Installing on AWS
      Back to home
      1. Zetaris Support - Knowledge Base
      2. Zetaris Deployment
      3. Installing on AWS

      Installing on AWS EKS using helm charts (non-marketplace)

      Please follow the below prerequisites and deployment instructions to deploy Zetaris on EKS environment. 

      Prerequisites: 

      You can ensure the CLI environment to run the helm and kubectl command. Either it can be on AWS Cloud Shell OR Local CLI (Shell, CMD or PowerShell) 

      • Helm 

      https://helm.sh/docs/intro/install/  

      • Kubectl CLI 

      https://kubernetes.io/docs/tasks/tools/#kubectl  

      • AWS CLI 

      https://docs.aws.amazon.com/cli/latest/userguide/getting-started-install.html#getting-started-install-instructions  

      • Configure AWS Credentials using “aws configure” OR “aws configure sso” 

      https://docs.aws.amazon.com/cli/latest/reference/configure/  

      https://medium.com/@pushkarjoshi0410/how-to-set-up-aws-cli-with-aws-single-sign-on-sso-acf4dd88e056 

      • DBeaver Installation OR  "postgres DB cli"

      https://dbeaver.io/download/ 

      This product requires an internet connection to deploy properly. The following packages are downloaded on deployment.
      This product is going to be hosted on secure a virtual private cloud network.
      If pre-requisites installation are failed, please don't continue to do further installation. You can reach out to Zetaris support team.
       If you get stuck on the helm installation or missed any steps, please don't continue to do further steps. Please reach out to Zetaris support team.

      Deployment Instructions: 

      1). Deploy the below CloudFormation template on your AWS account. Update require parameters such as KeyName, DBUsername, DBPassword..etc 

      https://zetarislogo.s3.ap-southeast-2.amazonaws.com/aws+zetaris.json

      Creating a new Cloudformation stack

      Sign in to the AWS Management Console and open the AWS CloudFormation console at https://console.aws.amazon.com/cloudformation.

      Choose Create Stack.

      In the Specify template section, select Amazon S3 Template URL and then choose Next:

      https://zetarislogo.s3.ap-southeast-2.amazonaws.com/aws+zetaris.json  

      In the Specify stack details section, enter a name in the Stack name field. For this example, use ZetarisStack. The stack name can't contain spaces.

      In the Parameters section, you ensure to fill all the values for all parameters including DBUsername, DBPassword, and KeyName. In the KeyName field, enter the name of a valid Amazon EC2 pair in the same region you are creating the stack.

      Choose Next.

      In this scenario, we won't add any tags. Choose Next. Tags, which are key-value pairs, can help you identify your stacks. 

      Review the information for the stack. When you're satisfied with the settings, choose Submit.

      2). Follow the instructions to add an OIDC provider for your cluster. 

      https://docs.aws.amazon.com/eks/latest/userguide/enable-iam-roles-for-service-accounts.html 

       

      3). Request for a public certificate for fully qualified domains from Zetaris and email us for approval. 

      (i.e) 

      uitest.enterprise.zetaris.com 

      apitest.enterprise.zetaris.com 

      airflow.enterprise.zetaris.com 

      privateai.enterprise.zetaris.com 

      avatar.enterprise.zetaris.com 

      zeppelin.ui.enterprise.zetaris.com

      opensearchdb.enterprise.zetaris.com

      opensearch.enterprise.zetaris.com

      Note the domain names and the cert ARN once its issued. 

      (i.e) 

      4). Connect your EKS cluster and ensure the nodes are in ready state. 

      aws eks --region <region-code> update-kubeconfig --name <cluster_name>

      https://medium.com/@kevken1000/eks-connecting-to-the-cluster-3c20e0e39a0f  

      5). Create service account for AWS Load balancer controller. 

      Replace the AWS ID with your AWS ID 

      cat >aws-load-balancer-controller-service-account.yaml <<EOF
      apiVersion: v1
      kind: ServiceAccount
      metadata:
        labels:
          app.kubernetes.io/component: controller
          app.kubernetes.io/name: aws-load-balancer-controller
        name: aws-load-balancer-controller
        namespace: kube-system
        annotations:
          eks.amazonaws.com/role-arn: arn:aws:iam::281917764651:role/aws-load-balancer-controller
      EOF
      kubectl apply -f aws-load-balancer-controller-service-account.yaml
      helm repo add eks https://aws.github.io/eks-charts
      helm repo update eks
      helm install aws-load-balancer-controller eks/aws-load-balancer-controller \
        -n kube-system \
        --set clusterName=Zetaris \
        --set serviceAccount.create=false \
        --set serviceAccount.name=aws-load-balancer-controller
      kubectl get deployment -n kube-system aws-load-balancer-controller

      Helm

      curl -fsSL https://raw.githubusercontent.com/helm/helm/main/scripts/get-helm-3 | bash

      3. Setup the ‘namespaces’

      · Login to OpenShift / Kubernetes cluster using kubeadmin credentials from oc/kubectl command line tool.

      kubectl create namespace zetaris
      kubectl create namespace airflow

      Set the below variables for the current version release in your terminal.

