Skip to main content Link Menu Expand (external link) Document Search Copy Copied
vPW

Getting Started with Helm

1. Usage

  • Helm must be installed to use the charts. Please refer to Helm’s documentation to get started.

  • Once Helm has been set up correctly, add the repo as follows:

    helm repo add viadee https://viadee.github.io/charts

  • If you had already added this repo earlier, run helm repo update to retrieve the latest versions of the packages. You can then run helm search repo viadee to see the charts.

  • To install the chart:

    helm install my-<chart-name> viadee/<chart-name>

  • To uninstall the chart:

    helm delete my-<chart-name>

2. Available Charts

The following charts are available:

It is recommended to install the vpw-chart, since as an umbrella chart it combines all vpw components (vpw-analyzer-chart, vpw-pipeline-chart, vpw-frontend-chart, vpw-polling-client) and therefore allows an easy deployment.

3. Configuration

The umbrella chart only contains the vpw components (vpw-analyzer, vpw-pipeline, vpw-frontend, vpw-polling-client). Therefore, you also need to deploy the following components on your own:

  • process engine from which the polling client can poll process information.
  • kafka with zookeeper to stream the process information from which the pipeline polls the process information.
  • elasticsearch to which the pipeline and the analyzer can connect.
  • database to which the analyzer can connect.

3.1 Process engine with polling client:

  • If you do not have a deployed instance of a process engine in your kubernetes cluster running, you need to deploy one:

  • Adjust the Values.yaml for the umbrella helm chart in the polling client section accordingly to the used process engine (here Camunda BPM Platform):

    vpw-polling-client.environment.secret.camundaPassword: "demo"

    vpw-polling-client.environment.configMap.camundaRestUrl: http://camunda-bpm-platform:8080/engine-rest/

    vpw-polling-client.environment.configMap.username: "demo"

3.2 Kafka with polling client and pipeline:

  • If you do not have an apache kafka instance deployed, you need to deploy one:

  • Adjust the Values.yaml for the umbrella chart in the polling client and pipeline section via connecting to the exposed service of kafka:

    vpw-polling-client.environment.configMap.kafkaBootstrapServers: http://kafka-zookeeper-kafka-bootstrap:9092

    vpw-pipeline.environment.configMap.kafka_bootstrap_servers: http://kafka-zookeeper-kafka-bootstrap:9192

    • Notice 0: You can find the custom resource definition used for our deployment in our charts repository.
    • Notice 1: The Kafka address depends on the selected Kafka configuration and may therefore differ from the address shown here.
    • Notice 2: The Kafka port differs for polling client and pipeline, since the polling client puts information on the kafka stream, and the pipeline consumes them. Refer to the kafka documentation for detailed information.

3.3 Elasticsearch with pipeline and analyzer:

  • If you do not have an elasticsearch instance running in you cluster, you need to deploy one.
  • You could use the following helm chart elastic chart
  • Alternatively, you can use the Elasticsearch operator. This procedure is also our recommendation and is briefly described below.
    • After deploying the operator, apply the custom resource definition to create an elasticsearch cluster.
    • You can find the custom resource definition used for our deployment in our charts repository.

    • Adjust the Values.yaml for the umbrella chart in the pipeline and analyzer section to connect to the elasticsearch instance:

      vpw-pipeline.environment.configMap.spring_elasticsearch_rest_uris: "http://elasticsearch-es-http:9200"

      vpw-analyzer.environment.configMap.spring_elasticsearch_rest_uris: "http://elasticsearch-es-http:9200"

      • Notice 0: You also need to set the user, given by the elasticsearch instance (or create a new user), for the analyzer and pipeline as plaintext or by referring to an external secret.
      • Notice 1: The elastic address depends on the selected elastic configuration and may therefore differ from the address shown here.

3.4 Database with analyzer

  • The analyzer requires a database to connect, e.g. postgresql.
  • Here it is also recommended using an operator to deploy a database instance.
  • Our custom resource definition used for our deployment can be found in our charts repository.

  • Adjust the Values.yaml for the umbrella chart in the analyzer section accordingly to your database settings:

    vpw-analyzer.environment.secret.postgresql_pwd: "vpw"

    vpw-analyzer.environment.configMap.postgresql_host: "analyzer-postgres-postgresql"

    vpw-analyzer.environment.configMap.postgresql_port: "5433"

    vpw-analyzer.environment.configMap.postgresql_databasename: "vpw"

    vpw-analyzer.environment.configMap.postgresql_user: "vpw"

3.5 Ingress Configuration and Basic Authentication

  • By default, ingress is enabled with some placeholder values (vpw-frontend.ingress.). You need to insert your domain and tls information so that the frontend is accessible from outside the cluster.
  • Alternatively, you can “port-forward” the service port of the frontend to your localhost.
  • By default, the nginx of the frontend is secured via basic authentication (demo:demo). This can be changed and is explained here.

4. Deploy Script

We provide a (static) script to deploy all components together in a kubernetes cluster. Please note, that this script & resulting deployment is not suitable for production environments as it serves only for testing and developing. When running the script, the namespace that will be created in the cluster must be specified via the command line. Please note, that we recommend to use the namespace “vpw”, because some connection names assume that the namespace is called in that way (see e.g. .kafka_bootstrap_servers). You also need to be on the same folder level as the “kafka-ephemeral-single.yaml” and “postgresql-secret.yaml” files.

sh .\k8_deploy_vpw_withComponents_script.sh "vpw"