LogScale exposes a variety of LogScale Metrics that can be used
to monitor its performance. If you use Prometheus, setting the
PROMETHEUS_METRICS_PORT configuration will open a TCP port
for Prometheus to scrape metrics from LogScale:
You can double-check that this is working by curling that port on your installation:
If it's working properly, you should see a long list of metrics returned.
You can use a fairly standard Prometheus scraper job for this. A very basic example Prometheus config for this is as follows:
global: scrape_interval: 1m scrape_timeout: 10s evaluation_interval: 1m scrape_configs: - job_name: humio bearer_token: scrape_interval: 30s scrape_timeout: 5s sample_limit: 1000 metrics_path: / scheme: http static_configs: - targets: ['10.0.0.10:8081', '10.0.0.11:8081', '10.0.0.12:8081'] labels: service: humio
The metrics need to be scraped from all LogScale nodes in your cluster. Prometheus will need access to whatever port you specify whether through a reverse proxy / load balancer or by directly hitting the LogScale nodes themselves.