prometheus_fdw 0.1.5

This Release
prometheus_fdw 0.1.5
Date
Status
Stable
Abstract
Postgres Foreign Data Wrapper for Prometheus Data
Description
Prometheus_fdw is an integration of Prometheus monitoring data into Postgres. It enables querying for Prometheus metrics directly within Postgres, bridging the gap between Prometheus monitoring and Postgres's robust database capabilities.
Released By
tembo
License
PostgreSQL
Resources
Special Files
Tags

Extensions

prometheus_fdw 0.1.5
Postgres Foreign Data Wrapper for Prometheus Data

Documentation

README
Practical example
basic_setup
Prometheus FDW basic demo
setup
Practical example
README
Prometheus FDW basic demo
index
index

README

Prometheus_fdw

Prometheus_fdw is an integration of Prometheus monitoring data into Postgres. It enables querying for Prometheus metrics directly within Postgres, bridging the gap between Prometheus monitoring and Postgresā€™s robust database capabilities. Static Badge PGXN version

Pre-requisistes

  • Install prometheus_fdw
  • (Optional) install pg_partman and pg_cron

Quick start

create extension prometheus_fdw;

Create the foreign data wrapper:

create foreign data wrapper prometheus_wrapper
  handler prometheus_fdw_handler
  validator prometheus_fdw_validator;

Create the server:

create server my_prometheus_server
  foreign data wrapper prometheus_wrapper
  options (
    base_url '<base prometheus url>');

Create Foreign Table:

CREATE FOREIGN TABLE IF NOT EXISTS metrics (
  metric_name TEXT,
  metric_labels JSONB,
  metric_time BIGINT,
  metric_value FLOAT8
  )
server my_prometheus_server
options (
  object 'metrics',
  step '10m'
);

Queries

To simply run the fdw and look at values

SELECT
  *
FROM metrics
WHERE
  metric_name='container_cpu_usage_seconds_total'
  AND metric_time > 1696046800 AND metric_time < 1696133000;

Examples

Please see the examples/ directory to find a basic example and a practical example. In the practical example, metrics are automatically synced into the database using pg_cron, and automatically expired using pg_partman. Performance is optimized using indexes and partitioning.