CORD Quick Start Guide using Physical Nodes

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.

NOTE: If you are new to CORD, you should start by bringing up a development POD on a single physical server to get familiar with the CORD deployment process. Instructions to do so are in quickstart.md.

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 fabric interfaces (typically 40G or 10G) are named eth0 and eth1.
  • Compute nodes POD management interfaces (typically 1G) 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 switches 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. In addition, the user should have password-less sudo permission.

Prerequisite: Vagrant is installed and operational. Note: This quick start guide has only been tested against Vagrant and VirtualBox, specially on MacOS.

Bootstrap 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 is deployed from a host outside the CORD POD (OtP).

Install Repo

Make sure you have a bin directory in your home directory and that it is included in your path:

mkdir ~/bin
PATH=~/bin:$PATH

(of course you can put repo wherever you want)

Download the Repo tool and ensure that it is executable:

curl https://storage.googleapis.com/git-repo-downloads/repo > ~/bin/repo
chmod a+x ~/bin/repo

Clone the Repository

To clone the repository, on your OtP build host issue the git command:

mkdir opencord && cd opencord
repo init -u https://gerrit.opencord.org/manifest -b master

NOTE: In the example above the OpenCORD version clones was the master branch of the source tree. If a different version is desired then master should be replaced with the name of the desired version, cord-2.0 for example.

Fetch the opencord source code

repo sync

Complete

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

ls -F
build/         incubator/     onos-apps/     orchestration/ test/

Create the Development Machine

The development environment is required for the tasks in this repository. This environment leverages Vagrant to install the tools required to build and deploy the CORD software.

To create the development machine the following Vagrant command can be used. This will create an Ubuntu 14.04 LTS based virtual machine and install some basic required packages, such as Docker, Docker Compose, and Oracle Java 8.

cd build
vagrant up corddev

NOTE: The VM will consume 2G RAM and about 12G disk space. Make sure it can obtain sufficient resources. It may takes several minutes for the first command vagrant up corddev to complete as it will include creating the VM as well as downloading and installing various software packages.

Complete

Once the Vagrant VM is created and provisioned, you will see output ending with:

==> corddev: PLAY RECAP *********************************************************************
==> corddev: localhost                  : ok=29   changed=25   unreachable=0    failed=0

The important thing is that the unreachable and failed counts are both zero.

Connect to the Development Machine

To connect to the development machine the following vagrant command can be used.

vagrant ssh corddev

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

cd /cord/build

Gradle

Gradle is the build tool that is used to help orchestrate the build and deployment of a POD. A launch script is included in the Vagrant machine that will automatically download and install gradle. The script is called gradlew and the download / install will be invoked on the first use of this script; thus the first use may take a little longer than subsequent invocations and requires a connection to the internet.

Complete

Once you have created and connected to the development environment this task is complete. The cord repository files can be found on the development machine under /cord. This directory is mounted from the host machine so changes made to files in this directory will be reflected on the host machine and vice-versa.

Fetch

The fetching phase of the deployment pulls Docker images from the public repository down to the local machine as well as clones any git submodules that are part of the project. This phase can be initiated with the following command:

./gradlew fetch

NOTE: The first time you run ./gradlew it will download the gradle binary from the Internet and install it locally. This is a one time operation, but may be time consuming depending on the speed of your Internet connection.

Complete

Once the fetch command has successfully been run, this step is complete. After this command completes you should be able to see the Docker images that were downloaded using the docker images command on the development machine:

docker images
REPOSITORY                  TAG                 IMAGE ID            CREATED             SIZE
opencord/onos               <none>              e1ade494f06e        3 days ago          936.5 MB
python                      2.7-alpine          c80455665c57        2 weeks ago         71.46 MB
xosproject/xos-base         <none>              2b791db4def0        4 weeks ago         756.4 MB
redis                       <none>              74b99a81add5        11 weeks ago        182.8 MB
xosproject/xos-postgres     <none>              95312a611414        11 weeks ago        393.8 MB
xosproject/cord-app-build   <none>              003a1c20e34a        5 months ago        1.108 GB
consul                      <none>              62f109a3299c        6 months ago        41.05 MB
swarm                       <none>              47dc182ea74b        8 months ago        19.32 MB
nginx                       <none>              3c69047c6034        8 months ago        182.7 MB
xosproject/vsg              <none>              dd026689aff3        9 months ago        336 MB

Build Images

Bare metal provisioning leverages utilities built and packaged as Docker container images. These utilities are:

  • cord-maas-bootstrap - (directory: bootstrap) run at MAAS installation time to customize the MAAS instance via REST interfaces
  • cord-maas-automation - (directory: automation) daemon on the head node to automate PXE booted servers through the MAAS bare metal deployment work flow
  • cord-maas-switchq - (directory: switchq) daemon on the head node that watches for new switches being added to the POD and triggers provisioning when a switch is identified (via the OUI on MAC address).
  • cord-maas-provisioner - (directory: provisioner) daemon on the head node to managing the execution of ansible playbooks against switches and compute nodes as they are added to the POD.
  • cord-ip-allocator - (directory: ip-allocator) daemon on the head node used to allocate IP address for the fabric interfaces.
  • cord-dhcp-harvester - (directory: harvester) run on the head node to facilitate CORD / DHCP / DNS integration so that all hosts can be resolved via DNS
  • opencord/mavenrepo - custom CORD maven repository image to support ONOS application loading from a local repository
  • cord-test/nose - container from which cord tester test cases originate and validate traffic through the CORD infrastructure
  • cord-test/quagga - BGP virtual router to support uplink from CORD fabric network to Internet
  • cord-test/radius - Radius server to support cord-tester capability
  • opencord/onos - custom version of ONOS for use within the CORD platform #XOSTBD

The images can be built by using the following command.

./gradlew buildImages

NOTE: The first time you run ./gradlew it will download the gradle binary from the Internet and install it locally. This is a one time operation, but may be time consuming depending on the speed of your Internet connection.

Complete

Once the buildImages command successfully runs this task is complete. The CORD artifacts have been built and the Docker images can be viewed by using the docker images command on the development machine.

docker images --format 'table {{.Repository}}\t{{.Tag}}\t{{.Size}}\t{{.ID}}'
REPOSITORY                  TAG                 SIZE                IMAGE ID
opencord/mavenrepo          latest              338.2 MB            2e29009df740
cord-maas-switchq           latest              337.7 MB            73b084b48796
cord-provisioner            latest              822.4 MB            bd26a7001dd8
cord-dhcp-harvester         latest              346.8 MB            d3cfa30cf38c
config-generator            latest              278.4 MB            e58059b1afb2
cord-maas-bootstrap         latest              359.4 MB            c70c437c6039
cord-maas-automation        latest              371.8 MB            9757ac34e7f6
cord-ip-allocator           latest              276.5 MB            0f399f8389aa
opencord/onos               <none>              936.5 MB            e1ade494f06e
python                      2.7-alpine          71.46 MB            c80455665c57
golang                      alpine              240.5 MB            00371bbb49d5
golang                      1.6-alpine          283 MB              1ea38172de32
nginx                       latest              181.6 MB            01f818af747d
xosproject/xos-base         <none>              756.4 MB            2b791db4def0
ubuntu                      14.04               187.9 MB            3f755ca42730
redis                       <none>              182.8 MB            74b99a81add5
xosproject/xos-postgres     <none>              393.8 MB            95312a611414
xosproject/cord-app-build   <none>              1.108 GB            003a1c20e34a
consul                      <none>              41.05 MB            62f109a3299c
swarm                       <none>              19.32 MB            47dc182ea74b
nginx                       <none>              182.7 MB            3c69047c6034
xosproject/vsg              <none>              336 MB              dd026689aff3

NOTE: Not all the above Docker images were built by the buildImages command. Some of them, list golang, are used as a base for other Docker images; and some, like abh1nav/dockerui were downloaded when the development machine was created with vagrant up.

Deployment Configuration File

The commands to deploy the POD can be customized via a deployment configuration file. The file is in YAML format.

To construct a configuration file for your physical POD, copy the sample deployment configuration found in config/sample.yml and modify the values to fit your physical deployment. Descriptions of the values can be found in the sample file.

