1. OPNFV config guide instructions for the Brahmaputra release of OPNFV when using installers to deploy onosfw

1.1. Abstract

This document describes how to config the Brahmaputra release of OPNFV when using installers as a deployment tool to deploy onosfw, covering it’s limitations, dependencies and required system resources.

1.2. License

Brahmaputra release of OPNFV when using installers to deploy onosfw Docs (c) by Henry (HUAWEI)

Brahmaputra release of OPNFV when using installers to deploy onosfw Docs are licensed under a Creative Commons Attribution 4.0 International License. You should have received a copy of the license along with this. If not, see <http://creativecommons.org/licenses/by/4.0/>.

1.3. Version history

Date Ver. Author Comment
2016-01-21 1.0.0 Henry (HUAWEI) Rewritten for ONOSFW B release
2016-01-20 0.0.2 Henry (HUAWEI) Minor changes & formatting
2016-01-19 0.0.1 Henry (HUAWEI) First draft

1.4. Introduction

ONOSFW need to deploy with several installers and each installer have differernt configs. Here is the scenarios ONOSFW need to supported

Apex Compass Fuel JOID
Virtual CentOS 7, ONOS with OpenStack neutron-l3-agent being disabled Virtual Ubuntu 14, ONOS with OpenStack neutron-l3-agent being disabled Virtual Ubuntu 14, ONOS with OpenStack neutron-l3-agent being disabled Virtual Ubuntu 14, ONOS with OpenStack neutron-l3-agent being disabled
NA BM Ubuntu 14, ONOS with OpenStack neutron-l3-agent being disabled NA NA

Below is the detail config for them:

1.5. Config for Installers

1.5.1. Config Documentation for onos with apex

  1. Pyhsical Requirement

    1.1 CentOS 7 (from ISO or self-installed).

    1.2 Root access.

    1.3 libvirt virtualization support.

    1.4 minimum 2 networks and maximum 6 networks, multiple NIC and/or VLAN combinations are supported. This is virtualized for a VM deployment.

    1.5 The Bramaputra Apex RPM.

    1.6 16 GB of RAM for a bare metal deployment, 56 GB of RAM for a VM deployment.

  2. How to add onos into apex

    2.1 Apex will download two images instack.qcow2 and overcloud-full.qcow2 when build apex rpm. The instack.qcow2 is used for installation of instack virtual machine. The overcloud-full.qcow2 is used for installation of openstack nodes. The opnfv-tripleo-heat-templates.patch will update tripleo-heat scripts in instack.qcow2. And it will call puppet deployment scripts in overcloud-full.qcow2 to finish deployment. Those two files will be patch up and store into the two images during apex rpm building process in instack.sh.

    below is the directory:

    ├── build
│   ├── instack.sh   # add onos build steps
│   ├── opnfv-tripleo-heat-templates.patch   # add onos deployment scripts
│
├── ci
│   └── deploy.sh   #add onos build steps inside

    2.2 Upload puppet-onos to github for apex iso/rpm building.

  3. Virtual deployment

    3.1 Install jumphost.

    3.2 Edit /etc/opnfv-apex/deploy_settings.yaml and change opendaylight into onos.

    3.3 Execute sudo opnfv-deploy –virtual [ –no-ha ] -d /etc/opnfv-apex/deploy_settings.yaml

  4. Baremetal deployment

    4.1 Install jumphost.

    4.2 Edit /etc/opnfv-apex/deploy_settings.yaml and change opendaylight into onos.

    4.3 Edit /etc/apex-opnfv/inventory.yaml and change mac_address, ipmi_ip, ipmi_user, ipmi_password etc base on your physical server and network.

    4.4 Execute sudo opnfv-deploy -d /etc/opnfv-apex/deploy_settings.yaml -i /etc/apex-opnfv/inventory.yaml

  5. Detail of apex installation Apex Installation.

1.5.2. Config Documentation for onos with Compass

  1. Pyhsical Requirement for install ONOS

    1.1 Ubuntu Server 14.04 LTS 64-bit (from ISO or self-installed).

    1.2 minimum 2GB RAM.

    1.3 minimum 2 processors.

    1.4 At least 5GB disk space.

    1.5 The ONOS version 1.4.

  2. How to add onos into compass

    2.1 the script that install onos service is added into the compass4nfv project. and the onos will be started when compass called the onos script. the script is included in the directory of compass4nfv project below:

    commpass4nfv
   ├── deploy
   │   ├── adapters
   │      ├── ansible
   │         ├── roles # include the sdn script
   │             ├── onos_cluster # include the onos script
   │                 ├── files # include the files of jdk and onos driver
   │                 ├── handlers # include the opertaion of restart onos service
   │                 ├── tasks # include the task of installing onos
   │                 ├── templates # include the templates of onos
   │                 ├── vars # include the var of onos used

  3. Virtual deployment

    3.1 Install jumphost

    3.2 Build iso of compass. Execute ./build.sh

    3.3 Execute ./deploy.sh virtual_cluster_onos

  4. Baremetal deployment

    4.1 Install jumphost

    4.2 Build iso of compass. Execute ./build.sh

    4.3 Config the envionment variables

    export WORKSPACE=”/home/jenkins/jenkins-slave/workspace/compass-deploy-bare-huawei-us-master”

    export BUILD_DIRECTORY=$WORKSPACE/build_output

    export CONFDIR=$WORKSPACE/deploy/conf/hardware_environment/huawei_us_lab/pod1

    export ISO_URL=file://$BUILD_DIRECTORY/compass.iso

    export EXTERNAL_NIC=eth0

    export INSTALL_NIC=eth1

    export OS_VERSION=trusty

    export OPENSTACK_VERSION=liberty

    4.4 Execute cd $WORKSPACE

    4.5 Execute ./deploy.sh –dha $CONFDIR/dha.yml –network $CONFDIR/network.yml

  5. Detail of compass installation `Compass Installation`_.

1.5.3. Config Documentation for onos with Fuel

  1. Pyhsical Requirement

    1.1 Linux , Microsoft or Mac OS.

    1.2 Root access or admin access.

    1.3 libvirt virtualization support.

