Slik implementerer du Oracle 12c Release 2 Grid og RAC-database på RHEL 7.x

       dell-redhat-rdbms-utilities-2018.06-1.el7.noarch.rpm is designed to do the following

            yum localinstall –y dell-redhat-rdbms-utilities-2018.06-1.el7.noarch.rpm

2.2. Setting up the Network

2.2.1. Public Network

 To configure the public network on each node

• Log in as root.
• Navigate to /etc/sysconfig/network-scripts and edit the ifcfg-em# file where # is the number of the network device

          NAME="Oracle Public"

DEVICE= "em1"
ONBOOT=yes
TYPE= Ethernet
BOOTPROTO=static
IPADDR=<routable IP address>
NETMASK=<netmask>
GATEWAY=<gateway_IP_address>

• Set the hostname via below command

         hostnamectl set-hostname <host-name>
          where <host-name> is the hostname that we are using for installation

• Type service network restart to restart the network service
• Type ifconfig to verify that the IP addresses are set correctly
• To check your network configuration, ping each public IP address from a client on the LAN that is not a part of the cluster
• Connect to each node to verify that the public network is functioning. Type ssh to verify that the secure shell (ssh) command is working 

2.2.2. Private Network

 The example below provides step-by-step instructions on enabling redundant interconnect using HAIP on a fresh Oracle 12c Grid Infrastructure installation

• Edit the file /etc/sysconfig/network-scripts/ifcfg-emX, where X is the number of the em device, ifcfg-emX configuration files of the network adapters to be used for your private interconnect.

                  DEVICE=em2
        BOOTPROTO=static
        HWADDR=
        ONBOOT=yes
        NM_CONTROLLED=yes
        IPADDR=192.168.1.140
        NETMASK=255.255.255.0
        DEVICE=em3

      HWADDR=
   BOOTPROTO=static
   ONBOOT=yes
   NM_CONTROLLED=yes
   IPADDR=192.168.1.141
   NETMASK=255.255.255.0

• Once you have saved both the configuration files, restart your network service using below commands

          nmcli connection reload

          nmcli device disconnect em2

          nmcli connection up em2

• Repeat the steps for each interface that has been modified.             
• The completion of the steps above have now prepared your system to enable HAIP using the Oracle Grid Infrastructure installer. When you have completed all the Oracle prerequisites and are ready to install Oracle, you will need to select em2 and em3 as 'private' interfaces at the 'Network Interface Usage' screen
• This step enables redundant interconnectivity once your Oracle Grid Infrastructure has successfully completed and is running

2.2.3. Oracle Flex ASM Network

 Oracle Flex ASM can use either the same private networks as Oracle Clusterware, or use its own dedicated private networks. Each network can be classified PUBLIC or PRIVATE+ASM or PRIVATE or ASM

2.2.4. IP Address and Name Resolution Requirements

We can configure IP address of the cluster nodes with one of the following option

• Domain Name Server (DNS)

2.2.4.1. Domain Name Server (DNS)

To set up an Oracle 12c RAC using Oracle (without GNS):

 The table below describes the different interfaces, IP address settings and the resolutions in a cluster      

Interface	Type	Resolution
Public	Static	DNS
Private	Static	Not required
ASM	Static	Not required
Node Virtual IP	Static	Not required
SCAN virtual IP	Static	Not required

 

Configuring a DNS Server

To configure changes on a DNS server for an Oracle 12c cluster using DNS (without GNS):

Configure SCAN NAME resolution on DNS server. A SCAN NAME configured on the DNS server using the Round Robin policy should resolve to three public IP addresses (recommended), however the minimum requirement is one public IP address

For example

scancluster IN A 192.0.2.1
IN A 192.0.2.2
IN A 192.0.2.3

Where scancluster is the SCAN NAME provided during Oracle Grid installation

Configuring a DNS Client

To configure the changes required on the cluster nodes for name resolution:

• You must configure the resolv.conf on the nodes in the cluster to contain name server entries that are resolvable to DNS server.

               nmcli connection modify ipv4.dns ipv4.dns-search

• Verify the order configuration /etc/nsswitch.conf controls the name service order. In some configurations, the NIS can cause issues with Oracle SCAN address resolution. It is recommended that you place the NIS entry at the end of the search list.

               For example, hosts: dns files nis

3. Preparing Shared Storage for Oracle RAC Installation

Oracle RAC requires shared LUNs for storing your Oracle Cluster Registry (OCR), voting disks, Oracle Database files, and Flash Recovery Area (FRA). The following table shows the typical recommended storage volume design for Oracle 12c Database
 

Database Volume Type/Purpose	Number of Volumes	Volume Size
OCR/VOTE	3	50GB each
DATA	4	250GB1 each
REDO2	2	At least 50GB each
FRA	1	100GB3
TEMP	1	100GB

1 – Adjust each volume size based on your database; 2 – At least two REDO ASM diskgroups are recommended, each with at least one storage volume; 3 – Ideally, the size should be 1.5x the size of the database if storage usable capacity permits 

 3.1. Setting up Device Mapper Multipath for XtremIO storage

 The purpose of Device Mapper Multipath is to enable multiple I/O paths to improve performance and provide consistent naming. Multipathing accomplishes this by combining your I/O paths into one device-mapper path and properly load balancing the I/O. This section will provide the best practices on how to setup your device mapper multipathing within your Dell PowerEdge server. Verify that your device-mapper and multipath driver are at least the version shown below or higher:

• rpm -qa | grep device-mapper-multipath

  device-mapper-multipath

• Enable multipath by  mpathconf –enable
• Configure XtremIO multipath by modifying /etc/multipath.conf with the following

 device {

            vendor                         XtremIO

            product                        XtremApp

            path_grouping_policy multibus

            path_checker              tur

            path_selector              "queue-length 0"

            rr_min_io_rq               1

           user_friendly_names  yes

           fast_io_fail_tmo         15

           failback                       immediate

}

• Add appropriate user friendly names to each volume with the corresponding scsi_id. We can get scsi_ids with the below command

     /usr/lib/udev/scsi_id -g -u -d /dev/sdX

• Locate the multipath section within your /etc/multipath.conf file. In this section you will provide the scsi_id of each volume and provide an alias in order to keep a consistent naming convention across all of your nodes. An example is shown below     

    multipaths {

     multipath {

wwid               <out put of step4 for volume1>

alias                 alias_of_volume1

}

multipath {

wwid         <out put of step4 for volume2>

alias alias_of_volume2
                 }

}

• Restart your multipath daemon service using

     Service multipathd restart

• Verify that your multipath volumes alias are displayed properly

    multipath -ll

• Repeat all trhe steps for all nodes

3.2. Partitioning the Shared Disk

     This section describes how to use parted utility to create a single partition on a volume/virtual disk that spans the entire disk.

     Partition each database volume that was setup using device-mapper by running the following command:

           $> parted -s /dev/mapper/<volume1> mklabel msdos
     $> parted -s /dev/mapper/<volume1> primary 2048s 100%

Repeat this for all the required volumes

• Repeat above steps for all volumes and restart multipathd on all other nodes

     systemctl restart multipathd.service

• Reboot the system if your newly created partition is not displayed properly