What does a scrape config actually do?

Each scrape_configs entry tells Prometheus a job_name (a label attached to every metric it collects from that target) and a static_configs.targets list of host:port addresses to poll on an HTTP /metrics endpoint at the interval you set globally, unless overridden per-job.

Do I need the alerting block if I'm not using Alertmanager?

No, leave it off. The alerting.alertmanagers block only matters if you're running Alertmanager and want Prometheus to forward firing alerts to it; without it Prometheus still evaluates alerting rules (if you load any) but has nowhere to send notifications.

What is remote_write for?

remote_write streams every scraped sample to an external URL in addition to storing it locally, the mechanism used to feed long-term storage systems like Thanos, Cortex, Mimir, or a hosted Prometheus-compatible backend.

Does this validate that my targets are actually reachable?

No. It validates job_name characters, hostname shape, and that ports are numeric in range, but it can't verify a target is actually listening since that requires a live network probe from wherever Prometheus itself runs.

Prometheus Config Generator - Build prometheus.yml

Overview

Introduction

A prometheus.yml file is small but unforgiving: a scrape config nested one level wrong, or a targets list written as a string instead of a list, and Prometheus refuses to start or silently scrapes nothing.

This tool builds that file from a form, letting you add scrape jobs, wire up Alertmanager, and set a remote_write endpoint without hand-indenting YAML.

What Is Prometheus Config Generator?

A visual builder for prometheus.yml, Prometheus's main configuration file, covering the global block, any number of static-target scrape_configs, an optional alerting.alertmanagers block, and an optional remote_write entry.

It mirrors the structure Prometheus itself expects: global settings apply to every job unless a job overrides them, and scrape_configs is the list of jobs Prometheus will poll.

How Prometheus Config Generator Works

The global block is emitted first with your scrape_interval and evaluation_interval (how often Prometheus polls targets and how often it evaluates alerting/recording rules).

Each scrape config becomes one entry under scrape_configs, with its job_name and a static_configs.targets list built from the host:port you provide. If alerting is enabled, an alerting.alertmanagers.static_configs block is added pointing at your Alertmanager instance; if remote_write is enabled, a remote_write list with your URL is appended.

When To Use Prometheus Config Generator

Use it when standing up a new Prometheus instance and you need a starting prometheus.yml with a handful of scrape targets, for example node_exporter, cAdvisor, or an app's own /metrics endpoint.

It's also useful for quickly wiring in an Alertmanager or remote_write endpoint to an existing setup without re-deriving the exact YAML shape from memory.

Often used alongside Alertmanager Config Generator, Grafana Dashboard Generator and OpenTelemetry Collector Config Generator.

Features

Advantages

Produces correctly nested YAML for scrape_configs, the section most hand-written configs get wrong (targets as a string instead of a list is a very common mistake).
Validates job names, hostnames, and port ranges before generating output, catching typos before they reach a running Prometheus instance.
Covers alerting and remote_write, letting you sketch a complete config in one pass instead of assembling it from multiple docs pages.

Limitations

Only supports static_configs targets, not service discovery mechanisms like Kubernetes SD, EC2 SD, or Consul SD; add those manually if you need dynamic discovery.
Doesn't verify targets are reachable, only that job names, hostnames, and ports are well-formed.
Doesn't generate alerting or recording rule files, only the alerting.alertmanagers pointer; write your rules separately and reference them with rule_files.

Examples

Scraping node_exporter with Alertmanager wired up

Input

scrape_interval: 15s, evaluation_interval: 15s, job_name: node, target: localhost:9100, alerting: alertmanager at localhost:9093

Output

global:
  scrape_interval: 15s
  evaluation_interval: 15s

scrape_configs:
  - job_name: "node"
    static_configs:
      - targets: ["localhost:9100"]

alerting:
  alertmanagers:
    - static_configs:
        - targets: ["localhost:9093"]

The global block, one scrape job, and the alerting pointer are all emitted with correct nesting.

Best Practices & Notes

Best Practices

Give each scrape job a descriptive job_name (e.g. node, api-server) since it becomes a label on every metric that job produces, and is how you filter in PromQL later.
Keep scrape_interval at 15s-60s for most workloads; shorter intervals increase storage and CPU cost quickly on high-cardinality targets.
Point remote_write at a durable long-term-storage backend if you need retention beyond Prometheus's local TSDB window.

Developer Notes

The generator is a pure string-building function emitting a fixed 2-space YAML indentation, with no YAML library dependency. job_name is validated against /^[a-zA-Z0-9_-]+$/, hostnames against /^[a-zA-Z0-9_.-]+$/, and ports against a numeric range of 1-65535; targets are always emitted as a single-element YAML flow-sequence (targets: ["host:port"]) to match Prometheus's expected static_configs shape.

Prometheus Config Generator Use Cases

Bootstrapping a new Prometheus instance's config with a handful of scrape targets
Adding an Alertmanager pointer to an existing prometheus.yml
Wiring a remote_write endpoint to forward samples to a long-term storage backend

Common Mistakes

Writing a scrape target as a bare string instead of a YAML list under static_configs.targets, which Prometheus rejects.
Setting evaluation_interval far apart from scrape_interval, causing alerting rules to evaluate against stale data.
Enabling remote_write with an internal-only URL that isn't reachable from wherever Prometheus actually runs.

Tips

Use the Alertmanager Config Generator next to build the alertmanager.yml that this config's alerting block points at.
Use the OpenTelemetry Collector Config Generator if you'd rather have an OTel Collector scrape and forward metrics to Prometheus instead of scraping targets directly.

Prometheus Config Generator