Once your pipeline is running, the natural next question is "how is it doing?" The nature of streaming pipelines is that they quietly just do their work; problems only come to light if they start to slow down the flow of data. Ideally, you can detect issues before they impact the stream itself. To help you do this, Active(SP) has metrics reporting built into the framework. All you need to do it turn it on.

If you set the Helm property monitoring.prometheus.enabled to true when starting the pipeline, it becomes a Prometheus client, reporting metrics that integrate with your existing Prometheus and Grafana infrastructure. For example:

$ helm install mydatapipe voltdb/voltsp                   \
  --set-file streaming.licenseXMLFile=${MY_VOLT_LICENSE}  \
  --set image.repository=${MY_DOCKER_REPO}                \
  --set image.tag=mypipe--latest                          \
  --values test/src/main/resources/mydatapipe.yaml        \
  --set monitoring.prometheus.enabled=true

If you do not have an existing Prometheus infrastructure or if you simply want to quickly evaluate the pipeline performance without having to design your own dashboard, Volt provides a package containing Prometheus, Grafana, and a custom Grafana dashboard for reporting on Active(SP) pipelines. To start the management console with your pipeline, set management-console.enabled to true when you start the pipeline:

$ helm install mydatapipe voltdb/voltsp                   \
  --set-file streaming.licenseXMLFile=${MY_VOLT_LICENSE}  \
  --set image.repository=${MY_DOCKER_REPO}                \
  --set image.tag=mypipe--latest                          \
  --values test/src/main/resources/mydatapipe.yaml        \
  --set monitoring.prometheus.enabled=true                \
  --set management-console.enabled=true

Note that one copy of the management console can report on multiple pipelines. So you only need to start it with one of the pipelines. Alternatively you can leave it out of the pipeline startup entirely and start the management console separately with its own helm install command:

$ helm install mydataconsole voltdb/volt-stream/charts/management-console