      Please contact Zetaris for the token below

      $zettoken=github_pat_XXXXXXX
      $zetrelease=helm_latest

      Set the env value to reflect the Infrastructure provider 

      like 'aws', 'azure', 'ibm', 'gcp', 'dell', 'hitachi', 'local' (minikube)

      $environment=local

      Postgres installation

      Please install Postgres external using the below two option

       

      A).Postgres on external source like RDS or other managed services. Go with the low cost one like Azure flexible server, AWS Aurora postgres..etc

      OR

      B) .Postgres on k8s inside the cluster.

      Run “kubectl get sc” to confirm the storage class

      Set appropriate “storageClassName”. “kubectl get sc”

      $storageclass=default
      helm repo add helm-postgres https://$zettoken@raw.githubusercontent.com/zetaris/openshift/$zetrelease/postgres
      helm upgrade --install postgres \

      helm-postgres/postgres --namespace zetaris \

      --set storageClassName=$storageclass \

      --set environment=$environment

      List out all generated pods

      kubectl get pods -n zetaris

      Get the pod name of postgres

      kubectl exec -it <postgres pod name> -n zetaris -- /bin/sh
      su - postgres

      The below command would ask password and enter “<Your Choice Password>”

      psql -h postgres -U rds_admin postgres
      create database metastore;
      create database auditlog;
      create database airflowdb;

       

      5. Install the Spark Operator and Opensearch

      · Add the spark operator, cert repo;

      helm repo add spark-operator https://kubeflow.github.io/spark-operator 

      helm repo add jetstack https://charts.jetstack.io
      helm repo update

      · Install the spark operator , cert manager;

      helm upgrade --install spark-operator spark-operator/spark-operator --namespace spark-operator --create-namespace --version=1.2.15 --set webhook.enable=true
      kubectl label namespace cert-manager certmanager.k8s.io/disable-validation=true
      kubectl create ns cert-manager

       
      helm install cert-manager jetstack/cert-manager --namespace cert-manager --version v1.7.0 --set installCRDs=true

      Install the opensearch;

      Option 1: Using helm charts

      helm uninstall opensearch -n zetaris
      helm repo add opensearch https://opensearch-project.github.io/helm-charts/
        
      helm upgrade --install opensearch opensearch/opensearch --namespace zetaris --set image.tag=2.11.0 --set serviceAccount.name=zetaris-sa

      Option 2:  Using git repo

      git clone https://github.com/zetaris/opensearch.git
      helm uninstall opensearch -n zetaris
      helm upgrade --install opensearch opensearch/hitachi --namespace zetaris --set image.tag=2.11.0 --set serviceAccount.name=zetaris-sa

      6. Lightning Deployments

      Add below helm repos,

      helm repo add helm-zetaris-lightning-solr https://$zettoken@raw.githubusercontent.com/zetaris/HelmDeployment/$zetrelease/solr/helm/
      helm repo add helm-zetaris-lightning-server https://$zettoken@raw.githubusercontent.com/zetaris/zetaris-lightning/$zetrelease/deployments/helm/
      helm repo add helm-zetaris-lightning-api https://$zettoken@raw.githubusercontent.com/zetaris/lightning-api/$zetrelease/deployments/helm/
      helm repo add helm-zetaris-lightning-gui https://$zettoken@raw.githubusercontent.com/zetaris/lightning-gui/$zetrelease/deployments/helm/
      helm repo add helm-zetaris-lightning-zeppelin https://$zettoken@raw.githubusercontent.com/zetaris/zetaris-zeppelin/$zetrelease/deployments/helm/
      helm repo add helm-zetaris-digiavatar https://$zettoken@raw.githubusercontent.com/zetaris/digiavatar/$zetrelease/deployments/helm/
      helm repo add helm-zetaris-privateai https://$zettoken@raw.githubusercontent.com/zetaris/privateai/$zetrelease/deployments/helm/
      helm repo add helm-zetaris-airflow-ing https://$zettoken@raw.githubusercontent.com/zetaris/HelmDeployment/$zetrelease/airflow-ing/helm/

