Oracle-Validated setting for memlock soft limit is 50000000
oracle   soft   memlock    50000000

# Oracle-Validated setting for memlock hard limit is 50000000
oracle   hard   memlock    50000000

Finally, we can see that the /etc/sysctl.conf has been updated as well with some “Oracle-Validated” settings.

# For 11g, Oracle-Validated setting for fs.file-max is 6815744
# For 10g, uncomment 'fs.file-max = 327679', and comment 'fs.file-max = 6553600' entry and re-run sysctl -p
# fs.file-max = 327679
fs.file-max = 6815744

# Oracle-Validated setting for kernel.msgmni is 2878
kernel.msgmni = 2878

# Oracle-Validated setting for kernel.msgmax is 8192
kernel.msgmax = 8192

# Oracle-Validated setting for kernel.msgmnb is 65536
kernel.msgmnb = 65536

# Oracle-Validated setting for kernel.sem is '250 32000 100 142'
kernel.sem = 250 32000 100 142

# Oracle-Validated setting for kernel.shmmni is 4096
kernel.shmmni = 4096

# Oracle-Validated setting for kernel.shmall is 1073741824
kernel.shmall = 1073741824

# Oracle-Validated setting for kernel.shmmax is 4398046511104 on x86_64 and 4294967295 on i386 architecture. Refer Note id 567506.1
kernel.shmmax = 4398046511104

# Oracle-Validated setting for kernel.sysrq is 1
kernel.sysrq = 1

# Oracle-Validated setting for net.core.rmem_default is 262144
net.core.rmem_default = 262144

# For 11g, Oracle-Validated setting for net.core.rmem_max is 4194304
# For 10g, uncomment 'net.core.rmem_max = 2097152', comment 'net.core.rmem_max = 4194304' entry and re-run sysctl -p
# net.core.rmem_max = 2097152
net.core.rmem_max = 4194304

# Oracle-Validated setting for net.core.wmem_default is 262144
net.core.wmem_default = 262144

# For 11g, Oracle-Validated setting for net.core.wmem_max is 1048576
# For 10g, uncomment 'net.core.wmem_max = 262144', comment 'net.core.wmem_max = 1048576' entry for this parameter and re-run sysctl -p
# net.core.wmem_max = 262144
net.core.wmem_max = 1048576

# Oracle-Validated setting for fs.aio-max-nr is 3145728
fs.aio-max-nr = 3145728

# For 11g, Oracle-Validated setting for net.ipv4.ip_local_port_range is 9000 65500
# For 10g, uncomment 'net.ipv4.ip_local_port_range = 1024 65000', comment 'net.ipv4.ip_local_port_range = 9000 65500' entry and re-run sysctl -p
# net.ipv4.ip_local_port_range = 1024 65000
net.ipv4.ip_local_port_range = 9000 65500

# Oracle-Validated setting for vm.min_free_kbytes is 51200 to avoid OOM killer
vm.min_free_kbytes = 51200

Let us today take a look at two other functions STRCAT and STREXT.

STRCAT basically enables us to concatenate two strings or two character columns.

The STREXT function on the other hand is used to extract a portion of a string.

For example in our source database we have a table MYPHONE with the following structure:

SQL> desc myphone
 Name                                      Null?    Type
 ----------------------------------------- -------- ----------------------------
 AREA_CODE                                          NUMBER
 PHONE_NO                                           NUMBER

In the target we have the same table MYPHONE with a different structure:

SQL> desc myphone
 Name                                      Null?    Type
 ----------------------------------------- -------- ----------------------------
 PHONE_NO                                           VARCHAR2(30)

So in the target database, the PHONE_NO column includes both the area code as well as the phone number with a ‘-‘ character as well between both the values, while in the source database the table has two columns – one for the area code and one for the phone number.

We will use SRTCAT function in this case.

Remember that since the source and target tables differ in structure, we need to run the DEFGEN utility in the source GoldenGate environment to generate the definitions file and then copy this across to the target GoldenGate environment..

