Prometheus: Difference between revisions
From wikinotes
(→Notes) |
(→Notes) |
||
Line 59: | Line 59: | ||
= Notes = | = Notes = | ||
<blockquote> | <blockquote> | ||
Setup | == Setup == | ||
{| class="wikitable" | {| class="wikitable" | ||
|- | |- | ||
Line 70: | Line 70: | ||
|} | |} | ||
Configuration | == Configuration == | ||
{| class="wikitable" | {| class="wikitable" | ||
|- | |- | ||
Line 79: | Line 79: | ||
|} | |} | ||
Usage | == Usage == | ||
{| class="wikitable" | {| class="wikitable" | ||
|- | |- | ||
Line 88: | Line 88: | ||
|} | |} | ||
Syntax | == Syntax == | ||
{| class="wikitable" | {| class="wikitable" | ||
|- | |- |
Revision as of 16:32, 1 October 2022
A system monitoring tool.
Metrics are polled/persisted over REST,
Documentation
home https://prometheus.io/
Tutorials
overview https://sbcode.net/prometheus/
Routes
9090
default web-ui http://${IP_ADDR}:9090
web-ui http://${IP_ADDR}:9090/targets
web-gui (health of queried metrics)
Locations
${PREFIX}/etc/prometheus.yml
main configuration /var/db/prometheus
data dir
Quickstart
- prometheus polls HTTP endpoints for collected metrics
- these endpoints are called exporters, and they are generated/exposed using clients
- you can configure rules to persist data, or emit alerts when a metric fails
xdg-open localhost:9090 # prometheus-ui promtool check rules /path/to/example.rules.yml # validate rule amtool check-config alertmanager.yml # validate alertmanager config
Notes
Setup
prometheus install prometheus exporters prometheus alerters Configuration
prometheus configuration prometheus rules Usage
prometheus web ui prometheus http api Syntax
prometheus promql prometheus metric syntax
Integrations
Dashboards
grafana metric/graph dashboards