CODEWITHSUNDEEP
laravelcentossupervisord
Matthew Weeks
Matthew Weeks

Reputation: 1028

Laravel notification jobs run multiple times by supervisor

So I have a laravel project running in production, and using supervisor to handle jobs and notifications.

The problem is, after supervisor runs for a couple hours, it starts to send duplicate notifications, and it seems that the longer it runs, the more times a notification will be sent out (It's happened where the same notification gets sent out 4 times).

My notification looks like the following:

class MessageNotification extends Notification implements ShouldQueue
{
    use Queueable;

    public function __construct($opts = [])
    {
        $this->connection = 'database';
        $this->queue = 'sendMail';
    }

    public function via($notifiable)
    {
        return ['mail', 'database', 'broadcast'];
    }

    public function toMail($notifiable)
    {
        $mesg = (new MailMessage)
            // ... defining message stuff

        return $mesg;
    }

    public function toArray($notifiable)
    {
        return [
          // .. defining array stuff
        ];
    }

    public function toBroadcast($notifiable){
        return (new BroadcastMessage($this->toArray($notifiable)));
    }
}

and my queue config looks like the following:

return [
    'default' => env('QUEUE_CONNECTION', 'sync'),

    'connections' => [

        'sync' => [
            'driver' => 'sync',
        ],

        'database' => [
            'driver' => 'database',
            'table' => 'jobs',
            'queue' => 'sendMail',
            'retry_after' => 90,
        ],
    ],


    'failed' => [
        'database' => env('DB_CONNECTION', 'mysql'),
        'table' => 'failed_jobs',
    ],

];

And my supervisor config is the following:

[program:laravel-worker]
command=php /app/artisan queue:work database --queue=sendMail --sleep=3 --tries=3 --timeout=75
process_name=%(program_name)s_%(process_num)02d
numprocs=8
priority=999
autostart=true
autorestart=true
startsecs=1
startretries=3
user=yastechco
redirect_stderr=true
stdout_logfile=/app/worker.log

This doesn't happen running in development on my local machine.

If it matters, the production server is running on CentOS 7.6.1810 and Supervisord 3.1.4

Upvotes: 0

Views: 988

Answers (1)

Matthew Weeks
Matthew Weeks

Reputation: 1028

I ended up configuring redis as my queue adapter, and it appears to have fixed the problem

Upvotes: 2

Related Questions