These are the contents of the extract parameter file:

extract ext
userid ggs_owner, password ggs_owner
rmthost venom , mgrport 7809
rmttrail ./dirdat/gg
table ggs_owner.myphone;

These are the contents of the replicat parameter file – note how the STRCAT function is being used here.

REPLICAT rep
sourcedefs /export/home/oracle/gg/dirsql/myphone.sql
USERID ggs_owner, PASSWORD ggs_owner
MAP ggs_owner.myphone, TARGET ggs_owner.myphone,
COLMAP (phone_no=@STRCAT("0",area_code,"-",phone_no));

We now start the extract and replicat processes and insert some records in the source table.

The source table now has the following rows:

SQL> select * from myphone;

 AREA_CODE   PHONE_NO
---------- ----------
         8   92471136
         4   17213124
         2     818188

Let us see how these rows have been transformed on the target database.

SQL> select * from myphone;

PHONE_NO
------------------------------
08-92471136
04-17213124
02-818188

We can extend this example and use both STRCAT and STREXT together to achieve the data transformation that we desire.

For example in our source MYSSID table, we have data like this:

SQL> select * from myssid;

      SSID
----------
 123456787
 123456788
 123456789

But on the target database, in the same table the data has been displayed like this:

SQL> select * from myssid;

SSID
--------------------
123-456-787
123-456-788
123-456-789

So how did we achieve this?

Have a look at the replicat parameter file and note how we have used first the STREXT function to extract 3 characters at a time – the first 3, then the next 3 and finally the last 3 and then used STRCAT to concatenate the result of the STREXT functions together with a ‘-‘ character acting as the string separator.

REPLICAT rep
sourcedefs /export/home/oracle/gg/dirsql/myssid.sql
USERID ggs_owner, PASSWORD ggs_owner
MAP ggs_owner.myssid, TARGET ggs_owner.myssid,
COLMAP (ssid=@STRCAT(@STREXT(ssid,1,3),"-",@STREXT(ssid,4,6),"-",@STREXT(ssid,7,9) ));

To illustrate this example, we have a source and target table called EMPSAL and the structure of the tables in both databases is slightly different.

In the source table we have the JOB column and in the target database we have a TAX_BRACKET column.

In the first example, we will use CASE function to populate the target database TAX_BRACKET column based on the value for the JOB column in the source database.

In the second example, we will use the EVAL function to populate the target database TAX_BRACKET column based on a particular range of values for the SAL column in the source table.

Remember that since the table structure in source and target are different, we need to run the DEFGEN utility to generate a definitions file which we will then copy to the target GoldenGate environment and instead of using the ASSUMEDEFS parameter in the replicate parameter file, we use the SOURCEDEFS parameter instead to indicate the location of the definitions file which we have generated and copied over from the source GoldenGate environment.

Source

SQL> desc empsal
 Name                                      Null?    Type
 ----------------------------------------- -------- ----------------------------
 ENAME                                              VARCHAR2(20)
 SAL                                                NUMBER
 JOB                                                VARCHAR2(12)

Target

SQL> desc empsal
 Name                                      Null?    Type
 ----------------------------------------- -------- ----------------------------
 ENAME                                              VARCHAR2(20)
 SAL                                                NUMBER
 TAX_BRACKET                                        VARCHAR2(10)

We now create the definitions file for the EMPSAL table and copy the generated definitions file to the target GoldenGate environment.

GGSCI (sunos1) 4> edit params defgen

"/export/home/oracle/gg/dirprm/defgen.prm" 3 lines, 97 characters
defsfile ./dirsql/empsal.sql
userid ggs_owner password ggs_owner
table ggs_owner.empsal;

sunos1:/export/home/oracle/gg $ ./defgen paramfile /export/home/oracle/gg/dirprm/defgen.prm

sunos1:/export/home/oracle/gg/dirsql $ scp -rp empsal.sql oracle@sunos2:/export/home/oracle/gg/dirsql

Let us have a look at the Extract parameter file – quite normal and straight forward here.

extract ext
userid ggs_owner, password ggs_owner
rmthost sunos2 , mgrport 7809
rmttrail ./dirdat/gg
table ggs_owner.empsal;

On the target, these are the contents of the replicat parameter file. Note the COLPMAP clause. Since both the tables have common columns ENAME and SAL, we do not have to explicitly map them and we can just specify USEDEFAULTS.

