Laravel Scheduler Monitoring

The problem

Laravel's task scheduler runs from a single cron entry — `* * * * * php artisan schedule:run` — and that is exactly why scheduler failures are so easy to miss. If that one line stops running (a deploy that drops the crontab, a server reboot, a full disk), every scheduled command silently stops with it and nothing alerts you. Even when the scheduler is running, an individual `Schedule::command()` can throw, exit non-zero, hang past its window, or get silently skipped by `withoutOverlapping`. Laravel's built-in `emailOutputOnFailure` only fires when the command actually ran and failed; it can't tell you about a job that never started or a scheduler that died. The result is the classic blind spot: your nightly backup, queue prune, or data sync quietly stops happening and you only find out when something downstream breaks.

How Pakyas helps

Pakyas is execution-signal based: each scheduled task proves it ran by sending a signal, and Pakyas compares those signals against the task's expected schedule. Laravel ships first-class hooks for exactly this — `pingBefore()`, `thenPing()`, `pingOnSuccess()`, and `pingOnFailure()` — so you wire a Pakyas check into the scheduler definition itself, no wrapper scripts and no SSH-ing into the box to edit crontabs. When a command starts, finishes, or fails, the scheduler pings Pakyas, and Pakyas turns silence and failure into distinct, actionable states: Missing (the signal never arrived, so the task — or the whole scheduler — never ran), Late (it ran outside its expected window), Overrunning (it started but never reported completion), and Error (it explicitly signaled failure with a non-zero exit code). Because the check is defined in source control alongside the task, the monitoring travels with your deploy.

Set it up

1. 1

   Confirm the scheduler cron entry is in place

   # Laravel needs one cron entry per server to drive schedule:run
   * * * * * cd /path-to-your-project && php artisan schedule:run >> /dev/null 2>&1

   This single line runs every minute and dispatches all your due tasks. If it ever stops, Pakyas catches the resulting silence as a Missing check (see the optional heartbeat snippet below).

2. 2

   Create a check in Pakyas and copy its ping URL

   # Sign up at https://pakyas.com and create a check for your scheduled task.
   # You get a unique ping URL using the check's public_id (a UUID):
   #
   #   start:    https://ping.pakyas.com/{public_id}/start
   #   success:  https://ping.pakyas.com/{public_id}
   #   fail:     https://ping.pakyas.com/{public_id}/fail
   #
   # Replace {public_id} below with the UUID Pakyas gives you.

   GET or POST both work. The bare URL signals success; append /start when the task begins and /fail on a non-zero exit.

3. 3

   Wire the ping into your scheduled task

   <?php
   
   use Illuminate\Support\Facades\Schedule;
   
   // Store the check's base URL in config/env so it isn't hard-coded.
   // Add your own key (e.g. PAKYAS_BACKUP_URL) to .env and config/services.php.
   $url = config('services.pakyas.backup_url'); // "https://ping.pakyas.com/{public_id}"
   
   Schedule::command('backup:run')
       ->dailyAt('02:00')
       ->withoutOverlapping()
       ->pingBefore($url . '/start')   // task is starting
       ->pingOnSuccess($url)           // exit code 0
       ->pingOnFailure($url . '/fail'); // non-zero exit code

   pingBefore / pingOnSuccess / pingOnFailure are built-in Laravel scheduler methods — no extra package required. Define this in routes/console.php (Laravel 11+) or your scheduler closure.

4. 4

   Optional: monitor the scheduler heartbeat itself

   <?php
   
   use Illuminate\Support\Facades\Http;
   use Illuminate\Support\Facades\Schedule;
   
   // A separate Pakyas check that pings every minute proves schedule:run is alive.
   // Supply your own scheduler check URL via config/env (a second check).
   $schedulerUrl = config('services.pakyas.scheduler_url');
   
   Schedule::call(fn () => Http::get($schedulerUrl))
       ->name('pakyas-scheduler-heartbeat')
       ->everyMinute();

   Set this Pakyas check's expected period to 1 minute. If the crontab dies, the heartbeat stops and Pakyas marks it Missing — catching the failure mode emailOutputOnFailure can never see.

Pricing

Pakyas is free for up to 10 checks ($0, no credit card) — enough to monitor the scheduler heartbeat plus your most important commands. Paid tiers add more checks and channels: Developer at $9/mo, Pro at $29/mo, and Business at $99/mo.

See the full breakdown on the pricing page.