promls : Explore Prometheus metrics

promls lets you explore prometheus metrics for your services. Target the metrics endpoint of your service and search for relevant metrics.

Installation

The promls cli is available in a Python package. You can install it directly with pipx:

pipx install alpacloud-promls

Usage

CLI

promls has cli modes. Filter for metrics by name, path, or any field.

> promls name --filter cpu http://localhost:9402/metrics
process_cpu_seconds_total (counter) Total user and system CPU time spent in seconds.

You can output the metrics in several formats:

• flat: a compact list, useful for humans

  process_cpu_seconds_total (counter) Total user and system CPU time spent in seconds.
  

• json: for scripting

  {
  "process_cpu_seconds_total": {
      "name": "process_cpu_seconds_total",
      "help": "Total user and system CPU time spent in seconds.",
      "type": "counter"
    }
  }
  

• tree: use the words as nodes in a tree

  process
    cpu
      seconds
        total : process_cpu_seconds_total (counter) Total user and system CPU time spent in seconds.
  

• full: like the prometheus metric format

  # HELP process_cpu_seconds_total Total user and system CPU time spent in seconds.
  # TYPE process_cpu_seconds_total counter
  process_cpu_seconds_total
  

TUI

Interactively filter and explore the metrics.

Bibliography

• Prometheus metric format : Old grammar for Prometheus metrics
• Prometheus metric format : the actual grammar