I.2.7 Preparation for Building Cluster

When using GLS, take the following steps to set up the initial settings of GLS for the network used as the public LAN (also used as the administrative LAN). For more information on each setting, refer to "PRIMECLUSTER Global Link Services Configuration and Administration Guide: Redundant Line Control Function."

Note

If the initial settings are not correct, you may not access the system. Take the snapshot of the system disk before applying the settings.

Set up the following settings in each node building the cluster:

Setting up the system
1. In the /etc/hosts file, specify both IP addresses and host names to be used.
  Example
```
172.16.0.10	node1	# node1 IP address
172.16.0.11	node2	# node2 IP address
172.16.0.100	takeover	# Takeover IP address
172.16.0.1	gw	# Gateway IP address
```
2. In the /etc/sysconfig/network-scripts/ifcfg-eth0 file, comment out TYPE, set "static" to BOOTPROTO and "no" to PEERDNS. Add "HOTPLUG=no" and "DEVICETYPE=hanet".
  - /etc/sysconfig/network-scripts/ifcfg-eth0
```
DEVICE=eth0
#TYPE=Ethernet
BOOTPROTO=static
UUID=<Fixed value depending on environment (no change necessary)>
HOTPLUG=no
ONBOOT=yes
DEVICETYPE=hanet
PEERDNS=no
```
3. In the /etc/sysconfig/network file, set "sha0" to GATEWAYDEV.
```
GATEWAYDEV=sha0
```

Creating the virtual interface

# /opt/FJSVhanet/usr/sbin/hanetconfig create -n sha0 -m v -t eth0

Setting up the virtual interface
In the /etc/sysconfig/network-scripts/ifcfg-sha0 file, comment out IPADDR and NETMASK, and set "dhcp" to BOOTPROTO. Add the settings of "PEERDNS=yes", DNS1, and DNS2.
- /etc/sysconfig/network-scripts/ifcfg-sha0
```
DEVICE=sha0
#IPADDR=
#NETMASK=
BOOTPROTO=dhcp
ONBOOT=yes
DEVICETYPE=sha
HOTPLUG=no
PEERDNS=yes
DNS1=<IP address of master DNS server>
DNS2=<IP address of sub DNS server>
```
Note
Do not set SHAMACADDR in the ifcfg-sha0 file.
Setting up the network monitoring function
Set up the virtual router as a monitoring target. Considering the possibility that the virtual router has stopped for a long time, set up to avoid a failover of the cluster when an error occurs in the transmission route.
Example
```
# /opt/FJSVhanet/usr/sbin/hanetpathmon target -n sha0 -p 172.16.0.1
# /opt/FJSVhanet/usr/sbin/hanetpathmon param -n sha0 -f no
```

Setting up the subnet mask of the takeover virtual interface

Example

# /opt/FJSVhanet/usr/sbin/hanetmask create -i 172.16.0.0 -m 255.255.255.0

Creating the takeover virtual interface

Example

# /opt/FJSVhanet/usr/sbin/hanethvrsc create -n sha0 -i 172.16.0.100

Confirming the settings

Make sure that the settings done from Step 3 to Step 6 are enabled.

Example

# /opt/FJSVhanet/usr/sbin/hanetconfig print
 [IPv4,Patrol / Virtual NIC]

 Name       Hostname      Mode Physical ipaddr    Interface List
+-----------+---------------+----+-----------------+---------------------------+
 sha0                         v                    eth0

[IPv6]

 Name        Hostname/prefix                  Mode Interface List
+-----------+---------------------------------+-----+---------------------------+

# /opt/FJSVhanet/usr/sbin/hanetpathmon target
[Target List]
Name   VID  Target
+-------+----+----------------------------------------------------------+
sha0    -    172.16.0.1

# /opt/FJSVhanet/usr/sbin/hanetpathmon param
[Parameter List]
Name   Monitoring Parameter
+-------+----------------------------------------------------------+
sha0    auto_startup        =    yes
         interval           =      3 sec
         times              =      5 times
         repair_times       =      2 times
         idle               =     45 sec
         Auto fail-back     =     no
         FAILOVER Status    =     no

# /opt/FJSVhanet/usr/sbin/hanetmask print
network-address  netmask
+---------------+---------------+
172.16.0.0       255.255.255.0

# /opt/FJSVhanet/usr/sbin/hanethvrsc print
 ifname     takeover-ipv4     takeover-ipv6     vlan-id/logical ip address list
+----------+----------------+----------------+--------------------------------+
 sha0:65    172.16.0.100     -                -

Restarting the system
Execute the following command and restart the system.
```
# /sbin/shutdown -r now
```

To operate the cluster system in RHOSP environment, take the following steps to create the RHOSP environment information file.

On all the nodes, create the /opt/SMAW/SMAWRrms/etc/os_endpoint.cfg file as follows.

DOMAIN_NAME=RHOSPDomainName
PROJECT_NAME=RHOSPProjectName
IDENTITY=IdentityURL
COMPUTE=ComputeURL

RHOSPDomainName : Domain name in RHOSP
RHOSPProjectName: Project name in which the cluster is built in RHOSP
IdentityURL     : URL of the Identity service endpoint used in RHOSP
                  Note: Do not include any character strings from "/vX.X" in URL.
ComputeURL      : URL of the Compute service endpoint used in RHOSP
                  Note: Do not include any character strings from "/vX.X" in URL.

Example

DOMAIN_NAME=primecluster_domain
PROJECT_NAME=primecluster_project
IDENTITY=https://192.168.11.11:5000
COMPUTE=https://192.168.11.11:8774

Set the owner, group, and the access authority as follows.

# chown root:root /opt/SMAW/SMAWRrms/etc/os_endpoint.cfg
# chmod 600 /opt/SMAW/SMAWRrms/etc/os_endpoint.cfg

I.2.7 Preparation for Building Cluster

I.2.7.1 Initial GLS Setup

I.2.7.2 Creating RHOSP Environment Information File

I.2.7.3 Preparation Prior to Building Cluster