Download the PHP package webrek/laravel-health-ui without Composer

Laravel Health UI

A production health dashboard for Laravel. Pluggable checks (database, cache, disk, external services), a JSON endpoint your uptime monitor can poll, and a clean status page — all from a single route.

Quickstart

Visit /health to see the status page, or poll it as JSON:

The endpoint returns 200 while the app is healthy and 503 the moment a check fails hard — exactly what an uptime monitor needs (Pingdom, UptimeRobot, a Kubernetes liveness probe).

What you get over Laravel's /up

Laravel's built-in /up route answers a single question: did the framework boot? That tells you nothing about whether your database is reachable, whether your cache is responding, whether the disk is filling up, or whether a third-party API you depend on is down. This package runs real checks, aggregates them, and shows you which one is sick — on a status page and as machine-readable JSON.

Status model

Each check returns one of three statuses, and the overall status is the worst of them:

Warnings keep the endpoint at 200 so you aren't paged for degraded-but-serving conditions — only hard failures flip it to 503.

Built-in checks

Enable and tune them in config/health-ui.php:

• database — runs select 1 on the connection.
• cache — writes a value and reads it back.
• disk_space — warns/fails when it crosses the configured used-percentage thresholds.
• debug_mode — warns when APP_DEBUG is enabled (a common slip in production).
• http — pings each external dependency and expects a 2xx.
• queue_failed_jobs — warns/fails as the failed-jobs backlog grows.
• schedule — fails when the scheduler heartbeat goes stale (see below).
• migrations — warns when there are migrations committed but not run here.
• certificates — warns before a host's TLS certificate expires.

Scheduler heartbeat

The schedule check reads a timestamp your scheduler keeps fresh. Add a frequent task that stamps it:

The check then fails if that heartbeat is older than max_age_minutes.

Writing your own check

Implement the Check contract. Return a Result, or simply throw an exception — the checker turns an exception into a failed result automatically.

Register it (e.g. in a service provider):

Command line

Run the checks from the CLI or a cron — it exits with a non-zero code when unhealthy, so it plugs straight into deployment gates and alerting:

Configuration highlights

Publish the config and views to customize them:

The endpoint can reveal internal state. In production, put it behind middleware (a token, a signed URL, or an internal network restriction).

Requirements

Testing

Contributing

See CONTRIBUTING.md.

Security

Please review the security policy before reporting a vulnerability.

License

The MIT License (MIT). See LICENSE.