On Thu, Jun 20, 2013 at 06:00:11AM -0700, emc_lab wrote:
>
> In answer#1, you mention workers outside of rails app. Is there any
> document or post explaining workers outside of an app (vs inside)? There is
> a snippet of code for workers outside of rails app in ruote document. But
> there is no explanation of what it is and what kind of impact it has on the
> app.
Hello,
it's a bit discouraging to read that after having written lots of
documentation. Is it my fault? Or is it your fault because you can't read
between the lines? All the time I invest in helping you, it will never pay
back. I give you precise help, you make vague affirmations about a "ruote
document" that exists somewhere and talks about worker outside of Rails apps
(can't remember writing it)...
Anyway...
To run workflows, a ruote engine (one could say a ruote system) has to have
at least one ruote worker running.
Ruby on Rails is used to build web applications. When it was a baby
framework, Ruote on Rails was running on top of Webrick, a pure Ruby web
server. Nowadays, there are many web servers to run Rails on top of. Mongrel,
Passenger, Puma, Unicorn, ...
Some of them fork Ruby processes while other use threads with a single Ruby
process, etc... Many choices, many sets of pluses/minuses.
What if you run a ruote worker inside of a Rails app? Well it steals some of
the resources of the rails app itself and the response time may go down or
not. If the Rails app is run on top of a forking server, you may end up
having one ruote worker per Rails app fork, etc...
Running ruote workers externally to the rails app has its advantages, as
said, ruote doesn't eat Rails resources (if we make a parallel, with most of
the database servers, your client doesn't run background indexation work,
your Rails app isn't running indexation work, etc).
Some web servers Rails run on, can go in some "quiet mode" when no process is
running (a new one will be spawned if the need arise), that means that if you
rely on such a web server and your run the ruote worker inside of the Rails
app, you could have times when no ruote worker is on, it shouldn't too bad,
expect when using schedules: if there is a timeout or some work scheduled at
time t and at that time there is no ruote worker, the work will not happen
{on time}.
Should I write documentation for all of this or should I expect the people
integrating my library to know that beforehand?
Are you entitled to such documentation or am I entitled to believe you [want
to] understand what you are building?
> Also any document or post or open source project which may help us to
> better understand ruote and its integration with rails are appreciated.
In order:
* http://www.ruby-lang.org/
* http://rubyonrails.org/
* http://guides.rubyonrails.org/
* http://ruote.rubyforge.org/
Best regards,
--
John Mettraux - http://lambda.io/jmettraux
--
--
you received this message because you are subscribed to the "ruote users" group.
to post : send email to [email protected]
to unsubscribe : send email to [email protected]
more options : http://groups.google.com/group/openwferu-users?hl=en
---
You received this message because you are subscribed to the Google Groups
"ruote" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
For more options, visit https://groups.google.com/groups/opt_out.
