The activity tile of the API window is limited to the last 10 request statuses.
However, when an API is enabled for a dataset, metrics are gathered, and can be consumed
by the monitoring dashboard of your choice.
For more information about the metrics available, see Metrics available for the API.
Procedure
-
In your Remote Engine Gen2 directory, add the following configuration to the
default/nginx/nginx.conf file to make the metrics
endpoint visible:
location /metrics {
proxy_pass http://dataset-api-server/api/v1/metrics;
}
This configuration can be added under the
location /apis
element in the file. For
example:
events { }
http {
server {
listen 9102;
location /apis {
proxy_pass http://dataset-api-server/apis;
}
location /metrics {
proxy_pass http://dataset-api-server/api/v1/metrics;
}
}
}
-
Restart your Remote Engine Gen2 instance to apply the new configuration.
-
In the monitoring system of your choice, use the following URL with the correct
parameters to get the metrics for the APIs running on your Remote Engine Gen2:
protocol://remote_engine_host:DATASET_API_SHARING_EXTERNAL_PORT/metrics
Different parameters that make the URL
Parameter |
Value |
protocol |
http by default, or
https if it is configured. |
remote_engine_host |
Host where your Remote Engine Gen2 is installed. This is specific to your infrastructure, if
you do not know the host, contact your
administrator. |
DATASET_API_SHARING_EXTERNAL_PORT |
Value of the
DATASET_API_SHARING_EXTERNAL_PORT
parameter in the default/.env file in
the Remote Engine Gen2 directory. |
-
Configure your monitoring system as needed.
Example
With Prometheus, you can use the following configuration. For more
information, see the Prometheus documentation.
- job_name: 'dataset_api_sharing'
scrape_interval: 10s
metrics_path: /metrics
static_configs:
- targets: [ 'localhost:8080' ]
Results
Examples of Grafana dashboards which query Prometheus data for memory usage, number
of requests per API, or average latency for example.