Re: [GridSim-users] AllocPolicy: sendFinishGridlet() (Katia Leal)

[Jawad A. <ja...@le...>]

Hi Katia
I think there is a possibility of conflict or race occurrence between events.
If you can send the code I might be able to tell more.

Regards,
Jawad



On 2 Feb 2012, at 10:52, gri...@li... wrote:

> Send Gridsim-users mailing list submissions to
> 	gri...@li...
> 
> To subscribe or unsubscribe via the World Wide Web, visit
> 	https://lists.sourceforge.net/lists/listinfo/gridsim-users
> or, via email, send a message with subject or body 'help' to
> 	gri...@li...
> 
> You can reach the person managing the list at
> 	gri...@li...
> 
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of Gridsim-users digest..."
> 
> 
> Today's Topics:
> 
>   1. AllocPolicy: sendFinishGridlet() (Katia Leal)
> 
> 
> ----------------------------------------------------------------------
> 
> Message: 1
> Date: Thu, 2 Feb 2012 11:52:29 +0100
> From: Katia Leal <kat...@gm...>
> Subject: [GridSim-users] AllocPolicy: sendFinishGridlet()
> To: gri...@li...
> Message-ID:
> 	<CA+...@ma...>
> Content-Type: text/plain; charset="iso-8859-1"
> 
> Hello,
> I'm usign GridSim5.2 to implement a generic federated grid.
> I have developed my own classes including the class AllocationPolicy, that
> extends AllocPolicy class.
> Instead of implement this new class from scrach, I have modified some of
> the SpaceShared class methods.
> However, other methods remain the same, for example, "private void *
> gridletFinish*(ResGridlet rgl, int status)".
> This method invokes super.*sendFinishGridlet*( rgl.getGridlet() ); method
> as soon as a gridlet finishes its execution.
> Finally, sendFinishGridlet method invokes the super.*sim_schedule*(outputPort_,
> 0, GridSimTags.GRIDLET_RETURN, obj);
> method of Sim_entity class.
> 
> However, I have discovered that in some cases there are gridlets whose
> ending is reported much later.
> Their status is SUCCESS, but their owners received the completion of the
> gridlet even 3000 seconds later (simulation time).
> Here you have part of the trace output of one simulation:
> 
> --> MetaScheduler-X Gridlet-1978 ASSIGNED to Y at 304800.0
> --> MetaScheduler-Y Gridlet-1978 ASSIGNED to M1 at *304830.0*
> --> MetaScheduler-X Gridlet-1978 is InExec at 308400.0
> --> MetaScheduler-X Gridlet-1978 is Success at *308700.0*
> --> MetaScheduler-X Gridlet-1978 is Success at 309000.0
> --> MetaScheduler-X Gridlet-1978 is Success at 309300.0
> --> MetaScheduler-X Gridlet-1978 is Success at 309600.0
> --> MetaScheduler-X Gridlet-1978 is Success at 309900.0
> --> MetaScheduler-X Gridlet-1978 is Success at 310200.0
> --> MetaScheduler-X Gridlet-1978 is Success at 310500.0
> --> MetaScheduler-X Gridlet-1978 is Success at 310800.0
> --> MetaScheduler-X Gridlet-1978 is Success at 311100.0
> --> MetaScheduler-X Gridlet-1978 is Success at 311400.0
> --> MetaScheduler-X Gridlet-1978 is Success at 311700.0
> --> MetaScheduler-Y received Gridlet-1978 from M1 at 311908.75
> --> MetaScheduler-X received Gridlet-1978 from Y at *311921.75*
> 
> As I said before, this only happens in few cases.
> However, this delays increase whole application makespam
> and spoil the final result.
> 
> Regards,
> Katia.
> -------------- next part --------------
> An HTML attachment was scrubbed...
> 
> ------------------------------
> 
> ------------------------------------------------------------------------------
> Keep Your Developer Skills Current with LearnDevNow!
> The most comprehensive online learning library for Microsoft developers
> is just $99.99! Visual Studio, SharePoint, SQL - plus HTML5, CSS3, MVC3,
> Metro Style Apps, more. Free future releases when you subscribe now!
> http://p.sf.net/sfu/learndevnow-d2d
> 
> ------------------------------
> 
> _______________________________________________
> Gridsim-users mailing list
> Gri...@li...
> https://lists.sourceforge.net/lists/listinfo/gridsim-users
> 
> 
> End of Gridsim-users Digest, Vol 221, Issue 1
> *********************************************