C2SIM Metrics¶
The C2SIM server exposes a /metrics endpoint that publishes runtime metrics using the OpenTelemetry standard.
Docker Test Environment¶
To support local development and metrics validation, the development Docker environment includes a complete observability stack.
Included Services
| System | Description |
|---|---|
| Prometheus | Open-source tool to scrapes the /metrics endpoint and stores time-series data in a database. |
| Grafana | Open-source visualization platform for exploring and displaying metric history |
This setup allows:
-
Verify that metrics are correctly exposed by the C2SIM server
-
Inspect real-time metric values
-
Analyze historical trends
-
Build dashboards for performance monitoring
Prometheus periodically collects metrics from the C2SIM /metrics endpoint, while Grafana provides a user-friendly web interface for querying and visualizing the collected data.
⚠ Development Use Only
This Docker setup is intended for development and testing purposes only.
Configuration¶
In docker\prometheus\prometheus.yml is configured that Prometheus should scrape the endpoint c2sim-server:7777 every 5 sec. Because the c2-sim server and Prometheus run the the same docker container this works with localhost and an internal port number. In docker\grafana\provisioning\datasources\datasource.yml a data source is created for Prometheus that uses Prometheus to query time serie-data. The Grafana dashboards are automatically loaded from the folder grafana\provisioning\dashboards.
Using Grafana dashboard¶
When the docker compose is started, the monitoring tools are automatically started. Use 'http:\\localhost:30001to openGrafana`:
By default the website is opened in readonly mode. Login as admin to modify the configuration.
In the file docker/.env the following configuration items apply:
| Config entry | Description | default |
|---|---|---|
| GRAFANA_PORT | The external port number for Grafana dashboard | 3000 |
| PROMETHEUS_PORT | Don't change, is for docker internally | 9090 |
| PASSWORD | The password for adminon Grafana. |
federates |
Javalin Micrometer plugin¶
The Javalin Micrometer plugin is used to collect and expose application metrics. This plugin is built on top of the Micrometer framework, which is primarily focused on exposing metrics to monitoring systems.
For broader observability capabilities—including metrics, logs, and traces—the OpenTelemetry framework is more suitable. OpenTelemetry provides a unified approach to collecting and exporting telemetry data across these domains.
The Javalin Micrometer plugin provides several metrics out of the box, such as Jetty server metrics, disk usage, processor statistics, and other system-level information. In addition to these built-in metrics, the MetricService class introduces custom C2SIM metrics specific to the application.. See C2SIM metrics
In the future, it may be possible to combine both libraries, using Micrometer for certain metrics while integrating Open Telemetry for a more comprehensive observability solution.
Metric datatype¶
In the Open Telemetry standard multiple datatypes for storing metric data are defined. The C2SIM server uses the following types:
| Type | Description |
|---|---|
| Counter | Is a metric that only increases over time (or resets to zero after a restart) |
| Gauge | Represents a value that can go up or down at any time. |
| History | A histogram measures the distribution of values by counting observations into configurable buckets. |
Metric attributes¶
Each metric can have zero, or multiple attributes (sometimes referred as tags or labels). The combination of all attributes in a metric is stored as a data series.
For the C2SIM-server the following attributes are added:
| Attribute (tag/label) | Description |
|---|---|
| c2sim.app | Fixed to value c2sim-server |
| c2sim.shared_session | The shared session name where the metric data was collected from. |
| c2sim.system_name | The system name of the system that is monitored. |
| c2sim.msg_kind | The category of the C2SIM message |
| c2sim.request_endpoint, c2sim.request_method, c2sim.request_status | All javalin endpoints the duration of the request is measured for each combination of method, endpoint and status |
| c2sim.error_kind | The category why request failed. |
Warning
Keep number of possible attribute combination low, or it will impact the performance of the database. For example: c2sim.shared_session (3 sessions) * c2sim.system_name (10 systems) * c2sim.msg_kind (5 categories) = 150 data series. Below ** 1000 ** is a safe number of data series.
Generate markdown from metric endpoint¶
The project porm2json can scrape a Prometheus client and dump the result as JSON. · With with linux tool jq this json can be converted to markdown notation.
docker run --rm \
--add-host=host.docker.internal:host-gateway \
prom/prom2json \
http://host.docker.internal:9999/metrics \
| jq -r '
sort_by(.name)
| group_by(.name)[]
| .[0] as $m
| ([ .[].metrics[].labels? | keys[] ] | unique) as $labels
| "## Metric \"" + $m.name + "\"\n\n"
+ "**Type:** " + ($m.type|ascii_downcase) + " \n"
+ "**Description:** " + ($m.help // "No description") + "\n"
+ (if ($labels|length)>0
then "\n**Labels:** " + ($labels | join(", ")) + "\n"
else ""
end)
' > markdown.md