Publish

Publishing consists of pushing the build docker images to the Docker repository on the target head node. This step can take a while as it has to transfer all the image from the development machine to the target head node. This step is started with the following command:

./gradlew -PdeployConfig=config/podX.yml publish

Complete

Once the publish command successfully runs this task is complete. When this step is complete a Docker registry and Docker registry mirror. It can be verified that these are running by using the docker ps command.

docker ps --format 'table {{.ID}}\t{{.Image}}\t{{.Command}}\t{{.CreatedAt}}'
CONTAINER ID        IMAGE               COMMAND                  CREATED AT
c8dd48fc9d18        registry:2.4.0      "/bin/registry serve "   2016-12-02 11:49:12 -0800 PST
e983d2e43760        registry:2.4.0      "/bin/registry serve "   2016-12-02 11:49:12 -0800 PST

We can also query the docker registry on the head node. We should be able to observe a list of docker images.

Note: the example below uses the commands curl and jq to retrieve data and pretty print JSON. If you system doesn't have curl or jq installed it can be installed using sudo apt-get install -y curl jq.

curl -sS http://head-node-ip-address:5000/v2/_catalog | jq .
{
  "repositories": [
    "config-generator",
    "consul",
    "cord-dhcp-harvester",
    "cord-ip-allocator",
    "cord-maas-automation",
    "cord-maas-bootstrap",
    "cord-maas-switchq",
    "cord-provisioner",
    "mavenrepo",
    "nginx",
    "opencord/onos",
    "redis",
    "swarm",
    "xosproject/cord-app-build",
    "xosproject/vsg",
    "xosproject/xos-base",
    "xosproject/xos-postgres"
  ]
}

Deploy Bare Metal Provisioning Capabilities

There are three parts to deploying bare metal: deploying the head node PXE server (MAAS), PXE booting a compute node, and post deployment provisioning of the compute node. These tasks are accomplished utilizing additionally Vagrant machines as well as executing gradle tasks in the Vagrant development machine. This task also provisions XOS. XOS provides service provisioning and orchestration for the CORD POD.

Deploy MAAS and XOS

Canonical MAAS provides the PXE and other bare metal provisioning services for CORD and will be deployed on the head node.

./gradlew -PdeployConfig=config/podX.yml deploy

This task can take some time so be patient. It should complete without errors, so if an error is encountered, something has gone Horribly Wrong(tm). See the Getting Help section.

Complete

This step is complete when the command successfully runs. The Web UI for MAAS can be viewed by browsing to the target machine using a URL of the form http://head-node-ip-address/MAAS. To login to the web page, use cord for username. If you have set a password in the deployment configuration password use that, else the password used can be found in your build directory under <base>/build/maas/passwords/maas_user.txt.

After the deployBase command install MAAS, it initiates the download of an Ubuntu 14.04 boot image that will be used to boot the other POD servers. This download can take some time and the process cannot continue until the download is complete. The status of the download can be verified through the UI by visiting the URL http://head-node-ip-address/MAAS/images/, or via the command line from head node via the following commands:

APIKEY=$(sudo maas-region-admin apikey --user=cord)
maas login cord http://localhost/MAAS/api/1.0 "$APIKEY"
maas cord boot-resources read | jq 'map(select(.type != "Synced"))'

It the output of of the above commands is not a empty list, [], then the images have not yet been completely downloaded. depending on your network speed this could take several minutes. Please wait and then attempt the last command again until the returned list is empty, []. When the list is empty you can proceed.

Browse around the UI and get familiar with MAAS via documentation at http://maas.io

The deployment of XOS includes a deployment of Open Stack.

Booting Compute Nodes

Network configuration

The CORD POD uses two core network interfaces, fabric and mgmtbr. The fabric interface will be used to bond all interfaces meant to be used for CORD data traffic and the mgmtbr will be used to bridge all interfaces used for POD management (signaling) traffic.

An additional interface of import on the head node is the external interface, or the interface through which the management net accesses upstream servers; such as the Internet.

How physical interfaces are identified and mapped to either the fabric or mgmtbr interface is a combination of their name, NIC driver, and/or bus type.

