[MTOS-dev] New Dispatcher On the Way

[Hirotaka Ogawa]

On Thu, Mar 6, 2008 at 4:36 PM, Reed A. Cartwright <reed at scit.us> wrote:
> I made some changes the mt-dispatch last weekend.  The current code can
>  be found here:
>
>  http://scit.us/cgi-bin/csvn/misc/mt-dispatch/current/mt-dispatch.fcgi
>
>  I've added support for autodetecting settings changes and restarting the
>  worker threads.  I'm thinking that it may even work with the upgrade
>  process, although I haven't tested it.
>
>  I've also changed the way the publishing queue is handled.  The script
>  now uses Sys::AlarmCall (available from CPAN) to balance waiting for a
>  request and processing a job from the publishing queue.
>
>  Minor changes have been made to the rc.d script as well.
>
>  I'll make an official release when I'm satisfied that it is running
>  smoothly on my two servers, but feel free to look at my code, linked
>  above and tell me what you think.

I think, TheSchwartz part can be separated from this dispatcher and
realized as a plugin which facilitates post_run callback.

Before merging this with MTOS, I hope that you make the dispatcher
core as simple as possible.

>  And finally, I'm willing to merge this dispatcher with MTOS for 4.2
>  release.  I can suggest some minor changes to make it work better like
>  removing need to read a file to know what app class to call, and instead
>  rely on the Applications registry key.
>
>  Down the road, I think that much of the code in the current
>  implementation should be scrapped for FCGI::Engine and
>  CGI::Application::Dispatch, like I wrote in my prototype many weeks ago.
>
>  --
>  *********************************************************
>  Reed A. Cartwright, PhD     http://scit.us/
>  Postdoctoral Researcher     http://www.dererumnatura.us/
>  Department of Genetics      http://www.pandasthumb.org/
>
>  Bioinformatics Research Center
>  North Carolina State University
>  Campus Box 7566
>  Raleigh, NC 27695-7566
>
>  Cuiusvis hominis est errare, nullius nisi
>  insipientis in errore perserverare. --Cicero
>  _______________________________________________
>  MTOS-dev mailing list
>  MTOS-dev at sixapart.com
>  http://www.sixapart.com/mailman/listinfo/mtos-dev
>



-- 
Hirotaka Ogawa makes no sense.
http://as-is.net/blog/