Versions Compared

Old Version 6

changes.mady.by.user Wenhui Zhang

Saved on

compared with

New Version Current

changes.mady.by.user Wenhui Zhang

Saved on

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

This section provides instructions to quickly bring up SEBA.

Contents

Table of Contents
excludeContents

Note: This Installation Guide assumes that prerequisite hardware is met and software (Akraino Stack with CORD Platform) have already been installed.

Specifically, wait for the three EtcdCluster CustomResourceDefinitions to appear in Kubernetes:

kubectl get crd | grep etcd | wc -l

Once the CRDs are present, proceed with the seba chart installation.

Overview

This page walks through the sequence of Helm operations needed to bring up the SEBA profile.

Prerequisites

It assumes the Akraino Stack with CORD Platform has already been installed.

Installation

Install components as a whole

Add the CORD repository and update indexes

Code Block
$ helm repo add cord https://charts.opencord.org
$ helm repo update

Install the CORD platform
Code Block
$ helm install -n cord-platform --version 6.1.0 cord/cord-platform

...

Code Block
$ kubectl get crd | grep -i etcd | wc -l

Install the SEBA profile
Code Block
$ helm install -n seba --version 1.0.0 cord/seba

...

This section provides instructions to quickly bring up SEBA.

Contents

Table of Contents
excludeContents


Note: This Installation Guide assumes that prerequisite hardware is met and software (Akraino Stack with CORD Platform) have already been installed.

Specifically, wait for the three EtcdCluster CustomResourceDefinitions to appear in Kubernetes:

kubectl get crd | grep etcd | wc -l

Once the CRDs are present, proceed with the seba chart installation.

Overview

This page walks through the sequence of Helm operations needed to bring up the SEBA profile.

Prerequisites

It assumes the Akraino Stack with CORD Platform has already been installed.

Installation

Install components as a whole

Add the CORD repository and update indexes

Code Block
$ helm install repo add cord https://charts.opencord.org
$ helm repo update

Install the CORD platform
Code Block
$ helm install -n attcord-workflowplatform --version 6.1.0.2 cord/attcord-workflow

Alternatively, install as separate components

...

platform

Wait until 3 etcd CRDs are present in Kubernetes
Code Block
$ helmkubectl get repocrd add| incubator http://storage.googleapis.com/kubernetes-charts-incubator
$ helm repo update

...

grep -i etcd | wc -l

Install the SEBA profile
Code Block
$ helm repo add cord https://charts.opencord.org
$ helm repo updateinstall -n seba --version 1.0.0 cord/seba


Install the CORD platform componentsAT&T workflow
Code Block
$ helm install -n onos cord/onos
$ helm install -n xos-coreatt-workflow --version 1.0.2 cord/xos-core
$ helm install --version 0.13.3 \
             --set configurationOverrides."offsets.topic.replication.factor"=1 \
             --set configurationOverrides."log.retention.hours"=4 \
             --set configurationOverrides."log.message.timestamp.type"="LogAppendTime" \
             --set replicas=1att-workflow

Alternatively, install as separate components

Add the official Kubernetes incubator repository (for Kafka) and update the indexes
Code Block
$ helm repo add incubator http://storage.googleapis.com/kubernetes-charts-incubator
$ helm repo update

Add the CORD repository and update the indexes
Code Block
$ helm repo add cord https://charts.opencord.org
$ helm repo update

Install the CORD platform components
Code Block
$ helm install -n onos cord/onos
$ helm install -n xos-core cord/xos-core
$ helm install --version 0.13.3 \
             --set persistence.enabled=falseconfigurationOverrides."offsets.topic.replication.factor"=1 \
             --set zookeeper.replicaCount=1configurationOverrides."log.retention.hours"=4 \
             --set zookeeper.persistence.enabled=falseconfigurationOverrides."log.message.timestamp.type"="LogAppendTime" \
             -n cord-kafka incubator/kafka

Optionally, install the logging and monitoring infrastructure components
Code Block
$ helm install -n nem-monitoring cord/nem-monitoring
$ helm installset replicas=1 \
             --set elasticsearch.cluster.env.MINIMUM_MASTER_NODES="1"persistence.enabled=false \
             --set elasticsearchzookeeper.client.replicasreplicaCount=1 \
             --set elasticsearchzookeeper.masterpersistence.replicasenabled=2false \
             -n cord-set elasticsearch.master.persistence.enabled=false \
             kafka incubator/kafka

Optionally, install the logging and monitoring infrastructure components
Code Block
$ helm install -n nem-monitoring cord/nem-monitoring
$ helm install --set elasticsearch.data.replicas=1cluster.env.MINIMUM_MASTER_NODES="1" \
             --set elasticsearch.dataclient.persistence.enabledreplicas=false1 \
             -n logging cord/logging

Install etcd-operator and wait until 3 etcd CRDs are present in Kubernetes
Code Block
$ helm install -n etcd-operator stable/etcd-operator --version 0.8.3
$ kubectl get crd | grep -i etcd | wc -l

Install the rest of the SEBA profile components
Code Block
$ helm install -n voltha cord/voltha
$ helm install -n seba-service cord/seba-services
$ helm install -n base-kubernetes cord/base-kubernetes

...

