What is the scrape_interval?

It is how often Prometheus pulls metrics from each target. The global value applies everywhere unless a job overrides it; 15 seconds is a common default balancing resolution against storage and load.

How do I point Prometheus at Alertmanager?

The alerting block lists Alertmanager instances under static_configs targets. Prometheus evaluates your alert rules and pushes firing alerts to those addresses, which then handle deduplication, grouping, and routing to receivers.

What goes in rule_files?

rule_files is a list of file globs containing recording and alerting rules, such as alerts/*.yml. Prometheus loads and evaluates them every evaluation_interval, separately from the scrape configuration.

Can a single job scrape many targets?

Yes. Each job's static_configs targets list can hold many host:port endpoints that share the same job label, metrics path, and scheme. This builder accepts a comma or space separated list per job.

When should a job override the global interval?

Override it when a specific exporter is expensive to scrape or needs higher resolution than the rest. Set a per-job scrape_interval to scrape that target more or less often without changing the global default.

Prometheus Scrape Config Builder

Email me this result

Get this tool's output sent to your inbox, plus one useful tool a week. No spam, unsubscribe any time.

A scrape config Prometheus will actually load

Prometheus is only as useful as its prometheus.yml, and a single mis-indented line can stop the whole server starting. This builder produces a valid config from a form: global intervals, an Alertmanager target, rule-file globs, and as many scrape jobs as you need — each with its own path, scheme, and target list.

How it works

The output follows Prometheus’ configuration schema. The global block sets scrape_interval and evaluation_interval. An optional alerting block lists Alertmanager endpoints under static_configs, and rule_files holds your alert and recording-rule globs. Under scrape_configs, each job emits a job_name, an optional metrics_path (only when it differs from the default /metrics), an optional scheme for HTTPS, an optional per-job scrape_interval, and a static_configs targets list parsed from your comma- or space-separated entries. Proper two-space YAML indentation is handled for you so the file loads on first try.

Tips and example

Always include a job that scrapes Prometheus itself (localhost:9090) so you can monitor the monitor.
Use a per-job scrape_interval for heavy exporters — scraping a slow endpoint every 15 seconds can overload it.
Keep alert and recording rules out of this file and in the globs under rule_files; that separation keeps the config readable.
Validate the result with promtool check config prometheus.yml before reloading in production.