By default any interface that has a module or kernel driver of tun, bridge, bonding, or veth will be ignored when selecting devices for the fabric and mgmtbr interfaces. As will any interface that is not associated with a bus type or has a bus type of N/A or tap. For your specific deployment you can verify the interface information using the ethtool -i <name> command on the linux prompt.

All other interfaces that are not ignored will be considered for selection to either the fabric or mbmtbr interface. By default, any interface that has a module or kernel driver of i40e or mlx4_en will be selected to the fabric interface and all others will be selected to the mgmtbr interface.

As the fabric interface is a bond the first interface, sorted alpha numerically by name, will be used as the primary interface.

For the management network an interface bond, mgmtbond, is created to provide redundant network for the physical interfaces. The bridge, mgmtbr, associates this bond interface and various other virtual interfaces together to enable management communication between compute nodes, containers, and virtual machines that make up the management software for CORD.

Customizing Network Configuration

The network configuration can be customized to your deployment using a set of variables that can be set in your deployment configuration file, e.g. podX.yml. There is a set of include, exclude, and ignore variables that operation on the interface name, module type, and bus type. By setting values on these variables it is fairly easy to customize the network settings.

The options are processed as following:

  1. If a given interface matches an ignore option, it is not available to be selected into either the fabric or mgmtbr interface and will not be modified in the /etc/network/interface.
  2. If no include criteria are specified and the given interfaces matches then exclude criteria then the interface will be set as manual configuration in the /etc/network/interface file and will not be auto activated
  3. If no include criteria are specified and the given interface does NOT match the exclude criteria then this interface will be included in either the frabric or mgmtbr interface
  4. If include criteria are specified and the given interface does not match the criteria then the interface will be ignored and its configuration will NOT be modified
  5. If include criteria are specified and the given interface matches the criteria then if the given interface also matches the exclude criteria then this interface will be set as manual configuration in the /etc/network/interface file and will not be auto activated
  6. If include criteria are specified and the given interface matches the criteria and if it does NOT match the exclude criteria then this interface will be included in either the frabric or mgmtbr interface

By default, the only criteria that are specified is the fabric include module types and they are set to i40e,mlx4_en (NOTE: the list is now comma separated and not vertical bar (|) separated.)

If the fabric include module types is specified and the management exclude module types are not specified, then by default the fabric include module types are used as the management exclude module types. This ensures that by default the fabric and the mgmtbr do not intersect on interface module types.

If an external interface is specified in the deployment configuration, this interface will be added to the farbric and management ignore names list.

Each of the criteria is specified as a comma separated list of regular expressions. Default

To set the variables you can use the seedServer.extraVars section in the deployment config file as follows:

seedServer:
  extraVars:
    - 'fabric_include_names=<name1>,<name2>'
    - 'fabric_include_module_types=<mod1>,<mod2>'
    - 'fabric_include_bus_types=<bus1>,<bus2>'
    - 'fabric_exclude_names=<name1>,<name2>'
    - 'fabric_exclude_module_types=<mod1>,<mod2>'
    - 'fabric_exclude_bus_types=<bus1>,<bus2>'
    - 'fabric_ignore_names=<name1>,<name2>'
    - 'fabric_ignore_module_types=<mod1>,<mod2>'
    - 'fabric_ignore_bus_types=<bus1>,<bus2>'
    - 'management_include_names=<name1>,<name2>'
    - 'management_include_module_types=<mod1>,<mod2>'
    - 'management_include_bus_types=<bus1>,<bus2>'
    - 'management_exclude_names=<name1>,<name2>'
    - 'management_exclude_module_types=<mod1>,<mod2>'
    - 'management_exclude_bus_types=<bus1>,<bus2>'
    - 'management_ignore_names=<name1>,<name2>'
    - 'management_ignore_module_types=<mod1>,<mod2>'
    - 'management_ignore_bus_types=<bus1>,<bus2>'

The Ansible scripts configure MAAS to support DHCP/DNS/PXE on the eth2 and mgmtbr interfaces.

Once it has been verified that the ubuntu boot image has been downloaded the compute nodes may be PXE booted.

