Quick Start for Physical CORD POD

This guide is meant to enable the user to utilize the artifacts of this repository to to deploy CORD on to a physical hardware rack. The artifacts in this repository will deploy CORD against a standard physical rack wired according to the best practices as defined in this document.

Physical configuration

Physical Hardware Connectivity

As depicted in the diagram above the base model for the CORD POD deployment contains:

  • 4 OF switches comprising the leaf - spine fabric utilized for data traffic
  • 4 compute nodes with with 2 40G ports and 2 1G ports
  • 1 top of rack (TOR) switch utilized for management communications

The best practices in terms of connecting the components of the CORD POD include:

  • Leaf nodes are connected to the spines nodes starting at the highest port number on the leaf.
  • For a given leaf node, its connection to the spine nodes terminate on the same port number on each spine.
  • Leaf n connections to spine nodes terminate at port n on each spine node.
  • Leaf spine switches are connected into the management TOR starting from the highest port number.
  • Compute nodes 40G interfaces are named eth0 and eth1.
  • Compute nodes 10G interfaces are named eth2 and eth3.
  • Compute node n is connected to the management TOR switch on port n, egressing from the compute node at eth2.
  • Compute node n is connected to its primary leaf, egressing at eth0 and terminating on the leaf at port n.
  • Compute node n is connected to its secondary leaf, egressing at eth1 and terminating on the leaf at port n.
  • eth3 on the head node is the uplink from the POD to the Internet.

The following assumptions are made about the phyical CORD POD being deployed:

  • The leaf - spine switchs are Accton 6712s
  • The compute nodes are using 40G Intel NIC cards
  • The compute node that is to be designated the head node has Ubuntu 14.04 LTS installed.

Bootstrapping the Head Node

The head node is the key to the physical deployment of a CORD POD. The automated deployment of the physical POD is designed such that the head node is manually deployed, with the aid of automation tools, such as Ansible and from this head node the rest of the POD deployment is automated.

The head node can be deployed either from a node outside the CORD POD or by deploying from the head the head node. The procedure in each scenario is slightly different because during the bootstrapping of the head node it is possible that the interfaces needed to be renamed and the system to be rebooted.

Bootstrapping the Head Node from Outside the POD (OtP)

To deploy the head node it is assumed that the node is reachable from outside the POD over its eth3 interface and that the machine from which you are bootstrapping the head node has Vagrant and git installed.

NOTE: This quick start walk through assumes that the head node is being deployed from the Vagrant machine that is defined within the repository. It is possible to deployment the head node from the cloned repository without using a Vagrant VM along as Ansible version > 2.0 is installed on the OtP host. When doing a deployment without the Vagrant VM, just invoke the given Ansible commands directly from on the OtP host.

Cloning the Repository

To clone the repository select a location on the outside the POD (OtP) host and issue the git command to download (clone) the repository.

$ git clone http://gerrit.opencord.org/maas

When this is complete, a listing (ls) of this directory should yield output similar to:

$ ls
QUICKSTART.md           automation/             doc/                    harvester/              roles/
QUICKSTART_PHYSICAL.md  bootstrap/              fabric.yml              head-node.yml           scripts/
README.md               build.gradle            gradle/                 head.yml
Vagrantfile             compute.yml             gradlew*                host_vars/
ansible/                dev-head-node.yml       gradlew.bat             hosts

Starting the Vagrant Machine

To start and connect to the the Vagrant machine, issue the following commands:

$ vagrant up maasdev
$ vagrant ssh maasdev

NOTE: It may have several minutes for the first command vagrant up maasdev to complete as it will include creating the VM as well as downloading and installing various software packages.

Once connected to the Vagrant machine, you can find the deployment artifacts in the /maasdev directory on the VM.

cd /maasdev

Invoke Bootstrapping

The head node will be bootstrapped using Ansible and the playbook head-node.yml. This playbook is small and defines the Ansible role for the head node:

- hosts: all
  serial: 1
  roles:
    - head-node

The head-node role depend on the Ansible compute-node role as well as others, the important point being that a head node is simply a compute node with some extra stuff installed.

To bootstrap the head node the following command can be issues:

$ ansible-playbook -i <ip-of-head-node>, --ask-pass --ask-sudo-pass \
  --user=<deployment-user-id> --extra-vars='fabric_ip=<fabric-ip> \
  management_ip=<management-ip> --exeternal_ip=<external-ip>' head-node.yml
Playbook Options

NOTE The comma (,) after the is important as it informs Ansible that the option is the IP of the head node and not an inventory file that contains the list of IPs for the Ansible managed nodes.

replace <ip-of-head-node> with the actually IP address to the host on which the head node is being deployed and <deployment-user-id> with a user ID on the host which can be used to ssh from the OtP host to the head node and has sudo rights on the head node.