The TAX_BRACKET column on the target table is being populated using the CASE function which evaluates the values for the JOB column in the source table.

If the value for JOB is ‘OPERATOR’, then the TAX_BRACKET column is assigned the value ‘LOW’. If it is ‘DBA’ then the TAX_BRACKET column is assigned the value ‘MID’ and if the value is ‘MANAGER’, then the TAX_BRACKET column is assigned the value ‘HIGH’. If the value for JOB is none of the values, then a default value of ‘UNKNOWN’ is assigned.

REPLICAT rep
sourcedefs /export/home/oracle/gg/dirsql/empsal.sql
USERID ggs_owner, PASSWORD ggs_owner
MAP ggs_owner.empsal, TARGET ggs_owner.empsal,
COLMAP (
usedefaults, tax_bracket=@case(JOB, "OPERATOR","LOW","DBA", "MID","MANAGER","HIGH", "UNKNOWN"));

Let us now insert some values into the source table.

SQL> insert into empsal values
  2  ('JOHN',10000, 'DBA');

1 row created.

SQL> insert into empsal values
  2  ('JACK',5000, 'OPERATOR');

1 row created.

SQL> insert into empsal values
  2  ('TOM',20000, 'MANAGER');

1 row created.

SQL>  insert into empsal values
  2  ('MARY', 6000, 'ANALYST');

1 row created.

SQL> commit;

Commit complete.

SQL> select * from empsal;

ENAME                       SAL JOB
-------------------- ---------- ------------
JOHN                      10000 DBA
JACK                       5000 OPERATOR
TOM                       20000 MANAGER
MARY                       6000 ANALYST

Let us now see how the data has been replicated and how the transformation has taken place on the target.

SQL> select * from empsal;

ENAME                       SAL TAX_BRACKE
-------------------- ---------- ----------
JOHN                      10000 MID
JACK                       5000 LOW
TOM                       20000 HIGH
MARY                       6000 UNKNOWN

Let us now truncate both the tables and see the next example using EVAL instead of CASE.
We stop both the extract and replicat processes and will now change the replicat parameter file on the target to include the EVAL function.
So this is how our replicat parameter file will look like.
REPLICAT rep
sourcedefs /export/home/oracle/gg/dirsql/empsal.sql
USERID ggs_owner, PASSWORD ggs_owner
MAP ggs_owner.empsal, TARGET ggs_owner.empsal,
COLMAP (
usedefaults, &
tax_bracket=@eval(sal < 10000,"LOW", sal >= 20000 , "HIGH",sal >= 10000, "MID"));

Start the extract and replicat processes again and let us now insert the same set of values in the source table as we did before and see how they are now replicated and transformed on the target.
SQL> select * from empsal;

ENAME                       SAL TAX_BRACKE
-------------------- ---------- ----------
JOHN                      10000 MID
JACK                       5000 LOW
TOM                       20000 HIGH
MARY                       6000 LOW

For those rows where the SAL value was less than 10000, a value of 'LOW' was assigned to the TAX_BRACKET column and if the value was equal to or greater than 10000 (but less than 20000), a value of 'MID' is assigned and all SAL values above 20000 are assigned 'HIGH'.
Have a read of the all the diferent column conversion functions mentioned in the GoldenGate reference guide and there are quite a few. Let me know if you found any others which are useful and kindly share your experience with the community!
]]>
			http://gavinsoorma.com/2011/08/using-the-goldengate-column-conversion-functions-case-and-eval/feed/
		0
		
		
		
		http://gavinsoorma.com/2011/08/passed-the-1z0-539-exam-am-now-an-oracle-certified-goldengate-implementation-specialist/
		http://gavinsoorma.com/2011/08/passed-the-1z0-539-exam-am-now-an-oracle-certified-goldengate-implementation-specialist/#comments
		Thu, 18 Aug 2011 02:44:21 +0000
		Gavin Soorma
				

		http://gavinsoorma.com/?p=3239
		 I have recently passed the Oracle GoldenGate 10 Essentials exam which recognizes you as a Oracle GoldenGate 10 Certified Implementation Specialist.

