You can subscribe to this list here.
2007 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
(5) |
Oct
(24) |
Nov
(5) |
Dec
(4) |
---|---|---|---|---|---|---|---|---|---|---|---|---|
2008 |
Jan
(14) |
Feb
(8) |
Mar
(6) |
Apr
(7) |
May
(1) |
Jun
(22) |
Jul
(6) |
Aug
(22) |
Sep
(22) |
Oct
(8) |
Nov
(10) |
Dec
(31) |
2009 |
Jan
(18) |
Feb
(29) |
Mar
(25) |
Apr
(11) |
May
(6) |
Jun
(7) |
Jul
(8) |
Aug
|
Sep
(3) |
Oct
(20) |
Nov
(11) |
Dec
(1) |
2010 |
Jan
(15) |
Feb
(33) |
Mar
(20) |
Apr
(13) |
May
(1) |
Jun
(4) |
Jul
(7) |
Aug
(11) |
Sep
(5) |
Oct
(6) |
Nov
(7) |
Dec
(12) |
2011 |
Jan
(6) |
Feb
(7) |
Mar
(1) |
Apr
(7) |
May
|
Jun
(10) |
Jul
(2) |
Aug
(2) |
Sep
(4) |
Oct
(3) |
Nov
(4) |
Dec
(1) |
2012 |
Jan
(3) |
Feb
(5) |
Mar
(15) |
Apr
(3) |
May
(2) |
Jun
(5) |
Jul
(5) |
Aug
(7) |
Sep
|
Oct
(2) |
Nov
(3) |
Dec
(1) |
2013 |
Jan
|
Feb
|
Mar
(4) |
Apr
(14) |
May
(8) |
Jun
(4) |
Jul
(3) |
Aug
|
Sep
|
Oct
(1) |
Nov
(1) |
Dec
|
2014 |
Jan
(2) |
Feb
(1) |
Mar
(6) |
Apr
(2) |
May
(1) |
Jun
|
Jul
(4) |
Aug
|
Sep
|
Oct
(3) |
Nov
(3) |
Dec
(2) |
2015 |
Jan
(1) |
Feb
|
Mar
|
Apr
(1) |
May
|
Jun
|
Jul
(1) |
Aug
|
Sep
|
Oct
|
Nov
(2) |
Dec
(1) |
2016 |
Jan
|
Feb
|
Mar
(1) |
Apr
|
May
|
Jun
|
Jul
|
Aug
(2) |
Sep
|
Oct
|
Nov
|
Dec
|
2017 |
Jan
|
Feb
|
Mar
|
Apr
(1) |
May
|
Jun
|
Jul
(1) |
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
2019 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
(1) |
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
2020 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
(1) |
Oct
|
Nov
|
Dec
|
2021 |
Jan
|
Feb
(1) |
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
2023 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
(1) |
Dec
|
2024 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
(1) |
Oct
|
Nov
|
Dec
|
From: Marcos D. de A. <ma...@cs...> - 2008-07-11 06:58:07
|
Dear Shuo Yang, I am wondering if you could explain in more detail what the 4,000 entities actually do. Are you using the Workload class for job submission? If yes, how many instances of it do you have? How many jobs in average are the user entities/workloads submitting? Have you enabled the trace mode (the simulation is generating log files)? How many resources do you have? What resource allocation policies are they using? Regards, Marcos On 10/07/2008, at 7:12 PM, Shuo Yang wrote: > Dear GridSim team, > > I'm Shuo Yang, a PhD student at University of Manchester, UK. I'm > doing > my research project with simulations based on GridSim. However I > encountered scalability issues in the simulation where there are > around > 4000 threads (entities) with Java exception > > 'java.lang.OutOfMemoryError: Java heap space' > > The simulation ran for ~48 hours with increasing memory usage and > ended > up with the OutOfMemoryError exceptions. Please could you help more > with > tuning? Here's more info: > > $ uname -a > Linux nimue.rcs.manchester.ac.uk 2.6.18-1.2798.fc6 #1 SMP Mon Oct 16 > 14:54:20 EDT 2006 i686 i686 i386 GNU/Linux > > $ java -version > java version "1.6.0_06" > Java(TM) SE Runtime Environment (build 1.6.0_06-b02) > Java HotSpot(TM) Server VM (build 10.0-b22, mixed mode) > > > $ ps aux | grep java > > yangs 12579 109 43.7 2915672 1473316 ? Sl 03:10 1011:15 java > -Xss128k -Xms1024m -Xmx1024m -cp gridsim-lib/gridsim.jar:. > topology/HierarchySimulation true > > > $ free -mt > total used free shared buffers > cached > Mem: 3290 2055 1235 0 83 > 363 > -/+ buffers/cache: 1608 1682 > Swap: 8001 0 8001 > Total: 11291 2055 9236 > > > $ top > PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ SWAP CODE > DATA COMMAND > 12579 yangs 17 0 2847m 1.4g 5624 S 100 43.7 1057:02 1.4g 40 > 2.8g java > > > > > Many thanks, > > Shuo 'Alex' Yang. > School of Computer Science > University of Manchester > > > ---------------------------------------------------------------------- > --- > Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW! > Studies have shown that voting for your favorite open source project, > along with a healthy diet, reduces your potential for chronic lameness > and boredom. Vote Now at http://www.sourceforge.net/community/cca08 > _______________________________________________ > Gridsim-users mailing list > Gri...@li... > https://lists.sourceforge.net/lists/listinfo/gridsim-users Marcos Dias de Assuncao Grid Computing and Distributed Systems (GRIDS) Laboratory Department of Computer Science and Software Engineering The University of Melbourne, Australia Email: ma...@cs... ------------- "I sincerely believe the banking institutions having the issuing power of money are more dangerous to liberty than standing armies" Thomas Jefferson |
From: Agustín C. H. <ag...@ds...> - 2008-07-10 12:03:21
