INF StarlingX O-Cloud Getting Started /Sample test process

Owned by Xiaohua Zhang
Last updated: Jul 31, 2024 by Jackie Huang
5 min read

This doc describes how to install and deploy INF bare metal All-in-one Duplex deployment configuration on two HP ProLiant DL380p Gen8 servers.

1 Installation for the first server from the O-RAN INF ISO image

 1.1 Burn the image to the USB device 

  • Assume the the usb device is /dev/sdX here 
  •  $ sudo dd if=/path/to/inf-image-aio-installer-intel-corei7-64.iso of=/dev/sdX bs=1M  

1.2 Install the first server (controller-0) 

  • Reboot the target from the USB device. 

  • Select “All-in-one Graphics console” or “All-in-one Serial console install” and press ENTER 

 

  • Start the auto installation 

 


  •  It will reboot automatically after installation

 

2 Configuration and initialize the bootstrap

2.1 First Login with "sysadmin/sysadmin" and change password  

 

2.2 Set OAM network before bootstrap 

 export OAM_DEV=eno3 

export CONTROLLER0_OAM_CIDR=128.224.210.110/24 

export DEFAULT_OAM_GATEWAY=128.224.210.1 

sudo ip address add $CONTROLLER0_OAM_CIDR dev $OAM_DEV 

sudo ip link set up dev $OAM_DEV 

sudo ip route add default via $DEFAULT_OAM_GATEWAY dev $OAM_DEV 

2.3 Login the server through SSH with "sysadmin"

2.4 Prepare the localhost.yml for bootstrap 

cat << EOF > localhost.yml 

system_mode: duplex 

management_subnet: 192.168.18.0/24 

management_start_address: 192.168.18.2 

management_end_address: 192.168.18.50 

management_gateway_address: 192.168.18.1 

external_oam_subnet: 128.224.210.0/24 

external_oam_gateway_address: 128.224.210.1 

external_oam_floating_address: 128.224.210.110 

external_oam_node_0_address: 128.224.210.111 

external_oam_node_1_address: 128.224.210.112 

EOF 

2.5 Run the ansible bootstrap 

ansible-playbook /usr/share/ansible/stx-ansible/playbooks/bootstrap.yml -vvv 

After the bootstrap successfully finish, it will show as following: 

2.6 Congiure controller-0 

  • Acquire admin credentials: 

source /etc/platform/openrc 

  • Configure the OAM and MGMT interfaces of controller-0 and specify the attached networks: 

OAM_IF=eno3 

MGMT_IF=eno1 

system host-if-modify controller-0 lo -c none 

IFNET_UUIDS=$(system interface-network-list controller-0 | awk '{if ($6=="lo") print $4;}') 

for UUID in $IFNET_UUIDS; do 

    system interface-network-remove ${UUID} 

done 

system host-if-modify controller-0 $OAM_IF -n oam0 

system host-if-modify controller-0 $MGMT_IF -n pxeboot0 

 

system host-if-modify controller-0 oam0 -c platform 

system interface-network-assign controller-0 oam0 oam 

 

system host-if-modify controller-0 pxeboot0 -c platform 

system interface-network-assign controller-0  pxeboot0 pxeboot 

 

system host-if-add         -V 18 controller-0 mgmt0 vlan pxeboot0 

system interface-network-assign controller-0 mgmt0 mgmt 

 system host-if-add         -V 19 controller-0 cluster0 vlan pxeboot0 

system interface-network-assign controller-0 cluster0 cluster-host 

  • Configure NTP Servers for network time synchronization: 

 system ntp-modify ntpservers=0.pool.ntp.org,1.pool.ntp.org 

Then check the inferfaces status

 

  • Add an OSD on controller-0 for Ceph: 

system host-disk-list controller-0 

system host-disk-list controller-0 | awk '/\/dev\/sdb/{print $2}' | xargs -i system host-stor-add controller-0 {} 

system host-disk-list controller-0 | awk '/\/dev\/sdc/{print $2}' | xargs -i system host-stor-add controller-0 {} 

system host-stor-list controller-0 

2.7 Unlock controller-0 

 system host-unlock controller-0 

Controller-0 will reboot to apply configuration changes and come into service. This can take 5-10 minutes, depending on the performance of the host machine. 

Once the controller comes back up, check the status of controller-0 

source /etc/platform/openrc 

system host-list 

 

2 Installation for the second server (controller-1)

2.1 Power on the controller-1 server and force it to network boot  

 

 2.2 As controller-1 boots, a message appears on its console instructing you to configure the personality of the node 

 

 2.3 On the console of controller-0, list hosts to see newly discovered controller-1 host (hostname=None) 

 system host-list 

 

 2.4 Using the host id, set the personality of this host to 'controller’: 

 system host-update 2 personality=controller 

 

 2.5 Wait for the software installation on controller-1 to complete, for controller-1 to reboot, and for controller-1 to show as locked/disabled/online in 'system host-list'.  

 This can take 5-10 minutes, depending on the performance of the host machine. 

 

2.6 Configure controller-1 

OAM_IF=eno3 

MGMT_IF=eno1 

system host-if-modify controller-1 $OAM_IF -n oam0 

system host-if-modify controller-1 oam0 -c platform 

system interface-network-assign controller-1 oam0 oam 

 system host-if-add         -V 19 controller-1 cluster0 vlan pxeboot0 

system interface-network-assign controller-1 cluster0 cluster-host 

 system host-if-list controller-1 

 system host-disk-list controller-1 

system host-disk-list controller-1 | awk '/\/dev\/sdb/{print $2}' | xargs -i system host-stor-add controller-1 {} 

system host-disk-list controller-1 | awk '/\/dev\/sdc/{print $2}' | xargs -i system host-stor-add controller-1 {} 

system host-stor-list controller-1 

2.7 Unlock controller-1 

system host-unlock controller-1 

 system host-list 

 3 High Availability status show

sm-dump

Use the sm-dump to show the status of 2 controller, it shows that services on controller-0 are in active mode, and services on controller-1 are in standby mode.