Prometheus — Monitoring

It is a CNCF-standard monitoring system that collects server and application metrics, stores them as time series, and sends alerts.

Basic Information

What Is This Program?

Prometheus is an open-source monitoring system that periodically receives numeric metrics from servers, containers, and applications and stores them in a time-series database. It collects not only system metrics such as CPU and memory, but also business metrics that apps expose directly.

As a CNCF graduated project, it has become the standard in container and Kubernetes environments. You can query it through the built-in UI, and dashboards are often used in combination with Grafana. Alerts are defined with the built-in rule language (PromQL) and sent to Alertmanager.

What Problems Does It Solve?

• In server monitoring, when you need to see CPU, disk, and process status over time, storing metrics as time series enables trend and anomaly detection.
• In incident response, when you need an alert when a threshold is exceeded, you can define alert conditions with PromQL rules to send notifications.
• In container environments, when you need to collect metrics at the pod or node level, service discovery automatically registers targets.

Key Features

Pull-based Collection

Periodically scrapes the /metrics endpoints exposed by targets

• Use Case: HTTP metric collection
• Result: Simple target registration and removal
• Limitations: Environments where pull is difficult require Pushgateway.

PromQL Queries

A native language for querying and aggregating time-series data

• Use Case: Graphs and alert rules
• Result: Label-based aggregation and rate calculation
• Limitations: PromQL has a learning curve.

Alert Rules and Alertmanager

Generates alerts based on thresholds and anomaly patterns, then groups and deduplicates them

• Use Case: Failure notifications
• Result: Delivered via chat, email, and PagerDuty
• Limitations: Too many alerts can cause notification fatigue.

Service Discovery

Automatically discovers targets through DNS, Kubernetes, Consul, etc.

• Use Case: Dynamic environments
• Result: Manage targets without manual registration
• Limitations: Discovery configuration errors can lead to missed collections.

Installation and Execution

Linux (Binary)

Download the appropriate architecture binary from the official download page and run it.

wget https://github.com/prometheus/prometheus/releases/download/v3.15.0/prometheus-3.15.0.linux-amd64.tar.gz
tar xvf prometheus-3.15.0.linux-amd64.tar.gz
cd prometheus-3.15.0.linux-amd64
./prometheus --config.file=prometheus.yml

Linux (systemd)

After installation, keep it running using a service unit file.

sudo useradd --no-create-home --shell /bin/false prometheus
sudo cp prometheus /usr/local/bin/
sudo cp prometheus.yml /etc/prometheus/
# After creating /etc/systemd/system/prometheus.service
sudo systemctl enable --now prometheus

Docker

Run it quickly with the official image.

docker run -d -p 9090:9090 \
  -v ./prometheus.yml:/etc/prometheus/prometheus.yml \
  prom/prometheus

Real-World Usage Examples

Node Metric Collection

Install node_exporter and register it as a target in prometheus.yml.

$ ./node_exporter &
# prometheus.yml
scrape_configs:
  - job_name: 'node'
    static_configs:
      - targets: ['localhost:9100']

PromQL Queries

Query metrics with PromQL in the UI (localhost:9090).

node_cpu_seconds_total{mode="idle"}
100 - (avg by(instance)(rate(node_cpu_seconds_total{mode="idle"}[5m])) * 100)

Defining Alert Rules

Define alert rules in a file and apply them with reload.

groups:
- name: alerts
  rules:
  - alert: HighCpu
    expr: 100 - (avg by(instance)(rate(node_cpu_seconds_total{mode="idle"}[5m]))*100) > 85
    for: 5m

Advantages and Limitations

Advantages

• Free and open source (Apache-2.0); easy to deploy as a single binary
• Pull model and service discovery make it well suited for container environments
• PromQL and alert rule language enable flexible queries and notifications
• Integrates very well with visualization tools such as Grafana

Considerations Before Use

• Long-term, large-scale time series may require integration with Thanos, Cortex, etc. due to local storage limitations.
• Because of the pull model, targets behind NAT or firewalls may require workarounds such as Pushgateway.
• Writing PromQL and alert rules requires learning.

Comparing Similar Programs

Official Resources and Further Learning

• Official Website — https://prometheus.io/
• Official Documentation / Knowledge Base — https://prometheus.io/docs/
• Official Download — https://prometheus.io/download/
• Release Notes — https://github.com/prometheus/prometheus/releases
• Source / Development Resources — https://github.com/prometheus/prometheus

Version and supported environment information is based on the official site as of 2026-10-18. Installation commands may vary depending on your distribution and version, so please also check the official installation documentation.

Actual Execution Screenshot

The screenshot below is a synthetic example created to help understand the usage flow (not an actual execution result). The screenshot is a synthetic example and not an actual execution result.

!Prometheus example: server startup log, health check

Servers in actual operation

Source (HTML): https://games.aidebatehub.com/programs/apps/monitoring/prometheus.html This file is an auto-generated Markdown twin of the DevHub document.