Hi everyone, I will try to be concise as possible to explain my
problem. I building a admin section on my site to manage all my ruby
scripts. I want to remove the script from cron and put them in
backgroundrb so every admin user can start/stop those scripts when
they want.
So the script start first from the scheduler(yml) to be sure they run
each day. The way I understand bgrb, less worker is better. So I would
like to reuse the worker started from the scheduler in my rails admin
to start again the task sync_ezdeal.
I have a worker started from backgroundrb.yml file like this:
---
:backgroundrb:
:port: 11006
:ip: 0.0.0.0
:result_storage: memcache
:memcache: "localhost:11211,localhost:11211"
:schedules:
:ezdeal_worker:
:sync_ezdeal:
:trigger_args: 0 30 3 * * * *
----------------
#This script run each night at 3:30.
My Worker:
class EzdealWorker < BackgrounDRb::MetaWorker
set_worker_name :ezdeal_worker
pool_size 1
def sync_ezdeal(args = nil)
.... #do job
cache[:filename] = xml_file # This is the filename of a generated xml
file that the job sync_ezdeal build and save on disk. This filename
must be unique each time the task sync_ezdeal is called. So i can look
at the file from the admin site.
end
My call in the controller to start a new job is that:
MiddleMan.worker(:ezdeal_worker).enq_sync_ezdeal(:job_key
=>current_time, :arg =>{:filename=>current_time,:local =>'/Users/carl/
gm_core/log/ezdeal/ezdeal.xml'})
The job is queued correctly but when I try to get the :filename that
doesnt work.
This is my view:
<%
@tasks = BdrbJobQueue.find(:all,:order =>'id DESC')
@tasks.each do |task|
%>
<tr>
<td width="5%" align="center"><%=h task.id%></td>
<td width="20%"><%=h task.worker_method %></td>
<td width="20%"><%=h time_ago_in_words(task.submitted_at) %></td>
<td width="20%"><%=h time_ago_in_words(task.started_at) unless
task.started_at.blank?%></td>
<td width="20%"><%=h time_ago_in_words(task.finished_at) if
task.finished? %></td>
<td width="20%"><%=h task.finished? %></td>
<td width="20%" align="right">
<!-- Action td -->
<%=link_to
'log_file
',ezdeal_log_file_admin_task_path
(MiddleMan.worker(task.worker_name).ask_result(:filename)) if
task.finished?%>
##### This is were i retreive the filename from the worker. This
should came from the database?no? like task.args[:filename]
</td>
</tr>
<%end%>
I don't known how to fetch the filename for each job (with the
job_key??). Because the ask_result is alway overright when I start a
new job. The way I understand it is the cache array is the same for
all workers of the same type. so the cache[:filename] is overight each
time the sync_ezdeal is called. I tried by creating a new worker each
time, it's working, but that look overkill to me.
Is there a way to retreive the good key?
I saw in the database the field args that old the data push in the
cache array. Is there a way to retreive it? Because currently the data
look to have strange character in it. In the database it's a binary
object.. Not sure what to do with that...
ex:args |
\004\010{\007:\015filename"\02414012009_155548:\012local"./Users/carl/
gm_core/log/ezdeal/ezdeal.xml
Thanks for any help!
I hope I am clear enough!
Carl Pelletier
_______________________________________________
Backgroundrb-devel mailing list
[email protected]
http://rubyforge.org/mailman/listinfo/backgroundrb-devel