During the bootstrapping of the nodes various network settings are modified. The extra-vars settings to the ansible-playbook command allow these settings to be specified. The values for the extra-vars can be one of the following:

  • dhcp - assumes that the address will be assigned via DHCP
  • manual - assumes that the address will be assigned manually
  • a.b.c.d/# - specifies the IP address of the interface as well as the number of bits in the netmask.

These values are used to configure the interfaces and will result in changes to the /etc/network/interface file on the head node.

If you do not wish for the deployment scripts to modify the network configuration of the head node you can substitute the ansible-playbook option --skip-tags=interface_config for the extra-vars options.

After you invoke the ansible-playbook you will be prompted for the ssh and sudo passwords for the the remote user. In most cases these are the same.

The ansible-playbook will take several minutes to complete as it does roughly the following:

  1. Download and install Docker and Docker Compose
  2. Rename and configured the network interfaces
  3. Reboot the system to apply the network changes
  4. Download boot images for the Accton switches
  5. Download and install Canonical's Metal as a Service (MAAS) software
  6. Configure MAAS
  7. Download and invoke Docker images to support automation of MAAS capabilities

Wait for Image Download

As part of the bootstrapping and configuration of MAAS, downloading of boot images for the other compute nodes is initiated. Before the other compute nodes can be booted the download of this image must be completed.

To verify the status of the image download you can visit the MAAS UI at http://<ip-of-head-node>/MAAS and select the Images tab. On this page the status of the download will be visible.

Complete

Once the download of boot image for the compute nodes is complete, the head node is boot strapped and you can proceed to the section Booting the Rest of the POD.

Bootstrapping the Head Node from the Head Node

In order to bootstrap the head node from the head node Ansible, version >= 2, must be installed on the head node. Additionally the following files / directories from the repository must be on head node:

  • roles (and all its sub-directories and files)
  • head-node.yml

Invoke on the Head Node (OtH) Bootstrapping

Once Ansible is installed on the head node and the proper files are availble, the head node can be bootstrapped. Because the bootstrapping capabilities modify the network interface configuration of the head node, when bootstrapping OtH must be done in two steps as there is a system reboot in the middle. The first step is provisioning the head node as as a compute node and the second is provisioning it with the head node capabilities.

Bootstrap Head Node as Compute Node

To complete the first phase of the bootstrapping the head node, provisioning as a compute node, the following command can be used:

$ ansible-playbook -i <ip-of-head-node>, --ask-pass --ask-sudo-pass \
  --user=<deployment-user-id> --extra-vars='fabric_ip=<fabric-ip> \
  management_ip=<management-ip> --exeternal_ip=<external-ip>' compute-node.yml

(see [Playbook Options](#Playbook Options) for a description of the parameters)

If you do not wish to have the Ansible playbook to modify the network configuration of the host you can add the --skip-tags=interface_config option to the ansible-playbook command line.

If you do not wish the system to auto reboot if the network configuration is modified you can add the --skip-tags=reboot option to the ansible-playbook command line.

NOTE: If the network parameters have changed the head node will likely need to be rebooted for those changes to take effect. If you would like to understand the changes before the reboot, you can specify the --skip-tags=reboot option and then run the follow diff command:

diff /etc/network/interfaces.1 /etc/network/interfaces

NOTE: Be sure to reboot the head node after network changes so that they will be applied.

Bootstrap Head Node as Head Node

Once the head node has been provisioned as a compute node the head node capabilities can be overlaid. This can be done using the following command line:

$ ansible-playbook -i <ip-of-head-node>, --ask-pass --ask-sudo-pass \
  --user=<deployment-user-id> --extra-vars='fabric_ip=<fabric-ip> \
  management_ip=<management-ip> --exeternal_ip=<external-ip>' --skip-tags=interface_config head-node.yml

(see [Playbook Options](#Playbook Options) for a description of the parameters)

The ansible-playbook will take several minutes to complete as it does roughly the following:

  1. Download and install Docker and Docker Compose
  2. Rename and configured the network interfaces
  3. Reboot the system to apply the network changes
  4. Download boot images for the Accton switches
  5. Download and install Canonical's Metal as a Service (MAAS) software
  6. Configure MAAS
  7. Download and invoke Docker images to support automation of MAAS capabilities

Wait for Image Download

As part of the bootstrapping and configuration of MAAS, downloading of boot images for the other compute nodes is initiated. Before the other compute nodes can be booted the download of this image must be completed.

To verify the status of the image download you can visit the MAAS UI at http://<ip-of-head-node>/MAAS and select the Images tab. On this page the status of the download will be visible.

Complete

Once the download of boot image for the compute nodes is complete, the head node is boot strapped and you can proceed to the section Booting the Rest of the POD.