Azure Deploy Aurelius Atlas

Getting started

Welcome to the Aurelius Atlas solution powered by Apache Atlas! Aurelius Atlas is an open-source Data Governance solution, based on a selection of open-source tools to facilitate business users to access governance information in an easy consumable way and meet the data governance demands of the distributed data world.

Here you will find the instillation instructions and the required setup of the kubernetes instructions, followed by how to deploy the chart in different namespaces.

Installation Requirements

This installation assumes that you have:

  • a kubernetes cluster running with 2 Node of CPU 4 and 16GB
  • Chosen Azure Cli installed
  • kubectl installed and linked to Azure Cli

Further you need the helm chart to deploy all services from https://github.com/aureliusenterprise/Aurelius-Atlas-helm-chart

Required Packages

The deployment requires the following packages:

  • Certificate Manager
    • To handel and manage the creation of certificates
    • Used in demo: cert-manager
  • Ingress Controller
    • Used to create an entry point to the cluster through an external IP.
    • Used in demo: Nginx Controller
  • Elastic
  • Reflector
    • Used to reflect secrets across namespaces
    • Used in demo to share the DNS certificate to different namespace

The steps on how to install the required packages

1. Install Certificate manager

Only install if you do not have a certificate manager. Please be aware if you use another manger, some commands later will need adjustments. The certificate manager here is cert-manager.

helm repo add jetstack https://charts.jetstack.io
helm repo update
helm install  cert-manager jetstack/cert-manager   --namespace cert-manager   --create-namespace   --version v1.9.1 --set installCRDs=true

2. Install Ingress Nginx Controller

Only install if you do not have an Ingress Controller.

helm repo add ingress-nginx https://kubernetes.github.io/ingress-nginx
helm repo update
helm install nginx-ingress ingress-nginx/ingress-nginx --set controller.publishService.enabled=true --set controller.service.annotations."service\.beta\.kubernetes\.io/azure-load-balancer-health-probe-request-path"=/healthz

It is also possible to set a DNS label to the ingress controller if you do not have a DNS by adding --set controller.service.annotations."service\.beta\.kubernetes\.io/azure-dns-label-name"=<label>

3. Install Elastic

kubectl create -f https://download.elastic.co/downloads/eck/2.3.0/crds.yaml
kubectl apply -f https://download.elastic.co/downloads/eck/2.3.0/operator.yaml

4. Install Reflector

helm repo add emberstack https://emberstack.github.io/helm-charts
helm repo update
helm upgrade --install reflector emberstack/reflector

Azure DNS Label

In Azure, it is possible to apply a DNS label to the ingress controller, if you do not have a DNS.

Edit the ingress controller deployment (if not set upon installation)

helm upgrade nginx-ingress ingress-nginx/ingress-nginx --reuse-values --set controller.service.annotations."service\.beta\.kubernetes\.io/azure-dns-label-name"=<label>

Save and exit. Resulting DSN will be <label>.westeurope.cloudapp.azure.com

Put ssl certificate in a Secret

Before you start, update zookeeper dependencies:

cd charts/zookeeper/
helm dependency update

Define a cluster issuer

This is needed if you installed cert-manager from the required packages.

Here we define a CLusterIssuer using cert-manager on the cert-manager namespace

  1. Move to the home directory of the chart helm-governance

  2. Uncomment templates/prod_issuer.yaml.

  3. Update the {{ .Values.ingress.email_address }} in values.yaml file and create the ClusterIssuer with the following command

    helm template -s templates/prod_issuer.yaml . | kubectl apply -f -
    
  4. comment out prod_issuer.yaml in templates Check that it is running:

    kubectl get clusterissuer -n cert-manager
    
  5. It is running when Ready is True.

    ../../_images/letsencrypt.png

Create ssl certificate

This is needed if you installed cert-manager from the required packages.

  1. Assumes you have a DNS linked to the external IP of the ingress controller

  2. Move to the home directory of the chart helm-governance

  3. Uncomment templates/certificate.yaml

  4. Update the values.yaml file {{ .Values.ingress.dns_url}} to your DNS name

  5. Create the certificate with the following command

    helm template -s templates/certificate.yaml . | kubectl apply -f -
    
  6. Comment out certificate.yaml in templates.

  7. Check that it is approved.

    kubectl get certificate -n cert-manager
    

It is running when Ready is True

../../_images/cert_aurelius_dev.png

Deploy Aurelius Atlas

  1. Create the namespace
kubectl create namespace <namespace>
  1. Update the values.yaml file
    • {{ .Values.keycloak.keycloakFrontendURL }} replace it to your DNS name
    • {{ .Values.kafka-ui. ... .bootstrapServers }} edit it with your <namespace>
    • {{ .Values.kafka-ui. ... .SERVER_SERVLET_CONTEXT_PATH }} edit it with your <namespace>
  2. Deploy the services
cd Aurelius-Atlas-helm-chart
helm dependency update
helm install --generate-name -n <namespace>  -f values.yaml .

Users with Randomized Passwords

In the helm chart 5 base users are created with randomized passwords stored as secrets on kubernetes.

The 5 base users are:

  1. Keycloak Admin User
  2. Atlas Admin User
  3. Atlas Data Steward User
  4. Atlas Data User
  5. Elastic User

To get the randomized passwords out of kubernetes there is a bash script get_passwords.

./get_passwords.sh <namespace>

The above command scans the given <namespace> and prints the usernames and randomized passwords as follows:

keycloak admin user pwd:
username: admin
vntoLefBekn3L767
----
keycloak Atlas admin user pwd:
username: atlas
QUVTj1QDKQWZpy27
----
keycloak Atlas data steward user pwd:
username: steward
XFlsi25Nz9h1VwQj
----
keycloak Atlas data user pwd:
username: scientist
PPv57ZvKHwxCUZOG
==========
elasticsearch elastic user pwd:
username: elastic
446PL2F2UF55a19haZtihRm5
----

Check that all pods are running

kubectl -n <namespace> get all # check that all pods are running

Aurelius Atlas is now accessible via reverse proxy at <DNS-url>/<namespace>/atlas/