    1.4 minimum 2 networks and maximum 4 networks, multiple NIC and/or VLAN combinations are supported.

    1.5 250G disk at least for no-ha virtual deployment

  2. How to add onos into Fuel

    2.1 Fuel provides an intuitive, GUI-driven experience for deployment and management of OpenStack, related community projects and plug-ins. Onos supplies plug-in to manage network of L2/L3.

    below is the directory:

    ├── build
│   ├──f_isoroot
│     ├── f_onosfwpluginbuild   # add onos build url
│
├── deploy
│   ├──scenario
│     ├── ha-onos_scenario.yaml   # add onos ha configuration
│       ├── noha-onos_scenario.yaml   # add onos noha configuration
├── ci
│   └── deploy.sh   #add onos scenarion steps inside

    2.2 Upload fuel-plugin-onos to git for fuel iso/rpm building.

  3. Automatic deployment

    3.1 Install jumphost.

    3.2 git clone https://gerrit.opnfv.org/gerrit/fuel

    3.3 In fuel/ci, exec ./deploy.sh. For virtual deployment, you can use -b file:///fuel/deploy/config -l devel-popeline -p huawei-ch -s no-ha_onos_heat_ceilmeter_scenario_0.0.0.2.yaml -i file://root/iso/fuel.iso. Fore bare metal deployment, change parameters correspondingly and use no-ha_onos scenario_0.0.1.yaml.

  4. Build onos plugin into rpm independently.

    4.1 Install fuel plugin builder( detailed steps can be found in https://wiki.openstack.org/wiki/Fuel/Plugin ).

    4.2 git clone git://git.openstack.org/openstack/fuel-plugin-onos. For Kilo deployment, use –b Kilo.

    4.3 fpb –build fuel-plugin-onos

    4.4 Move onos*.rpm in to master and fuel plugins –install onos*.rpm.

    4.5 Create a new environment and select onos plugin in settings table. If the Public Ethernet is not eth3, please change it with your actual configuration.

    4.6 Select a node with the role of controller and onos( onos must collocate with a controller).

    4.7 Deploy changes.

  5. Related url for fuel and onos.

    Fuel: https://wiki.openstack.org/wiki/Fuel

    Fuel plugin: https://wiki.openstack.org/wiki/Fuel/Plugins

    Fuel codes: https://gerrit.opnfv.org/gerrit/fuel

    Fuel iso: http://build.opnfv.org/artifacts/

    Fuel-plugin-onos: http://git.openstack.org/cgit/openstack/fuel-plugin-onos/

1.5.4. Config Documentation for onos with JOID

1、Virtual Machine Deployment

1.1、 Hardware Requirement:

OS: Ubuntu Trusty 14.04 LTS

Memory: 48 GB +

CPU: 24 cores +

Hard disk: 1T

1.2、Get the joid code from gerrit https://gerrit.opnfv.org/gerrit/p/joid.git

1.3、Suggest to create a user ubuntu and use this user, if not,you should edit the file：joid/ci/maas/default/deployment.yaml. Find Virsh power settings and change ubuntu to your own user name.

1.4、Deploy Maas
$ cd joid/ci/ $ ./02-maasdeploy.sh
1.5、Deploy OPNFV:
For liberty openstack, ONOS SDN, HA mode

$ ./deploy.sh -o liberty -s onos -t ha

2、Bare Metal Deployment

2.1、Pre Requisite:

  1. have a single node install with Ubuntu OS 14.04 LTS
  2. Minimum four nodes exist and should have been preconfigured and integrated with JOID please have look into this wiki page https://wiki.opnfv.org/joid/get_started

2.2、Get the joid code from gerrit : https://gerrit.opnfv.org/gerrit/p/joid.git

2.3、Suggest to create a user ubuntu and use this user, if not,you should edit the file：joid/ci/maas/default/deployment.yaml.
Find Virsh power settings and change ubuntu to your own user name.

2.4、Deploy MAAS:

$ ./02-maasdeploy.sh <lab and pod name i.e. intelpod5>

2.5、Deploy OPNFV:

For liberty openstack, ONOS SDN, HA mode in intel pod5 $ ./deploy.sh -o liberty -s onos -t ha -l intelpod5

3、How to add onos into joid

create a dir onos as below:

--onos
├── 01-deploybundle.sh  # deploy bundle define
├── juju-deployer
│   ├── ovs-onos-ha.yaml # openstack type ha feature define
│   ├── ovs-onos-nonha.yaml # openstack type nosha feature define
│   ├── ovs-onos-tip.yaml # openstack type tip feature define
├── juju_test_prepare.sh  # create ext-net and update gw_mac
└── README  # description

Revision:

Author:Henry(jiangrui1@huawei.com)

Build date: February 14, 2016