Skip to content

METRICS.md

Latest commit

 

History

History
56 lines (44 loc) · 1.83 KB

METRICS.md

File metadata and controls

56 lines (44 loc) · 1.83 KB

Prometheus Metrics

The biliproxy application exposes Prometheus metrics at the /metrics endpoint. These metrics can be scraped by a Prometheus server and visualized in Grafana.

Exposed Metrics

All metrics are prefixed with biliproxy_.

Default Metrics

The application exports default Node.js metrics provided by prom-client, including:

  • biliproxy_process_cpu_user_seconds_total: Total user CPU time spent in seconds.
  • biliproxy_process_cpu_system_seconds_total: Total system CPU time spent in seconds.
  • biliproxy_process_resident_memory_bytes: Resident memory size in bytes.
  • biliproxy_nodejs_eventloop_lag_seconds: Lag of event loop in seconds.
  • ...and others.

Custom Metrics

biliproxy_http_request_duration_ms

  • Type: Histogram
  • Description: Duration of HTTP requests in milliseconds.
  • Labels:
    • method: HTTP method (e.g., GET, POST)
    • route: The route path (e.g., /x/web-interface/nav)
    • code: HTTP status code (e.g., 200, 404)
  • Buckets: 100, 300, 500, 700, 1000, 3000, 5000, 7000, 10000

biliproxy_http_response_bytes_total

  • Type: Counter
  • Description: Total number of bytes sent in responses.
  • Labels:
    • method: HTTP method
    • route: The route path
    • code: HTTP status code

Example PromQL Queries

Request Rate (Requests per second)

rate(biliproxy_http_request_duration_ms_count[1m])

Error Rate (Percentage of non-200 responses)

sum(rate(biliproxy_http_request_duration_ms_count{code!="200"}[1m])) / sum(rate(biliproxy_http_request_duration_ms_count[1m])) * 100

99th Percentile Request Latency

histogram_quantile(0.99, sum(rate(biliproxy_http_request_duration_ms_bucket[1m])) by (le))

Throughput (Megabits per second)

rate(biliproxy_http_response_bytes_total[1m]) * 8 / 1000 / 1000