Note that the exam is currently based on GoldenGate 10.4 and not the 11g version of GoldenGate which has been out now for quite some time.
I must point out that while there are a number of ‘brain dumps’ like Pass4Sure and Test King to ‘help’ you pass the exam, I have seen the questions and answers that are provided by them and in most cases I found the answers to be provided to be grossly incorrect and wrong. So beware!
The best preparation for the exam is to go through the following sets of documentation:
Oracle Installation and Setup Guide
Reference Guide
Administration Guide
These and other 10.4 GoldenGate documentation can be obtained online  here …
I am now available to provide my services in any GoldenGate implementation project around the world as well 
Continue reading Passed the 1Z0-539 Exam – am now an Oracle Certified GoldenGate Implementation Specialist

]]>
			 I have recently passed the Oracle GoldenGate 10 Essentials exam which recognizes you as a Oracle GoldenGate 10 Certified Implementation Specialist.

Note that the exam is currently based on GoldenGate 10.4 and not the 11g version of GoldenGate which has been out now for quite some time.
I must point out that while there are a number of ‘brain dumps’ like Pass4Sure and Test King to ‘help’ you pass the exam, I have seen the questions and answers that are provided by them and in most cases I found the answers to be provided to be grossly incorrect and wrong. So beware!
The best preparation for the exam is to go through the following sets of documentation:
Oracle Installation and Setup Guide

Reference Guide

Administration Guide
These and other 10.4 GoldenGate documentation can be obtained online  here …
I am now available to provide my services in any GoldenGate implementation project around the world as well as for consultancy on any GoldenGate related subjects.
You can contact me at +61417713124 or via email to gavin.soorma@oncalldba.com.au to discuss anything further.
Happy GoldenGating!
]]>
			http://gavinsoorma.com/2011/08/passed-the-1z0-539-exam-am-now-an-oracle-certified-goldengate-implementation-specialist/feed/
		0
		
		
		
		http://gavinsoorma.com/2011/08/does-goldengate-support-oracle-flashback/
		http://gavinsoorma.com/2011/08/does-goldengate-support-oracle-flashback/#comments
		Wed, 17 Aug 2011 03:16:21 +0000
		Gavin Soorma
				

		http://gavinsoorma.com/?p=3235
		There is an Oracle Metalink Note (Does Goldengate Support The New Oracle 10g Flashback Feature? [ID 966212.1]) which discusses this aspect, but I thought that I would like to see for myself how does the Extract and Replicat processes behave if we issue any FLASHBACK command on the source database.

The Metalink note states that GoldenGate supports the following Flashback operations:
Flashback Table to SCN
Flashback Table to timestamp
Flashback Table to restore point
and does not support operations like:
Flashback Table to before drop
Flashback database
I have tested both the Flashback Table To Timestamp as well as the Flashback Table To Before Drop operations and find that both seem to be supported by GoldenGate and replication continues inspite of these commands being issued on the source database.
Let us see the examples of the same.
I have an Oracle 11g Goldengate (11.1.1) environment configured with the source and target databases of version 11.2.0.2.
Extract and Replicat processes are running 
Continue reading Does GoldenGate support Oracle Flashback?

]]>
			There is an Oracle Metalink Note (Does Goldengate Support The New Oracle 10g Flashback Feature? [ID 966212.1]) which discusses this aspect, but I thought that I would like to see for myself how does the Extract and Replicat processes behave if we issue any FLASHBACK command on the source database.

