Grafana

Akka Insights provides pre-defined dashboards for Grafana, a leading open source software time series analytics.

Grafana integrates with several time-series databases, e.g. Prometheus, Elasticsearch, InfluxDB, etc.

This section describes how to download, install, and use the dashboards in Grafana.

Grafana dashboards

Akka Insights provides several predefined dashboard templates including:

Akka metrics : metrics for actors, remote actors, cluster, and more.
Akka events : actor events (such as dead letters or unhandled messages), cluster events, and more.
Akka HTTP server and client metrics.
Akka Stream core and extended metrics.
Stopwatch metrics : information about Stopwatches that have been programmatically added.
JVM information : heap and garbage collection information.
… and more.

Note

The dashboards have been tested with Grafana v4 and v5.

Download dashboards

The example dashboards are packaged in zip files to download:

Below you can see the actor and dispatcher dashboards:

Actors dashboard

Dispatchers dashboard

Running Grafana

The simplest way to get started with Grafana is to use the Grafana Docker image.

Login

Once Grafana is started you should be able to login (default user/password is admin/admin), and define datasource so Grafana.

Define a data source

Akka Insights provides dashboards for three different time-series databases; Prometheus, Elasticsearch, and Graphite. Please follow the links below for how how to configure them:

Note

The data source must be correctly named for the templates to work.

Prometheus - data source name to use: Cinnamon Prometheus
Elasticsearch - data source name to use: Cinnamon Elasticsearch
Graphite - data source name to use: Cinnamon Graphite

Import dashboards

First download dashboards for the data source you’re using, either Prometheus, Elasticsearch, or Graphite. Follow these steps to import dashboards.

Run your application

With Grafana running, a defined data source, and imported dashboards it is time to run the application, have it configured to export into the time-series database of choice, and the graphs will be populated with data points.