Note: In order to ensure that the compute node PXE boot the bios settings may have to be adjusted. Additionally, the remote power management on the compute nodes must be enabled.

The compute node will boot, register with MAAS, and then be shut off. After this is complete an entry for the node will be in the MAAS UI at http://head-node-ip-address/MAAS/#/nodes. It will be given a random hostname, in the Canonical way, of a adjective and an noun, such as popular-feast.cord.lab. The name will be different for every deployment. The new node will be in the New state.

As the machines boot they should be automatically transitioned from New through the states of Commissioning and Acquired to Deployed.

Post Deployment Provisioning of the Compute Node

Once the node is in the Deployed state, it will be provisioned for use in a CORD POD by the execution of an Ansible playbook.

Complete

Once the compute node is in the Deployed state and post deployment provisioning on the compute node is complete, this task is complete.

Logs of the post deployment provisioning of the compute nodes can be found in /etc/maas/ansible/logs on the head node.

Additionally, the post deployment provisioning of the compute nodes can be queried using the command cord prov list

cord prov list
ID                                         NAME                   MAC                IP          STATUS      MESSAGE
node-c22534a2-bd0f-11e6-a36d-2c600ce3c239  steel-ghost.cord.lab   2c:60:0c:cb:00:3c  10.6.0.107  Complete
node-c238ea9c-bd0f-11e6-8206-2c600ce3c239  feline-shirt.cord.lab  2c:60:0c:e3:c4:2e  10.6.0.108  Complete
node-c25713c8-bd0f-11e6-96dd-2c600ce3c239  yellow-plot.cord.lab   2c:60:0c:cb:00:f0  10.6.0.109  Complete

The following status values are defined for the provisioning status:

  • Pending- the request has been accepted by the provisioner but not yet started
  • Processing - the request is being processed and the node is being provisioned
  • Complete - the provisioning has been completed successfully
  • Error - the provisioning has failed and the message will be populated with the exit message from provisioning.

Please refer to Re-provision Compute Nodes and Switches if you want to restart this process or re-provision a initialized compute node.

Booting OpenFlow switches

Once the compute nodes have begun their boot process you may also boot the switches that support the leaf spine fabric. These switches should ONIE install boot and download their boot image from MAAS.

Complete

This step is complete when the command completes successfully. You can verify the provisioning of the switches by querying the provisioning service using cord prov list which will show the status of the switches as well as the compute nodes. Switches can be easily identified as their ID will be the MAC address of the switch management interface.

cord prov list
ID                                         NAME                   MAC                IP          STATUS      MESSAGE
cc:37:ab:7c:b7:4c                          spine-1                cc:37:ab:7c:b7:4c  10.6.0.23   Complete
cc:37:ab:7c:ba:58                          leaf-2                 cc:37:ab:7c:ba:58  10.6.0.20   Complete
cc:37:ab:7c:bd:e6                          leaf-1                 cc:37:ab:7c:bd:e6  10.6.0.52   Complete
cc:37:ab:7c:bf:6c                          spine-2                cc:37:ab:7c:bf:6c  10.6.0.22   Complete
node-c22534a2-bd0f-11e6-a36d-2c600ce3c239  steel-ghost.cord.lab   2c:60:0c:cb:00:3c  10.6.0.107  Complete
node-c238ea9c-bd0f-11e6-8206-2c600ce3c239  feline-shirt.cord.lab  2c:60:0c:e3:c4:2e  10.6.0.108  Complete
node-c25713c8-bd0f-11e6-96dd-2c600ce3c239  yellow-plot.cord.lab   2c:60:0c:cb:00:f0  10.6.0.109  Complete

The following status values are defined for the provisioning status:

  • Pending- the request has been accepted by the provisioner but not yet started
  • Processing - the request is being processed and the node is being provisioned
  • Complete - the provisioning has been completed successfully
  • Error - the provisioning has failed and the message will be populated with the exit message from provisioning.

Please refer to Re-provision Compute Nodes and Switches if you want to restart this process or re-provision a initialized switch.