|
Dear Shuo Yang, The problem u r facing is quite difficult, and I think that the best way of tackling it is by tuning ur code. This is, make sure u don't waste memory unnecessarily. So, for example, every time that some object is not needed any more, delete it (set it to null so that the garbage collector releases the memory). Also, u can scale ur experiments, so that they become smaller. Hope this helps, Agustin Shuo Yang escribió: > Dear GridSim team, > > I'm Shuo Yang, a PhD student at University of Manchester, UK. I'm doing > my research project with simulations based on GridSim. However I > encountered scalability issues in the simulation where there are around > 4000 threads (entities) with Java exception > > 'java.lang.OutOfMemoryError: Java heap space' > > The simulation ran for ~48 hours with increasing memory usage and ended > up with the OutOfMemoryError exceptions. Please could you help more with > tuning? Here's more info: > > $ uname -a > Linux nimue.rcs.manchester.ac.uk 2.6.18-1.2798.fc6 #1 SMP Mon Oct 16 > 14:54:20 EDT 2006 i686 i686 i386 GNU/Linux > > $ java -version > java version "1.6.0_06" > Java(TM) SE Runtime Environment (build 1.6.0_06-b02) > Java HotSpot(TM) Server VM (build 10.0-b22, mixed mode) > > > $ ps aux | grep java > > yangs 12579 109 43.7 2915672 1473316 ? Sl 03:10 1011:15 java > -Xss128k -Xms1024m -Xmx1024m -cp gridsim-lib/gridsim.jar:. > topology/HierarchySimulation true > > > $ free -mt > total used free shared buffers > cached > Mem: 3290 2055 1235 0 83 > 363 > -/+ buffers/cache: 1608 1682 > Swap: 8001 0 8001 > Total: 11291 2055 9236 > > > $ top > PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ SWAP CODE > DATA COMMAND > 12579 yangs 17 0 2847m 1.4g 5624 S 100 43.7 1057:02 1.4g 40 > 2.8g java > > > > > Many thanks, > > Shuo 'Alex' Yang. > School of Computer Science > University of Manchester > > > ------------------------------------------------------------------------- > Sponsored by: SourceForge.net Community Choice Awards: VOTE NOW! > Studies have shown that voting for your favorite open source project, > along with a healthy diet, reduces your potential for chronic lameness > and boredom. Vote Now at http://www.sourceforge.net/community/cca08 > _______________________________________________ > Gridsim-users mailing list > Gri...@li... > https://lists.sourceforge.net/lists/listinfo/gridsim-users > -- =============================================== Agustin Caminero Research Assistant. Computing Systems Department. Albacete Research Institute of Informatics (I3A). The University of Castilla La Mancha. Campus Universitario s/n, 02071, Albacete. Spain. Phone: +34 967 599 200 Ext. 2693. Fax: +34 967 599 343 http://www.i3a.uclm.es/ =============================================== |
From: Shuo Y. <ya...@cs...> - 2008-07-10 09:25:53
|
Dear GridSim team, I'm Shuo Yang, a PhD student at University of Manchester, UK. I'm doing my research project with simulations based on GridSim. However I encountered scalability issues in the simulation where there are around 4000 threads (entities) with Java exception 'java.lang.OutOfMemoryError: Java heap space' The simulation ran for ~48 hours with increasing memory usage and ended up with the OutOfMemoryError exceptions. Please could you help more with tuning? Here's more info: $ uname -a Linux nimue.rcs.manchester.ac.uk 2.6.18-1.2798.fc6 #1 SMP Mon Oct 16 14:54:20 EDT 2006 i686 i686 i386 GNU/Linux $ java -version java version "1.6.0_06" Java(TM) SE Runtime Environment (build 1.6.0_06-b02) Java HotSpot(TM) Server VM (build 10.0-b22, mixed mode) $ ps aux | grep java yangs 12579 109 43.7 2915672 1473316 ? Sl 03:10 1011:15 java -Xss128k -Xms1024m -Xmx1024m -cp gridsim-lib/gridsim.jar:. topology/HierarchySimulation true $ free -mt total used free shared buffers cached Mem: 3290 2055 1235 0 83 363 -/+ buffers/cache: 1608 1682 Swap: 8001 0 8001 Total: 11291 2055 9236 $ top PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ SWAP CODE DATA COMMAND 12579 yangs 17 0 2847m 1.4g 5624 S 100 43.7 1057:02 1.4g 40 2.8g java Many thanks, Shuo 'Alex' Yang. School of Computer Science University of Manchester |
From: Vinicius P. <vin...@gm...> - 2008-06-27 23:42:29
|
Hi everyone, Thanks in advance for spending your time reading my e-mail. I'm new to this list but I've read almost all the e-mails so far in order to check if a similar problem was arised. As I could find anything that could help me, here it goes my question: I'm trying to use the method getGridletFinishedSoFar of the class Gridlet, but all i get is 0.0 ou the total lenght (when the gridlet is already completed). I execute the method several times during the execution of the gridlet using gridSimHold(interval) between the calls. Anyone could help me? Best regards, Vinicius Pinheiro |