-set elasticsearch.master.replicas=2 \
             --set elasticsearch.master.persistence.enabled=false \
             --set elasticsearch.data.replicas=1 \
             --set elasticsearch.data.persistence.enabled=false \
             -n logging cord/logging

Install etcd-operator and wait until 3 etcd CRDs are present in Kubernetes
Code Block
$ helm install -n att-workflow etcd-operator stable/etcd-operator --version 10.08.2 cord/att-workflow

Verify your installation and next steps

Once the installation completes, monitor your setup using kubectl get pods. Wait until all pods are in Running state and “tosca-loader” pods are in Completed state.

Note: The tosca-loader pods may periodically transition into error state. This is expected. They will retry and eventually get to the desired state. Note: Depending on the profile you're installing, you may need to check also different namespaces (for example, check the voltha namespace if you're installing SEBA with kubectl get pods -n voltha)

Your POD is now installed and ready for use.

POD Configuration

Once all the components needed for the SEBA profile are up and running on your POD, you will need to configure it. This is typically done using TOSCA.

In this page we are describing the process as a three steps process:

  • Fabric Setup
  • OLT Provisioning
  • Subscriber Provisioning

as that is what logically makes sense, but be aware that all the configurations can be unified in a single TOSCA file.

This configuration is environment specific, so you will need to create your own, but the following can serve as a reference:

Fabric Setup

Code Block
languageyml
tosca_definitions_version: tosca_simple_yaml_1_0
imports:
 - custom_types/switch.yaml
 - custom_types/switchport.yaml
 - custom_types/portinterface.yaml
 - custom_types/bngportmapping.yaml
 - custom_types/attworkflowdriverwhitelistentry.yaml
 - custom_types/attworkflowdriverservice.yaml
 - custom_types/serviceinstanceattribute3
$ kubectl get crd | grep -i etcd | wc -l

Install the rest of the SEBA profile components
Code Block
$ helm install -n voltha cord/voltha
$ helm install -n seba-service cord/seba-services
$ helm install -n base-kubernetes cord/base-kubernetes

Install the AT&T workflow
Code Block
$ helm install -n att-workflow --version 1.0.2 cord/att-workflow

Verify your installation and next steps

Once the installation completes, monitor your setup using kubectl get pods. Wait until all pods are in Running state and “tosca-loader” pods are in Completed state.

Note: The tosca-loader pods may periodically transition into error state. This is expected. They will retry and eventually get to the desired state. Note: Depending on the profile you're installing, you may need to check also different namespaces (for example, check the voltha namespace if you're installing SEBA with kubectl get pods -n voltha)

Your POD is now installed and ready for use.

POD Configuration

Once all the components needed for the SEBA profile are up and running on your POD, you will need to configure it. This is typically done using TOSCA.

In this page we are describing the process as a three steps process:

  • Fabric Setup
  • OLT Provisioning
  • Subscriber Provisioning

as that is what logically makes sense, but be aware that all the configurations can be unified in a single TOSCA file.

This configuration is environment specific, so you will need to create your own, but the following can serve as a reference:

Fabric Setup

Code Block
languageyml
tosca_definitions_version: tosca_simple_yaml_1_0
imports:
 - custom_types/switch.yaml
 - custom_types/onosappswitchport.yaml
 description: Configures a full SEBA POD

topology_- custom_types/portinterface.yaml
 - custom_types/bngportmapping.yaml
 - custom_types/attworkflowdriverwhitelistentry.yaml
 - custom_types/attworkflowdriverservice.yaml
 - custom_types/serviceinstanceattribute.yaml
 - custom_types/onosapp.yaml

description: Configures a full SEBA POD

topology_template:
 node_templates:
    # Fabric configuration
    switch#leaf_1:
 type: tosca.nodes.Switch
 properties:
 driver: ofdpa3
 ipv4Loopback: 192.168.0.201
 ipv4NodeSid: 17
 isEdgeRouter: True
 name: AGG_SWITCH
 ofId: of:0000000000000001
 routerMac: 00:00:02:01:06:01

    # Setup the OLT switch port
    port#olt_port:
 type: tosca.nodes.SwitchPort
 properties:
 portId: 1
 host_learning: false
 requirements:
 - switch:
 node: switch#leaf_1
 relationship: tosca.relationships.BelongsToOne

    # Port connected to the BNG
    port#bng_port:
 type: tosca.nodes.SwitchPort
 properties:
 portId: 31
 requirements:
 - switch:
 node: switch#leaf_1
 relationship: tosca.relationships.BelongsToOne