Post Deployment Configuration of XOS / ONOS VTN app

The compute node provisioning process described above (under Booting Compute Nodes) will install the servers as OpenStack compute nodes. You should be able to see them on the CORD head node by running the following commands:

source ~/admin-openrc.sh
nova hypervisor-list

You will see output like the following (showing each of the nodes you have provisioned):

+----+--------------------------+
| ID | Hypervisor hostname      |
+----+--------------------------+
| 1  | sturdy-baseball.cord.lab |
+----+--------------------------+

This step performs a small amount of manual configuration to tell VTN how to route external traffic, and verifies that the new nodes were added to the ONOS VTN app by XOS.

Fabric Gateway Configuration

First, login to the CORD head node and change to the service-profile/cord-pod directory.

To configure the fabric gateway, you will need to edit the file cord-services.yaml. You will see a section that looks like this:

    addresses_vsg:
      type: tosca.nodes.AddressPool
      properties:
        addresses: 10.6.1.128/26
        gateway_ip: 10.6.1.129
        gateway_mac: 02:42:0a:06:01:01

Edit this section so that it reflects the fabric's address block assigned to the vSGs, as well as the gateway IP and MAC address that the vSGs should use to reach the Internet.

Once the cord-services.yaml file has been edited as described above, push it to XOS by running the following:

make cord

Complete

This step is complete once you see the correct information in the VTN app configuration in XOS and ONOS.

To check the VTN configuration maintained by XOS:

  • Go to the "ONOS apps" page in the CORD GUI:
    • URL: http://<head-node>/admin/onos/onosapp/
    • Username: padmin@vicci.org
    • Password: letmein
  • Select VTN_ONOS_app in the table
  • Verify that the Backend status text has a green check with the message successfully enacted
  • Select Attributes tab
  • Look for the rest_onos/v1/network/configuration/ attribute. Verify that its value looks correct for the VTN app's network configuration.

To check that the network configuration has been successfully pushed to the ONOS VTN app and processed by it:

  • Log into ONOS from the head node
    • Command: ssh -p 8102 onos@onos-cord
    • Password: rocks
  • Run the cordvtn-nodes command
  • Verify that the information for all nodes is correct
  • Verify that the initialization status of all nodes is COMPLETE

This will look like the following:

$ ssh -p 8102 onos@onos-cord
Password authentication
Password: # the password is 'rocks'
Welcome to Open Network Operating System (ONOS)!
     ____  _  ______  ____
    / __ \/ |/ / __ \/ __/
   / /_/ /    / /_/ /\ \
   \____/_/|_/\____/___/

Documentation: wiki.onosproject.org
Tutorials:     tutorials.onosproject.org
Mailing lists: lists.onosproject.org

Come help out! Find out how at: contribute.onosproject.org

Hit '<tab>' for a list of available commands
and '[cmd] --help' for help on a specific command.
Hit '<ctrl-d>' or type 'system:shutdown' or 'logout' to shutdown ONOS.

onos> cordvtn-nodes
Hostname                      Management IP       Data IP             Data Iface     Br-int                  State
sturdy-baseball               10.1.0.14/24        10.6.1.2/24         fabric         of:0000525400d7cf3c     COMPLETE
Total 1 nodes

Troubleshoot

If the compute node is not initialized properly (i.e. not in the COMPLETE state): On the compute node, run

sudo ovs-vsctl del-br br-int

On the head node, run

ssh onos@onos-cord -p 8102

(password is 'rocks') and then in the ONOS CLI, run

cordvtn-node-init <compute-node-name>

(name is something like "sturdy-baseball")

Post Deployment Configuration of the ONOS Fabric

Manully Configure Routes on the Compute Node br-int Interface

The routes on the compute node br-int interface need to be manually configured now. Run the following command on compute-1 and compute-2 (nodes in 10.6.1.0/24)

sudo ip route add 10.6.2.0/24 via 10.6.1.254

Run the following command on compute-3 and compute-4 (nodes in 10.6.2.0/24)

sudo ip route add 10.6.1.0/24 via 10.6.2.254

Modify and Apply Fabric Configuration

Configuring the switching fabric for use with CORD is documented in the Fabric Configuration Guide on the OpenCORD wiki.

