Thanks for the reply,
I was looking at the same classes for possible implementations. I was just
hoping there was a breakdown of the classes somewhere so that I wouldnt have to
wade through all the code looking for. As for the examples..... they all work,
I managed to build and run them in a matter of minutes, and gave me enough info
to set my own project up and running within a very short time.
Well it is a shame to see a project like this just laying around. Quite a
bit
of work has allready been spent on it. As for changes that I make to the
source, Ill post back to the board, though my own programming abilities are
meager, Ive never been taught about programming and I just do this for fun.
Anyways, Ill probably bug you from time to time since you seem to be the
reigning authority for the project.
Thanks,
Richard
________________________________
From: Jonathan Mitchem <[email protected]>
To: Richard Foreman <[email protected]>
Cc: [email protected]
Sent: Wed, March 16, 2011 1:23:47 PM
Subject: Re: [Alchemi-users] A few questions
Nobody is working on this project. I'm still moderating the listserv but
that's
about it.
The original project was a Masters or PhD thesis project, then opensourced
under
the "management" of the University of Melbourne's Gridbus group. They seemed
to
have lost interest. There were a couple of us making some active work around
2008 or so, but Matt got hired by Microsoft and as for me, life got in the way.
Not exactly sure what happened to the others. Anything else in the codebase
(e.g., WCF support) has been contributions from users, but there hasn't been
any
active mainline development for a very long time.
All the real documentation that I know of hasn't really been touched since
about
2004. As for basic usage, some (most?) of the example projects should still
work.
IIRC, the functionality that you're looking for would be something built into
the "scheduler", but I never took a look at that implementation. All that's
there now -- again, as far as I know -- is a basic FIFO scheduler with no
intelligence of jobs or executor resources. We did some brainstorming awhile
back, but never ended up with any code.
http://tools.assembla.com/alchemireloaded
Let me take a quick peek at the code to see if I can at least point you in a
slightly better direction.
-- [looking] --
Ok, so it seems to work like this. There's Alchemi.Manager.IScheduler,
Alchemi.Manager.SchedulerFactory, and Alchemi.Manager.DefaultScheduler.
I believe if you implement the following method in IScheduler:
ThreadIdentifierScheduleNonDedicated(stringexecutorId);
Then you should be able to use that for any more "intelligent" decision making
based upon an Application <-> Executor mapping.
Also I just found MappingScheduler.cs which may provide some/all of what you
need. I didn't really read through the code yet.
http://alchemi.svn.sourceforge.net/viewvc/alchemi/trunk/alchemi-devel/src/Alchemi.Manager/MappingScheduler.cs?revision=289&view=markup
Alchemi.Manager root
http://alchemi.svn.sourceforge.net/viewvc/alchemi/trunk/alchemi-devel/src/Alchemi.Manager/
Hope that helps some,
Jonathan
On Wed, Mar 16, 2011 at 4:32 AM, Richard Foreman <[email protected]> wrote:
Is anyone still working on this project? I see very few postings on this
>board (none on the developer list since 2008), no new updates or anything. Ive
>been looking around and so far the only reference Ive found will get you far
>enough to get a project going but not much more. So for starters, is there a
>reference that shows a break down of the library? So far the closest I have
>come to is in the object browser thanks to the developers having documented the
>code, but that really is kind of a pain in the butt to wade through to figure
>out how to do something. And for a specific question, is there a way to
>deauthorize an executor for a single project while leaving it free to tackle
>other projects? I have my application watching the thread finished and
>threadfailed events and serving out the next threads based on the returned
>results. However, sometimes Ill start a project and Ill get one computer out
of
>all of them that wont run the thread for one reason or another, and the manager
>just dishes out thread after thread to the executor. I see where (I believe) I
>should insert some code into Alchemis source code to stop this, but I have a
>hard time believing that this ability was not part of the original design.
>
> By the way, I see the beginings to an excellent framework here, why is it so
>dead around here? Did BOINC kill it?
>
>Thanks, Richard
>
>
>
>
>------------------------------------------------------------------------------
>Colocation vs. Managed Hosting
>A question and answer guide to determining the best fit
>for your organization - today and in the future.
>http://p.sf.net/sfu/internap-sfd2d
>_______________________________________________
>alchemi-users mailing list
>[email protected]
>https://lists.sourceforge.net/lists/listinfo/alchemi-users
>
------------------------------------------------------------------------------
Colocation vs. Managed Hosting
A question and answer guide to determining the best fit
for your organization - today and in the future.
http://p.sf.net/sfu/internap-sfd2d
_______________________________________________
alchemi-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/alchemi-users
