Horizon

Level 1

Yes — that happens sometimes with Laravel Horizon if something in the setup isn’t right. A crash when Horizon starts usually points to one of these common issues:

Redis connection overload / wrong config
- Horizon uses Redis heavily. If your Redis connection details in config/database.php are wrong, or Redis is maxing out memory/CPU, Horizon can cause the server to choke.
- Check Redis logs (/var/log/redis/redis-server.log) and Laravel logs (storage/logs/laravel.log).
Supervisor misconfiguration (if you use it to keep Horizon alive)
- If Supervisor is restarting Horizon too fast (e.g., config loops), it can spike CPU and crash the box.
- Run supervisorctl status to check.
Horizon dashboard polling
- Horizon’s UI polls Redis for metrics — on a weak server (low memory or shared hosting), that can crash it.
PHP memory limit
- Horizon spawns multiple worker processes. If php.ini has a low memory limit, each worker eats memory and the server crashes.
Too many Horizon workers configured
- In config/horizon.php, check supervisor definitions. If you’ve set processes too high (like 50 on a small server), the server may run out of RAM instantly.

Check logs first:

tail -f storage/logs/laravel.log
tail -f /var/log/syslog
tail -f /var/log/redis/redis-server.log

Lower worker count in config/horizon.php:

'supervisors' => [
    'default' => [
        'connection' => 'redis',
        'queue' => ['default'],
        'balance' => 'auto',
        'maxProcesses' => 1,  // try small first
        'processes' => 1,
        'tries' => 3,
    ],
],

Restart Redis to clear stuck connections:
```
sudo systemctl restart redis
```
Clear Horizon cache:
```
php artisan horizon:clear
```
Test Horizon manually without Supervisor:
```
php artisan horizon
```
→ If it runs, the issue is probably Supervisor configuration.

In most cases, the culprit is too many Horizon workers for the server resources. Start with just 1 worker, confirm stability, then scale up gradually.

1 like