    # Setup the fabric switch port where the external
    # router is connected to
 bngmapping:
 type: tosca.nodes.BNGPortMapping
 properties:
 s_tag: any
 switch_port: 31

    # DHCP L2 Relay config
    onos_app#dhcpl2relay:
 type: tosca.nodes.ONOSApp
 properties:
 name: dhcpl2relay
 must-exist: true

 dhcpl2relay-config-attr:
 type: tosca.nodes.ServiceInstanceAttribute
 properties:
 name: /onos/v1/network/configuration/apps/org.opencord.dhcpl2relay
 value: > { "dhcpl2relay" : { "useOltUplinkForServerPktInOut" : false, "dhcpServerConnectPoints" : [ "of:0000000000000001/31" ] } }  requirements:
 - service_instance:
 node: onos_app#dhcpl2relay
 relationship: tosca.relationships.BelongsToOne

...

Code Block
(voltha) device 00015698e67dc060
(device 00015698e67dc060) show
Device 00015698e67dc060
+------------------------------+------------------+
|                        field |            value |
+------------------------------+------------------+
|                           id | 00015698e67dc060 |
|                         type |     broadcom_onu |
|                         root |             True |
|                    parent_id | 0001941bd45e71d8 |
|                       vendor |         Broadcom |
|                        model |              n/a |
|             hardware_version |     to be filled |
|             firmware_version |     to be filled |
|                 images.image |        1 item(s) |
|                serial_number |     BRCM22222222 |
+------------------------------+------------------+
|                      adapter |     broadcom_onu |
|                  admin_state |                3 |
|                  oper_status |                4 |
|               connect_status |                2 |
|      proxy_address.device_id | 0001941bd45e71d8 |
|         proxy_address.onu_id |                1 |
| proxy_address.onu_session_id |                1 |
|               parent_port_no |        536870912 |
|                    vendor_id |             BRCM |
|                        ports |        2 item(s) |
+------------------------------+------------------+
|                  flows.items |        5 item(s) |
+------------------------------+------------------+
to find the correct serial number.

Push a Subscriber into CORD

Once you have this information, you can create the subscriber by customizing the following TOSCA and passing it into the POD:

Code Block
tosca_definitions_version: tosca_simple_yaml_1_0
imports:
 - custom_types/rcordsubscriber.yaml
description: Create a test subscriber
topology_template:
 node_templates:
    # A subscriber
 my_house:
 type: tosca.nodes.RCORDSubscriber
 properties:
 name: My House
 c_tag: 111
 s_tag: 222
 onu_device: BRCM1234 # Serial Number of the ONU Device to which this subscriber is connected

Using TOSCA to push to CORD

Once CORD is up and running, a node can be added to a POD using the TOSCA interface by uploading the following recipe:

Code Block
tosca_definitions_version: tosca_simple_yaml_1_0

description: Load a compute node in XOS

imports:
 - custom_types/node.yaml

topology_template:
 node_templates:

    # A compute node
 GratefulVest:
 type: tosca.nodes.Node
 properties:
 name: Grateful Vest

In TOSCA terminology, the above would be called a TOSCA node template.

Where to find the generated specs?

On any running CORD POD, the TOSCA apis are accessible as:

Code Block
$ curl http://<head-node-ip>:<head-node-port>/xos-----------+
to find the correct serial number.

Push a Subscriber into CORD

Once you have this information, you can create the subscriber by customizing the following TOSCA and passing it into the POD:

Code Block
tosca_definitions_version: tosca_simple_yaml_1_0
imports:
 - custom_types/rcordsubscriber.yaml
description: Create a test subscriber
topology_template:
 node_templates:
    # A subscriber
 my_house:
 type: tosca.nodes.RCORDSubscriber
 properties:
 name: My House
 c_tag: 111
 s_tag: 222
 onu_device: BRCM1234 # Serial Number of the ONU Device to which this subscriber is connectedtosca | python -m json.tool


And it will return a list of all the recipes with the related url:

{
  "image": "/custom_type/image",
  "site": "/custom_type/site",
  ...
}

For examples, to site the TOSCA spec of the Site model, you can use the URL:

Code Block
$ curl http://<head-node-ip>:<head-node-port>/xos-tosca/custom_type/site


If you have a running xos-tosca container you can also find generated copies of the specs in /opt/xos-tosca/src/tosca/custom_types.

How to load a TOSCA recipe in the system

The xos-tosca container exposes two endpoint:

Code Block
POST http://<cluster-ip>:<tosca-port>/run
POST http://<cluster-ip>:<tosca-port>/delete


To load a recipe via curl you can use this command:

Code Block
$ curl -H "xos-username: xosadmin@opencord.org" -H "xos-password: <xos-password>" -X POST --data-binary @<path/to/file> http://<cluster-ip>:<tosca-port>/run

If you installed the xos-core charts without modifications, the tosca-port is 30007.

References