Oracle Grid Infrastructure for a Standalone Server 12C – Installation

Hi, everyone.

Today in this article I will show you how to install and configure Oracle Grid Infrastructure for a Standalone Server 12c.
In my previous post, I showed how to install and prepare the Operating System to install Oracle Database.

If you don’t even have your environment yet, please start from here: Oracle Linux 6.7 Installation – VM VirtualBox

1. After install and configure OEL 6.7 (Oracle Enterprise Linux), we need the Oracle Grid Infrastructure for a Standalone Server 12c installation packages, for more details and download: Oracle Database 12c (Linux x86_64)

When download installation packages finish, we need to upload to the OS (Operating System) all packages, if you are using Windows platform I recommend use WinSCP, it’s a free (GPL License) graphic tool that use SSH protocol to transfer through network using TCP/IP.

But if your are using Linux or Mac, it will be simple, because SSH protocol comes natively in most of OS versions and SCP program it’s part of OpenSSH package.

When upload packages has finished, we will unpack using unzip program:

Unpack package called: linuxamd64_12c_grid_1of2.zip and after that, unpack package called: linuxamd64_12c_grid_2of2.zip

Execute: unzip linuxamd64_12c_grid_1of2.zip and unzip linuxamd64_12c_grid_2of2.zip

It will be created a directory called grid with the following content:

2. Before start install Oracle Grid Infrastructure for a Standalone Server, we need to be sure about every prerequisites (Preinstallation Tasks), if you are using Windows to follow this “HowTo” you will need a client to access OS using SSH protocol (PuTTY) and it will be needed a program to export the display (Xming) from VM to physical machine (desktop).

In the “PuTTY Configuration“, expand “Connection” menu, expand “SSH” menu, go to the “X11” option and check box “Enable X11 forwarding” and start Xming program:

PS:

Now you will be able to export the display of your VM to your physical machine (desktop).

3. In this step I will guide you how to check prerequisites before start installation setup. It is not uncommon run the installation software and after made every configuration the installation setup perform the prerequisites check with “Failed” status. To prevent it to happen, it’s more easy to run first the setup installation with parameter “-executePrereqs“, it will “jump” to the prerequisites check and after everything is compliance we can perform installation definitely.

If you followed my previous post or if you are installing own your own and followed the best practices reading the official documentation, you will have the same result as me “No results” for “Show Failed” option.

But if there is some prerequisites with failed status you need to fix and run “Check Again” to be sure that there is no more complaints.

PS:
To call the installation setup utility, execute: ./runInstaller -executePrereqs

3.1 After match every prerequisites, call again the installation setup utility but without “-executePrereqs” parameter, select “Install and Configure Oracle Grid Infrastructure for a Standalone Server“

3.2 Select your preferred language –> Next.

3.3 In the step 3, we need to fill the field “Disk group name“, choose the type of “Redundancy“, configure “Allocation Unit Size” (Keep it default 1 MB), let’s selected “Candidate Disks“, in the option “Change Discovery Path” type “/dev/sd*“, then OK –> Next.

PS:

[ERRATA] (01/05/2018), in some cases the use of /dev/* or /dev in the “Disk Discovery Path” or “asm_diskstring” parameter may hang the node. This is a known issue and it is reported in the following docs:

– Known Issues When asm_diskstring is Set to /dev/* (Doc ID 1921561.1)

– root.sh Hangs or Reboots Node on Non First Node if ASM Discovery String asm_diskstring is Set to /dev/ or /dev/* (Doc ID 1356567.1)

The first disk group “DG_BLOG_DATA” it will be used to store data, datafiles and etc: /dev/sdb1.

Selecting “External” redundancy we assume that disks are redundant by Storage Server or anything else.

“Disk Discovery Path” is the same value of “ASM_DISKSTRING” parameter and points to the directory that “ASM” uses to discover candidates disks or disks that is already in use by some disk group.

When “Candidate Disks” option is selected only disks that are not in use and has the right permissions are presented to use, preventing us to make any mistake.

3.4 Set SYS and ASMSMP users password:

3.5 If in your environment you already have an “Oracle Enterprise Manager Cloud Control 12c“, you can set up your grid/database to be managed and monitored:

But if you are not using, just click –> Next.

3.6 In the step 6, we are prompted to choose some OS groups. If we intend to use “Job Role Separation“, now is the time to determine each job role responsibility:

PS:

To this purpose we are not gonna configure separated user roles, this is more used in a big and huge companies, that has a good division of teams: OS, backup, storage, database, middleware, network team and etc. Inside of each team each one has separated duties.

3.7 In the field “Oracle base” type “/opt/oracle” and “Software location” field type “/opt/oracle/product/12.1.0/grid“:

3.8 On step 8 the “Inventory Directory” field, type: “/opt/oracle/oraInventory“

3.9 I really prefer to take control about every action in the installation process, but you can just check box “Automatically run configuration scripts” and give one of the two alternatives:

Use “root” user credential, for that you need to type root user password in this field

Use “sudo“, you will need to configure /etc/sudoers and add sudo permissions

I will not check or type any information now, just let unchecked and click: Next:

3.10 Now you can see that the installation will execute “Perform Prerequisite Checks“, imagine that after make all configuration this step complain about something…

But we are “lucky” and everything is fine.

3.11 It will be presented an installation summary and an option to “Save Response File” to install Oracle software in other environments using silent mode installation, click “Install“:

3.12 Finally the installation has started and in some moment we will be prompted to execute two script files as “root” user:

After complete the execution of these two scripts, click “OK” and the installation will proceed and finish:

PS:

Now we have an ASM instance up running and one new disk group: “DG_BLOG_DATA“.

4. I will show how to create the other two disk groups in two different ways:

Using sqlplus utility to create: DG_BLOG_REDO (/dev/sdc1)

PS:

Set ORACLE_SID and ORACLE_HOME environment variable:

export ORACLE_SID=+ASM

export ORACLE_HOME=/opt/oracle/product/12.1.0/grid

Connect to the ASM instance using sqlplus: sqlplus ‘/ as sysasm’

4.1 The another way to create disk group is using “Oracle ASM Configuration Assistant (ASMCA)“:

Execute: asmca: select “Create” –> Type “Disk Group Name” –> Rundancy “External” –> Select “Show Eligible” –> Check box “Disk Path” – “/dev/sdd1” –> OK.