      Replace the below values and apply helm install,

      Update image and domain configs.

      serverImage=zetregistry.azurecr.io/lightning-server:v2.4.1.0-latest

      apiImage=zetregistry.azurecr.io/lightning-api:v2.4.1.0-latest

      guiImage=zetregistry.azurecr.io/lightning-gui:v2.4.1.0-latest

      computesparkimage=zetregistry.azurecr.io/lightning-server:v2.4.1.0-latest

      computeprestoimageRepo=zetregistry.azurecr.io/lightning-presto-app

      computeprestoimageTag=zv2.4.1.0-latest

      zeppelinImage=zetregistry.azurecr.io/lightning-zeppelin-app:zv2.4.1.0-latest

       

      Update postgres details here,

      dbdns=postgres

      dbport=5432

      metadb=metastore

      auditdb=audit_log

      dbuser=rds_admin

      dbpwd=<Your chosen password in Step 4>

      sslmode=require

      metastoreJdbcUrl="jdbc:postgresql://$dbdns:$dbport/$metadb?user=$dbuser&password=$dbpwd&sslmode=$sslmode"

      auditLogJdbcUrl="jdbc:postgresql://$dbdns:$dbport/$auditdb?user=$dbuser&password=$dbpwd&sslmode=$sslmode"

      For AWS EFS as storage

      awsefsid="". <EFS Name> 

      awsefsdata="" <EFS Access Point  -create for data> 

      awsefszeppelin="". <EFS Access Point  -create for Zeppelin with 1001 and 0777 permission>  

      tls_cert_arn="" <Available from Cert Manager once certs are ready>

      Set the protocol (In case the customer wants to proceed without certificate then)

      dnsprotocol=http

      For guidance on creating public and private keys, please refer to the following resource: https://stackoverflow.com/questions/44474516/how-to-create-public-and-private-key-with-openssl.

      openssl genpkey -algorithm RSA -out private_key.pem -pkeyopt rsa_keygen_bits:2048
      openssl rsa -pubout -in private_key.pem -out public_key.pem
      openssl rsa -in private_key.pem -outform DER -out private_key.der
      openssl rsa -in private_key.pem -pubout -outform DER -out public_key.der
      base64 private_key.der > private_key_der.txt
      base64 private_key.der > private_key_der.txt

      privateKeyDer=$(openssl rsa -in private_key.pem -outform DER | base64 -w0)

      publicKeyDer==$(openssl rsa -in private_key.pem -pubout -outform DER | base64 -w0)

      Lightning solr deployment

      helm upgrade --install lightning-solr \
      helm-zetaris-lightning-solr/solr \
      --namespace zetaris \
      --set storageClassName=$storageclass \
      --set environment=$environment

      Lightning server deployment

      helm upgrade --install lightning-server \
      helm-zetaris-lightning-server/lightning-server \
      --namespace zetaris \
      --set db.metastore.jdbcUrl=$metastorejdbcUrl \
      --set db.auditLog.dbcUrl=$auditLogjdbcUrl \
      --set storage.storageClass.name=$storageclass \
      --set environment=$environment \
      --set encryption.privateKeyDer=$privateKeyDer \
      --set encryption.publicKeyDer=$publicKeyDer \
      --set storage.aws.efs.id=$awsefsid \
      --set storage.aws.efs.data=$awsefsdata \
      --set serverImage=$serverImage \
      --set storage.storageClass.create=false \
      --set storage.aws.efs.driver=efs.csi.aws.com

      Lightning API deployment

      helm upgrade --install lightning-api \
      helm-zetaris-lightning-api/lightning-api \
      --namespace zetaris \
      --set environment=$environment \
      --set apiImage=$apiImage \
      --set ingress.protocol=$dnsprotocol \
      --set aws.ingress.tls_cert_arn=$tls_cert_arn \
      --set ingress.baseDomain=$basednsname \
      --set db.metastore.jdbcUrl=$metastorejdbcUrl \
      --set db.auditLog.jdbcUrl=$auditLogjdbcUrl \
      --set compute.spark.image=$computesparkimage \
      --set compute.presto.imageRepo=$computeprestoimageRepo \
      --set compute.presto.imageTag=$computeprestoimageTag

