Phobos + Prometheus InfluxDB Dashboard for Grafana
- The overall state of the cluster - the number of members of each status, each role type, and the total number of unreachable nodes over the past arbitrary window of time.
- Error rates and logging activity across the cluster - what do the global logging rates look like and which cluster nodes and logging which types of errors over how much time?
- Total actor counts, start / stop rates, crash rates, and message processing - this paints a complete picture of how many actors are alive throughout the cluster, what types of messages are processing, how are actors churning (starts / stops,) and what types of exceptions are being thrown by which actor types?
If you want to try this dashboard out without having to install Phobos into your own application yet, clone our https://github.com/petabridge/Petabridge.Phobos.Web.InfluxDb repository and build + run it locally using your Phobos NuGet Feed.
To install this dashboard, you will need a InfluxDB data source connected to your local Grafana instance.
This dashboard was built for InfluxDB 1.8+ data sources using InfluxQL - this should also be supported in InfluxDB 2.0+ but if you run into an issue please file an issue in the
phobos-dashboards repository here: https://github.com/petabridge/phobos-dashboards/issues
Once you have that up and running, you can install the Akka.NET Cluster + Phobos Metrics (Influx Data Source) dashboard via Grafana's Dashboard Import dialog and you'll be able to select your InfluxDB data source during that import dialog.
Once you've successfully imported the dashboard, you should be able to see the following graphs: