Molicious and Minion help request

[Árpád Hajdú]

Hello,

I need some help about the interaction when Mojolicious and Minion working together.

I have the full Mojolicious code where I load the Minion as a plugin and using Pg as a backend. 

sub status {

        my $c = shift;

        $c->render(json =>  $c->app->minion->stats, status => 200 );

        return;

}

This is showing me the following output when the worker is not running:

{"active_jobs":0,"active_workers":0,"delayed_jobs":0,"enqueued_jobs":43,"failed_jobs":0,"finished_jobs":0,"inactive_jobs":7,"inactive_workers":0}

After I start the minion worker in a different terminal on the same host and code:

{"active_jobs":0,"active_workers":0,"delayed_jobs":0,"enqueued_jobs":43,"failed_jobs":0,"finished_jobs":0,"inactive_jobs":7,"inactive_workers":1}

Inactive worker number increases as expected. 

I am planning to use the following endpoint to post a costly long running background job:

sub rescan {

my $c = shift;

my $minion = $c->app->minion;

$minion->add_task( 'rescan' => sub{

my $job = shift;

$job->app->SCSICleanup($job);

});

$minion->enqueue('rescan');

$c->render(json => {Service => 'Rescan'}, status => 200 );

return;

}

After I called the rescan function I have the following job in the database: (select * from minion_jobs order by id desc limit 1 ;)

44;"[]";"2017-06-02 16:58:52.252232+01";"2017-06-02 16:58:52.252232+01";"";0;"";"";0;"";"inactive";"rescan";;"default";1;"{}"

The job remain inactive even if the worker is running.

In the development.log I have the following - worker restarted after code changes:

[Fri Jun  2 17:04:09 2017] [debug] Loading Plugin Helpers

[Fri Jun  2 17:04:09 2017] [debug] Loading Plugin Database

[Fri Jun  2 17:04:09 2017] [debug] Initialise database connection

[Fri Jun  2 17:04:09 2017] [debug] Database Version: PostgreSQL 9.5.7 on x86_64-pc-linux-gnu, compiled by gcc (Ubuntu 5.4.0-6ubuntu

1~16.04.4) 5.4.0 20160609, 64-bit

[Fri Jun  2 17:04:09 2017] [debug] Loading Plugin Minion

[Fri Jun  2 17:04:09 2017] [debug] Loading Plugin SCSI

[Fri Jun  2 17:04:09 2017] [debug] Loading Plugin Jobs

[Fri Jun  2 17:04:09 2017] [debug] Worker 4127 started

[Fri Jun  2 17:04:09 2017] [debug] Checking worker registry and job queue

I tried the following modification:

sub rescan {

my $c = shift;

my $minion = $c->app->minion;

$minion->add_task( 'rescan' => sub{

my $job = shift;

$job->app->SCSICleanup($job);

});

my id = $minion->enqueue('rescan');

$minion->job($id)->start;

$c->render(json => {Service => 'Rescan'}, status => 200 );

return;

}

After the modification the situation is exactly the same:

45;"[]";"2017-06-02 17:04:12.067348+01";"2017-06-02 17:04:12.067348+01";"";0;"";"";0;"";"inactive";"rescan";;"default";1;"{}"

Could you please help me what I am doing wrong or what I missed ? 

Regards,

Arpad

[Sebastian Riedel]

> Could you please help me what I am doing wrong or what I missed ?

Add your task during application startup.

--
Sebastian Riedel
http://mojolicio.us
http://github.com/kraih
http://twitter.com/kraih

[Sebastian Riedel]

Actually, it might be best if you start experimenting with the example
from the Minion documentation.

[Árpád Hajdú]

By adding the tasks in startup solved my issue ... Thanks.

I won't say the example is useful. It is a performance comparison only. It is not a real example.

[sri]

I won't say the example is useful. It is a performance comparison only. It is not a real example.

I have no idea what you're talking about, how is this not a real example?

    http://mojolicious.org/perldoc/Minion#DESCRIPTION

--

sebastian