      Lightning GUI deployment

      helm upgrade --install lightning-gui \
      helm-zetaris-lightning-gui/lightning-gui \
      --namespace zetaris \
      --set guiImage=$guiImage \
      --set ingress.protocol=$dnsprotocol \
      --set aws.ingress.tls_cert_arn=$tls_cert_arn \
      --set ingress.baseDomain=$basednsname \
      --set environment=$environment

      Lightning Zeppelin deployment

      helm upgrade --install lightning-zeppelin \
      helm-zetaris-lightning-zeppelin/lightning-zeppelin \
      --namespace zetaris \
      --set ingress.protocol=$dnsprotocol \
      --set ingress.baseDomain=$basednsname \
      --set ingress.aws.tls_cert_arn=$tls_cert_arn \
      --set storage.storageClass.name=$storageclass \
      --set environment=$environment \
      --set storage.aws.efs.id=$awsefsid \
      --set storage.aws.efs.zeppelin=$awsefszeppelin \
      --set zeppelin.image=$zeppelinImage

      Digiavatar deployment

      helm upgrade --install digiavatar helm-zetaris-digiavatar/digiavatar --namespace zetaris \
      --set ingressprotocol=$dnsprotocol --set ingress.baseDomain=$basednsname \
      --set environment=$environment \
      --set serviceaccount=zetaris-sa

      Private AI deployment

      efsDriver=efs.csi.aws.com
      efsAccesspointPrivateai= <Access Point ID from EFS Instance for PrivateAI>

      helm upgrade --install privateai helm-zetaris-privateai/privateai --namespace zetaris \
      --set ingress.baseDomain=$basednsname \
      --set ingressprotocol=$dnsprotocol \
      --set environment=$environment \
      --set gpuenabled=false \
      --set storageclass=$storageclass \
      --set serviceaccount.name=zetaris-sa \
      --set ingress.aws.tls_cert_arn=$tlsCertArn \
      --set aws.efs.id=$efsId \
      --set aws.efs.driver=$efsDriver \
      --set aws.efs.privateai=$efsAccesspointPrivateai

      Airflow Deployment

      efsAccesspointDags= <Access Point ID from EFS Instance for Airflow> 
      helm upgrade --install airflow-ing \
      helm-zetaris-airflow-ing/airflow-ing \
      --namespace airflow \
      --set environment=$environment \
      --set storage.aws.efs.id=$efsId \
      --set storage.aws.efs.driver=$efsDriver \
      --set storage.aws.efs.data=$efsAccesspointDags \
      --set ingress.baseDomain=$basednsname \
      --set ingress.aws.tls_cert_arn=$tlsCertArn \
      --set  aws.dataVolume=pv-dags-aws

      Final deployment should contain running pods as shown below;

      NAME                                                  READY   STATUS             RESTARTS       AGE
      digiavatar-5f69b44d68-w2tfs                           1/1     Running            0              12h
      lightning-api-988b7cbb6-nhwnk                         1/1     Running            0              12h
      lightning-gui-7f9b6bbb4d-h52bn                        1/1     Running            0              179mlightning-server-5439ec967ddf56d2-exec-1              0/1     Pending            0              174m
      lightning-server-5439ec967ddf56d2-exec-2              0/1     Pending            0              174m
      lightning-server-5439ec967ddf56d2-exec-3              0/1     Pending            0              174m
      lightning-server-driver                               1/1     Running            0              13h
      lightning-solr-555664b544-fpwlc                       1/1     Running            0              12h
      lightning-zeppelin-7975969f5f-fl6mj                   1/1     Running            0              12h
      ollama-6977d564d8-l495p                               1/1     Running            0              68m
      opensearch-cluster-master-0                           1/1     Running            0              179m
      opensearch-cluster-master-1                           1/1     Running            0              179m
      os-opensearch-dashboards-57f49f85fc-zpfsn             1/1     Running            0              12h
      privateai-566df5d786-68f49                            1/1     Running            0              162m

      8. Create Lightning user account

      · Login to ‘lightning-server-driver’ pod

      kubectl exec --tty --stdin lightning-server-driver -n zetaris /bin/sh

      · Run the dev-account.sh script

      cd /home/zetaris/lightning/bin/

      ./dev-account.sh

      9. Launch Zetaris

      Please Open browser window (preferably chrome) from any computer that is connected to VM and can access the machine on which Zetaris is installed. 

      Use URL from route created earlier (e.g. http://zetaris-zetaris.apps.zetaris.iac.ssc) to launch Zetaris from browser window.