On the head node is a service that will generate an ONOS network configuration for the leaf/spine network fabric. This configuration is generating by querying ONOS for the known switches and compute nodes and producing a JSON structure that can be POSTed to ONOS to implement the fabric.

The configuration generator can be invoked using the cord generate command. The configuration will be generated to stdout.

Before generating a configuration you need to make sure that the instance of ONOS controlling the fabric doesn't contain any stale data and has has processed a packet from each of the switches and computes nodes. ONOS needs to process a packet because it does not have a mechanism to discover the network, thus to be aware of a device on the network ONOS needs to first receive a packet from it.

To remove stale data from ONOS, the ONOS CLI wipe-out command can be used:

ssh -p 8101 onos@onos-fabric wipe-out -r -j please
Warning: Permanently added '[onos-fabric]:8101,[10.6.0.1]:8101' (RSA) to the list of known hosts.
Password authentication
Password:  # password is 'rocks'
Wiping intents
Wiping hosts
Wiping Flows
Wiping groups
Wiping devices
Wiping links
Wiping UI layouts
Wiping regions

To ensure ONOS is aware of all the switches and the compute nodes, you must have each switch "connect" to the controller and have each compute node ping over its fabric interface to the controller.

If the switches are not already connected the following commands will initiate a connection.

for s in $(cord switch list | grep -v IP | awk '{print $3}'); do
ssh -qftn root@$s ./connect -bg 2>&1  > $s.log
done

You can verify ONOS has recognized the devices using the following command:

ssh -p 8101 onos@onos-fabric devices
Warning: Permanently added '[onos-fabric]:8101,[10.6.0.1]:8101' (RSA) to the list of known hosts.
Password authentication
Password:     # password is 'rocks'
id=of:0000cc37ab7cb74c, available=true, role=MASTER, type=SWITCH, mfr=Broadcom Corp., hw=OF-DPA 2.0, sw=OF-DPA 2.0, serial=, driver=ofdpa, channelId=10.6.0.23:58739, managementAddress=10.6.0.23, protocol=OF_13
id=of:0000cc37ab7cba58, available=true, role=MASTER, type=SWITCH, mfr=Broadcom Corp., hw=OF-DPA 2.0, sw=OF-DPA 2.0, serial=, driver=ofdpa, channelId=10.6.0.20:33326, managementAddress=10.6.0.20, protocol=OF_13
id=of:0000cc37ab7cbde6, available=true, role=MASTER, type=SWITCH, mfr=Broadcom Corp., hw=OF-DPA 2.0, sw=OF-DPA 2.0, serial=, driver=ofdpa, channelId=10.6.0.52:37009, managementAddress=10.6.0.52, protocol=OF_13
id=of:0000cc37ab7cbf6c, available=true, role=MASTER, type=SWITCH, mfr=Broadcom Corp., hw=OF-DPA 2.0, sw=OF-DPA 2.0, serial=, driver=ofdpa, channelId=10.6.0.22:44136, managementAddress=10.6.0.22, protocol=OF_13

To make sure that ONOS is aware of the compute nodes the follow command will a ping over the fabric interface on each of the compute nodes.

for h in localhost $(cord prov list | grep "^node" | awk '{print $4}'); do
ssh -qftn $h ping -c 1 -I fabric 8.8.8.8;
done

You can verify ONOS has recognized the devices using the following command:

ssh -p 8101 onos@onos-fabric hosts
Warning: Permanently added '[onos-fabric]:8101,[10.6.0.1]:8101' (RSA) to the list of known hosts.
Password authentication
Password:     # password is 'rocks'
id=00:16:3E:DF:89:0E/None, mac=00:16:3E:DF:89:0E, location=of:0000cc37ab7cba58/3, vlan=None, ip(s)=[10.6.0.54], configured=false
id=3C:FD:FE:9E:94:28/None, mac=3C:FD:FE:9E:94:28, location=of:0000cc37ab7cba58/4, vlan=None, ip(s)=[10.6.0.53], configured=false
id=3C:FD:FE:9E:94:30/None, mac=3C:FD:FE:9E:94:30, location=of:0000cc37ab7cbde6/1, vlan=None, ip(s)=[10.6.1.1], configured=false
id=3C:FD:FE:9E:98:69/None, mac=3C:FD:FE:9E:98:69, location=of:0000cc37ab7cbde6/2, vlan=None, ip(s)=[10.6.0.5], configured=false

