Deploy

Learn how to deploy into a private edition environment.

Related documentation:

• How Digital Channels works
• Suite-level guide
• All services
• Digital Channels Release Notes

RSS:

• For private edition

Assumptions

• The instructions on this page assume you are deploying the service in a service-specific namespace, named in accordance with the requirements on Creating namespaces. If you are using a single namespace for all private edition services, replace the namespace element in the commands on this page with the name of your single namespace or project.
• Similarly, the configuration and environment setup instructions assume you need to create namespace-specific (in other words, service-specific) secrets. If you are using a single namespace for all private edition services, you might not need to create separate secrets for each service, depending on your credentials management requirements. However, if you do create service-specific secrets in a single namespace, be sure to avoid naming conflicts.

Deploy AI Connector

Important

Make sure to review Before you begin for the full list of prerequisites required to deploy Digital Channels AI Connector.

Prepare your environment

To prepare your environment for the deployment, complete the steps in this section for either OpenShift or Google Kubernetes Engine (GKE).

OpenShift

Log in to the OpenShift cluster from the host where you will run the deployment:

oc login --token <token> --server <url of api server>

First confirm the cluster is running:

oc get clusterversion

Create a new project for Digital Channels:

oc new-project athena

GKE

Log in to the GKE cluster from the host where you will run the deployment:

gcloud container clusters get-credentials <cluster>

Create a JSON file called create-nexus-namespace.json with the following content:

{
  "apiVersion": "v1",
  "kind": "Namespace",
  "metadata": {
    "name": "athena",
    "labels": {
      "name": "athena"
    }
  }
}

Use the JSON file to create a new namespace for Digital Channels AI Connector:

kubectl apply -f apply create-athena-namespace.json

Now confirm the namespace was created:

kubectl describe namespace athena

Configure a secret to access JFrog

If you haven't done so already, create a secret for accessing the JFrog registry:

kubectl create secret docker-registry <credential-name> --docker-server=<docker repo> --docker-username=<jfrog_username> --docker-password=<API_key_from_jfrog>

For OpenShift deployments, map the secret to the default service account (this is an optional step):

oc secrets link default <credential-name> --for=pull

Deploy

To deploy AI Connector, you'll need the Helm package and override files you downloaded in a previous step. Copy values.yaml and the Helm package (athena-<version>.tgz) to the installation location.

You must override the following key sections in values.yaml:

• image.*
• athena.nexus.*
• athena.redis.*
• athena.db.*
• ingress.*

Here's an example of how your values.yaml file might look:

# Default values for athena.
# This is a YAML-formatted file.
# Declare variables to be passed into your templates.
  
version: "100.0.124.3419" # AI Connector Version
  
nameOverride: ""
  
fullnameOverride: ""
  
replicaCount: 1
  
image:
  registry: "pureengage-docker-staging.jfrog.io"
  repository: nexus/athena
  pullPolicy: IfNotPresent
  pullSecrets:
    - name: <pullsecret>
  
serviceAccount:
  create: false
  name: ""
  annotations: {}
  
podAnnotations: {}
  
podLabels: {}
  
podSecurityContext:
  runAsNonRoot: true
  runAsUser: 500
  runAsGroup: 500
  fsGroup: 500
  
securityContext: {}
  
configChecksum: true
  
secretChecksum: true
  
containerPort: 4084
  
service:
  enabled: true
  type: ClusterIP
  annotations: {}
  port: 80
  
ingress:
  enabled: false
  annotations: {}
  hosts:
    - host: athena.local
      paths: []
  tls: []
  #  - secretName: athena-tls-secret
  #    hosts:
  #      - athena.local
  
resources: {}
  # limits:
  #   cpu: 100m
  #   memory: 128Mi
  # requests:
  #   cpu: 100m
  #   memory: 128Mi
  
nodeSelector: {}
  
tolerations: []
  
affinity: {}
  
priorityClassName: ""
  
dnsPolicy: ClusterFirst
  
dnsConfig:
  options:
    - name: ndots
      value: "3"
  
monitoring:
  enabled: false
  
athena:
  server:
    apiPrefix: "/nexus/v3"
  nexus:
    url: "<nexus-url>"
    apiPrefix: "/nexus/v3"
    apiKey: "<cluster-api-key>"
    timeout: 10000
  db:
    host: "<postgres-host>"
    port: 5432
    user: "<user>"
    password: "<password>"
    database: "<dbname>"
    ssl: false
  redis:
    nodes: "redis://<redis-host>:6379"
    password: "<password>"
    cluster: true
    tls: false
  google:
    speechApiKey: ""

Run the following command to install AI Connector:

helm upgrade --install <deployment-name> <helm-repo>/athena-<version>.tgz -f values.yaml

Validate the deployment

To validate the deployment, first run the following code snippet

kubectl port-forward service/athena <port>:80

and then send GET request on the following URL

$athenaURL/health/detail 

Where $athenaURL is the fully qualified domain name (FQDN) for AI Connector.

The response should look like this:

{
    "buildInfo": {
        "version": "100.0.001.97446",
        "changeset": "565f432fa8f4555276b55e8237cebcfb201b986e",
        "timestamp": "Mon Jan 17 10:21:22 UTC 2022"
    },
    "startTime": "2022-03-17T13:15:22.873Z",
    "upTime": 49338032,
    "os": {
        "hostname": "athena-6bb9c5c68f-bz449",
        "upTime": 52366.39,
        "freemem": 1316397056,
        "loadavg": [0.35, 0.37, 0.63],
        "totalmem": 4124729344
    },
    "memoryUsage": {
        "rss": 178757632,
        "heapTotal": 83382272,
        "heapUsed": 80987072,
        "external": 1890524,
        "arrayBuffers": 126610
    },
    "redis": {
        "state": "READY",
        "latency": 5
    },
    "db": {
        "latency": 202
    },
    "isReady": true
}

The deployment was successful if state="green". You can also confirm that db.ready=true and redis.ready=true.

Uninstall

Execute the following command to uninstall AI Connector:

helm delete <deployment-name> -n <namespace>

Next steps

Complete the steps in Provisioning overview to finish deploying AI Connector.