Skip to content

Latest commit

 

History

History
108 lines (81 loc) · 3.55 KB

README.md

File metadata and controls

108 lines (81 loc) · 3.55 KB
Raw

License CI Pipeline Workflow

Prombox

Sandbox environment for editing and testing prometheus configuration on the fly.

Motivation

Prometheus users often have trouble configuring new targets or creating alerts and recording rules because getting the YAML syntax exactly right can be a struggle. When running Prometheus locally this isn't as big of a challenge because you have direct access to the config file and can make changes quickly. Once Prometheus is deployed into a production environment, however, this becomes challenging because you don't have the ability to re-create the same combination of configuration and environment.

This is especially challenging when you're taking advantage of some service discovery methods provided by Prometheus, like AWS EC2 discovery. You may be able to discover your instance hosts via the AWS EC2 API, but accessing the endpoint your application exposes metrics on may not work due to firewall rules. Prometheus needs to live within your AWS VPC or private network in order for the metric scraping to occur successfully.

For these reasons, we built "Prombox", a lightweight UI that allows users to build the Prometheus config file and alerts/recording rules. This could be deployed within your environment (AWS VPC, etc) along with an instance of Prometheus that mimics your production instance. This would create a "sandbox" that allows you to "play" with new rules and configuration options to see how they would work.

Build and Run Docker image

Build Prombox

make build

Build Docker Image

docker build -t prombox -f build/Dockerfile .

Run docker-compose

docker-compose up

Access UI

The Prombox UI is accessible at http://localhost:3000

The Prometheus UI is accessible at http://localhost:9090

Run and Test Locally

Set Environment Variables

Required

export PROMETHEUS_ADDRESS=http://localhost:9090
export PROMETHEUS_FRAME_ADDRESS=http://localhost:9090
export PROMETHEUS_CONFIG=$(pwd)/prometheus/prometheus.yml

Optional

export PORT=3000 (default: 3000)
export CORS_ALLOW_ORIGIN=http://localhost:8080 (default: empty)

Run

  1. Run Prometheus at :9090 (using configuration in /prometheus/prometheus.yml)

    docker run \
    -p 9090:9090 \
    --mount type=bind,source="$(pwd)"/prometheus/prometheus.yml,target=/etc/prometheus/prometheus.yml \
    prom/prometheus \
    --web.enable-lifecycle \
    --config.file=/etc/prometheus/prometheus.yml \
    --storage.tsdb.path=/prometheus \
    --web.console.libraries=/etc/prometheus/console_libraries \
    --web.console.templates=/etc/prometheus/consoles

  2. Run API at :3000

    export PROMETHEUS_ADDRESS=http://localhost:9090
export PROMETHEUS_FRAME_ADDRESS=http://localhost:9090
export PROMETHEUS_CONFIG=$(pwd)/prometheus/prometheus.yml
export CORS_ALLOW_ORIGIN=http://localhost:8080
make run-api-dev

  3. Run ui with hot-reloading at :8080

    make run-ui-dev

Lint

Linting go (go fmt + go vet + golint) and vue (npm lint) source code

make lint

Test

Unit Tests

Run unit tests for ui

make test-ui

Run unit tests for go

make test-api

Run unit tests for both ui and go

make test