Prometheus metrics for LandSandBoat
Clone this repository (as is or as a submodule) into the modules directory of your server.
git clone [email protected]:sruon/LandSandBoat-prometheus.git
# or
git submodule add [email protected]:sruon/LandSandBoat-prometheus.git
Modify the init.txt file in the root of the modules folder:
LandSandBoat-prometheus
The module can be configured by modifying the settings/map.lua.
This is not required if the default address and port work for you (127.0.0.0:9865)
xi.settings.map =
{
[...]
-- Address to bind the exposer on
PROMETHEUS_LISTEN_ADRESS = '0.0.0.0'
-- Port to bind the exposer on
PROMETHEUS_LISTEN_PORT = 2112
}Start the server and you should see this message in the logs
[map][error] Settings: Failed to look up key: map.PROMETHEUS_PORT, using default value: "0" (settings::get:157)
[map][warn] [prometheus] No listen port defined, listening on default port: 9865 (PrometheusModule::OnInit:74)
[map][error] Settings: Failed to look up key: map.PROMETHEUS_LISTEN_ADDRESS, using default value: "" (settings::get:157)
[map][warn] [prometheus] No listen address defined, listening on default address: 127.0.0.1 (PrometheusModule::OnInit:80)
[map][info] [prometheus] Exposer started. Visit http://127.0.0.1:9865/metrics (PrometheusModule::OnInit:86)
The metrics are exposed on the /metrics endpoint of the server. You can access them by visiting http://<ip>:<port>/metrics in your browser.
By default, this is http://127.0.0.1:9865/metrics
# HELP outgoing_packet_size_counter Counter for outgoing_packet_size
# TYPE outgoing_packet_size_counter counter
outgoing_packet_size_counter{packetType="37"} 672
outgoing_packet_size_counter{packetType="5E"} 180
outgoing_packet_size_counter{packetType="CC"} 176
# HELP packet_pushes_counter Counter for packet_pushes
# TYPE packet_pushes_counter counter
packet_pushes_counter{packetType="37"} 7
packet_pushes_counter{packetType="DF"} 5
packet_pushes_counter{packetType="D2"} 10
# HELP time_server_ticks_counter Counter for time_server_ticks
# TYPE time_server_ticks_counter counter
time_server_ticks_counter 230
# HELP zone_ticks_counter Counter for zone_ticks
# TYPE zone_ticks_counter counter
zone_ticks_counter{zoneName="Upper_Jeuno"} 26
# HELP char_zone_gauge Gauge for char_zone
# TYPE char_zone_gauge gauge
char_zone_gauge{zoneName="Upper_Jeuno"} 1
You can now configure any collector that supports Prometheus to scrape the metrics from your server and store them in your favorite datastore.
How to collect and visualize the metrics is out of the scope of this module, but you can find many resources online on how to do this.
An example showing how to use Grafana Alloy with Grafana Cloud is provided for convenience.
Examples show usage on Windows but should be extendable to other platforms with minor tweaks.
- Head over to Grafana Cloud and sign up for a free account
- Follow instructions until you get to your dashboard
- Click on the blue "Launch" button to access services
- Click on "Send Metrics" below "Prometheus" to get the URL and API key
- Keep this information handy for the next steps
Refer to Grafana Alloy's documentation for more information.
- Head over to the releases page
- Download the file named "alloy-installer-windows-amd64.exe.zip"
- Extract the zip file
- Run the installer
Grafana Alloy should now be running as a background service.
The debug UI should be available at http://localhost:12345/
Now we need to tell Grafana Alloy where to scrape the metrics from.
- Open the file
%PROGRAMFILES%\GrafanaLabs\Alloy\config.alloy- You may need to delete the existing file and replace it with a new one.
Use the following configuration, and edit the appropriate fields with the information obtained from Grafana Cloud.
prometheus.remote_write "grafanacloud" {
endpoint {
url = "https://prometheus-xxx.grafana.net/api/prom/push"
basic_auth {
username = "xxx"
password = "xxx"
}
}
}
prometheus.scrape "lsb" {
targets = [{
__address__ = "127.0.0.1:9865",
}]
scrape_interval = "10s"
forward_to = [prometheus.remote_write.grafanacloud.receiver]
}
Using the Service Manager, restart the Alloy service.
Head over to the Alloy debug UI and check the status of the scrape.
From Grafana Cloud, click the "Launch" button next to "Grafana".
After a couple moments, your Grafana instance should be ready.
Heading over to the "Explore" tab, you should be able to query the metrics scraped from your server.
Import dashboard.json into Grafana to create a Dashboard showcasing basic concepts and displaying basic metrics exported by this project.
Dashboards can be imported in the "New Dashboard" dialog.
From there you may want to do the following:
- Learn more Prometheus query language to get the most out of your metrics
- Create dashboards, mixing metrics as you like
- Set up alerts
- Tweak the scrape rate to submit metrics more frequently (default every minute)
- Support for multiple map processes
- Lua bindings to create arbitrary metrics
- Hooks into global Lua functions
- Incoming Packets
- Only a handful of labels are record at the time