To modify the fabric configuration for your environment, on the head node, generate a new network configuration using the following commands:

cp ~/xos_services/fabric/config/network-cfg-quickstart.json{,.$(date +%Y%m%d-%H%M%S)}
cord generate > ~/xos_services/fabric/config/network-cfg-quickstart.json

Once these steps are done, delete old configuration, apply new configuration, and restart apps in ONOS (still in the XOS VM):

cd ~/service-profile/cord-pod
make delete_fabric_config
make fabric
make reactivate_fabric_apps

To verify that XOS has pushed the configuration to ONOS, log into ONOS in the onos-fabric VM and run netcfg:

$ ssh -p 8101 onos@onos-fabric
Password authentication
Password:     # password is 'rocks'
Welcome to Open Network Operating System (ONOS)!
     ____  _  ______  ____
    / __ \/ |/ / __ \/ __/
   / /_/ /    / /_/ /\ \
   \____/_/|_/\____/___/

Documentation: wiki.onosproject.org
Tutorials:     tutorials.onosproject.org
Mailing lists: lists.onosproject.org

Come help out! Find out how at: contribute.onosproject.org

Hit '<tab>' for a list of available commands
and '[cmd] --help' for help on a specific command.
Hit '<ctrl-d>' or type 'system:shutdown' or 'logout' to shutdown ONOS.

onos> netcfg
{
  "hosts" : {
    "00:00:00:00:00:04/None" : {
      "basic" : {
        "ips" : [ "10.6.2.2" ],
        "location" : "of:0000000000000002/4"
      }
    },
    "00:00:00:00:00:03/None" : {
      "basic" : {
        "ips" : [ "10.6.2.1" ],
        "location" : "of:0000000000000002/3"
      }
    },
... etc.

Update physical host locations in XOS

To correctly configure the fabric when VMs and containers are created on a physical host, XOS needs to associate the location tag of each physical host (from the fabric configuration) with its Node object in XOS. This step needs to be done after new physical compute nodes are provisioned on the POD. To update the node locations in XOS, login to the head node and run the following:

cd ~/service-profile/cord-pod
rm fabric.yaml
make fabric.yaml

Edit fabric.yaml, updating the value field under <hostname>_location_tag to reflect the location of the host from the fabric configuration created in the previous step. Then run:

make fabric

Connect Switches to the controller

We need to manually connect the switches to ONOS after the network config is applied. This can be done by running following ansible script on the head node.

ansible-playbook /etc/maas/ansible/connect-switch.yml

This ansible script will automatically locate all switches in DHCP harvest and connect them to the controller.

Complete

This step is complete when each compute node can ping the fabric IP address of all the other nodes.

Getting Help

If it seems that something has gone wrong with your setup, there are a number of ways that you can get help -- in the documentation on the OpenCORD wiki, on the OpenCORD Slack channel (get an invitation here), or on the CORD-discuss mailing list.

See the How to Contribute to CORD wiki page for more information.

Re-provision Compute Nodes and Switches

If you would like to re-provision a switch or a compute node the cord prov delete command can be used. This command takes one or more IDs as parameters and will delete the provisioning records for these devices. This will cause the provisioner to re-provision them.

You can also use the argument --all, which will delete all known provisioning records.

cord prov delete node-c22534a2-bd0f-11e6-a36d-2c600ce3c239
node-c22534a2-bd0f-11e6-a36d-2c600ce3c239 DELETED
cord prov delete --all
cc:37:ab:7c:b7:4c DELETED
cc:37:ab:7c:ba:58 DELETED
cc:37:ab:7c:bd:e6 DELETED
cc:37:ab:7c:bf:6c DELETED
node-c22534a2-bd0f-11e6-a36d-2c600ce3c239 DELETED
node-c238ea9c-bd0f-11e6-8206-2c600ce3c239 DELETED
node-c25713c8-bd0f-11e6-96dd-2c600ce3c239 DELETED