The Metalink note states that GoldenGate supports the following Flashback operations:
Flashback Table to SCN

Flashback Table to timestamp

Flashback Table to restore point
and does not support operations like:
Flashback Table to before drop

Flashback database
I have tested both the Flashback Table To Timestamp as well as the Flashback Table To Before Drop operations and find that both seem to be supported by GoldenGate and replication continues inspite of these commands being issued on the source database.
Let us see the examples of the same.
I have an Oracle 11g Goldengate (11.1.1) environment configured with the source and target databases of version 11.2.0.2.
Extract and Replicat processes are running and I am not using Data Pump in this case. The table configured for replication is called MYOBJECTS.
To test the Flashback, I am deleting all the rows from the table and then check on the target database if this delete operation has been propagated. 
We will then flashback to a point in time when all the rows were present in the table and then check both the source and target database to verify if the delete operation has been flashed back on both databases.
Source
SQL> select count(*) from myobjects;
  COUNT(*)

———-

     56687
SQL> select to_char(sysdate,’dd-mon-yyyy hh24:mi:ss’) from dual;
TO_CHAR(SYSDATE,’DD-

——————–

17-aug-2011 10:23:07
SQL> delete myobjects;
56687 rows deleted.
SQL> commit;
Commit complete.
Target 
SQL> select count(*) from myobjects;
  COUNT(*)

———-

         0
Source
SQL> conn / as sysdba

Connected.

SQL> alter table ggs_owner.myobjects enable row movement;
Table altered.
SQL>   flashback table ggs_owner.myobjects to timestamp to_timestamp (’17-aug-2011 10:23:07′,’dd-mon-yyyy hh24:mi:ss’);
Flashback complete.
SQL> select count(*) from ggs_owner.myobjects;
  COUNT(*)

———-

     56687
Target 
SQL> select count(*) from myobjects;
  COUNT(*)

———-

     56687
If we check the Extract and Replicat process reports, we can see that around the time we have performed the database flashback operation, there has been both extract and replicat activity 
Source
2011-08-17 10:23:58  INFO    OGG-01026  Rolling over remote file ./dirdat/yy000004.

2011-08-17 10:25:53  INFO    OGG-01026  Rolling over remote file ./dirdat/yy000005.

2011-08-17 10:25:58  INFO    OGG-01026  Rolling over remote file ./dirdat/yy000006.
Target
Switching to next trail file ./dirdat/yy000004 at 2011-08-17 10:23:59 due to EOF, with current RBA 9999940

Opened trail file ./dirdat/yy000004 at 2011-08-17 10:23:59
Switching to next trail file ./dirdat/yy000005 at 2011-08-17 10:25:54 due to EOF, with current RBA 9999930

Opened trail file ./dirdat/yy000005 at 2011-08-17 10:25:54
Switching to next trail file ./dirdat/yy000006 at 2011-08-17 10:25:58 due to EOF, with current RBA 9999882

Opened trail file ./dirdat/yy000006 at 2011-08-17 10:25:58
Let us now look at the Flashback Table To Before Drop operation. 
The Metalink note states that  this operation is NOT supported, but I found that in my case, GoldenGate was still peforming the online change synchronization even after we dropped the table and flashed it back.
Would like others in the community to try this out themselves and provide some feedback to me on how it went for them.
Here is the example.
We will use the same MYOBJECTS table and drop it and then flash it back from the recycle bin. Note – if you are using DDL replication as well, then recycle bin has to be turned off and in that case you will not be able to flashback the table from the recycle bin using the Flashback Table To Before Drop command.
Source
SQL> drop table myobjects;
Table dropped.
SQL> flashback table myobjects to before drop;
Flashback complete.
SQL> update myobjects set owner=’GAVIN’;
56687 rows updated.
SQL> commit;
Commit complete.
SQL> select distinct owner from myobjects;
OWNER

——————————

GAVIN
Target
SQL> select distinct owner from myobjects;
OWNER

——————————

GAVIN
Changes have been replicated!
]]>
			http://gavinsoorma.com/2011/08/does-goldengate-support-oracle-flashback/feed/
		0