From: Vinicius P. <vin...@gm...> - 2008-06-27 22:45:53
|
Hi everyone, Thanks in advance for spending your time reading my e-mail. I'm new to this list but I've read almost all the e-mails so far in order to check if a similar problem was arised. As I could find anything that could help me, here it goes my question: I'm trying to use the method getGridletFinishedSoFar of the class Gridlet, but all i get is 0.0 ou the total lenght (when the gridlet is already completed). I execute the method several times during the execution of the gridlet using gridSimHold(interval) between the calls. Anyone could help me? Best regards, Vinicius Pinheiro |
From: Azin M. <az...@ma...> - 2008-06-26 22:49:00
|
Hi, I am simulating a scheduling algorithm with Gridsim. I need to know what is the load of a Grid resource during time stamps in my simulation. ResourceCharectristic Class provides the cose of using a resource to me however this is a static information about the resource. what I need to know is how much is the load of a resource in different times of the simulation regarding all the gridlets which are assigned to it. How can I have this information ? Thank you for your good supports, azin |
From: Srikumar V. <sri...@gm...> - 2008-06-26 00:54:57
|
Hi Beth, >> first, if I want to do some research in Replica mechanism, do I need use >> the 'gridbroker'? >> You dont need to use the gridbroker if you are doing research on replicas. However, if you performing scheduling using replica locations, then you may want to use it. >> second, should I change the version of gridsim 4.0 to gridsim 2.1 to >> satisfy the requirement of gridbroker? >> No, the grid broker included in gridsim 4.0 should be fine. Or, you could create your own gridbroker based on the one already provided If you are using GridSim, please join the gridsim-users mailing list on Sourceforge so that many people can help you with problems. Cheers Srikumar. >> I will thank you deeply, if you answer my question quickly. >> >> Beth >> >> |
From: Srikumar V. <sri...@gm...> - 2008-06-25 08:04:13
|
Hi Arif, Can you show a code sample of how you modified the existing classes ? In my opinion and experience, the best way to do this is to extend/inherit from the existing classes. HTH Cheers Srikumar. On Wed, Jun 25, 2008 at 12:07 AM, Mehtab Arif <mus...@ho...> wrote: > Hi all > > i have a problem in context of existing classes. i modify some existing > classes like "MessageBid" by adding modules and add these classes to project > and compile this modified class,but when i compile other classes in a > project that use the object of modify class it give an error message for all > module that i added in existing class. > > > Thnx > Mehtab Arif > > ________________________________ > Discover the new Windows Vista Learn more! > ------------------------------------------------------------------------- > Check out the new SourceForge.net Marketplace. > It's the best place to buy or sell services for > just about anything Open Source. > http://sourceforge.net/services/buy/index.php > _______________________________________________ > Gridsim-users mailing list > Gri...@li... > https://lists.sourceforge.net/lists/listinfo/gridsim-users > > |
From: Mehtab A. <mus...@ho...> - 2008-06-24 14:07:11
|
Hi all i have a problem in context of existing classes. i modify some existing classes like "MessageBid" by adding modules and add these classes to project and compile this modified class,but when i compile other classes in a project that use the object of modify class it give an error message for all module that i added in existing class. Thnx Mehtab Arif _________________________________________________________________ Discover the new Windows Vista http://search.msn.com/results.aspx?q=windows+vista&mkt=en-US&form=QBRE |
From: Marcos D. de A. <ma...@cs...> - 2008-06-24 00:11:48
|
Hi Azin, sorry for the delay in my response. I don't exactly how you are implementing your simulation because I don't have access to its source code, but this is probably not the way to do it. You should always create an entity by extending either Sim_entity of GridSim classes. Sim_entity is a thread, but it differs from a normal java thread in the sense that is aware of the simulation events. Internally, SimJava sets threads to sleep and wakes them up according to simulation events inserted in the queue. If you simply extends a thread that does not take this mechanism into account, the results may vary, but certainly the simulation will not run as expected. Regards, Marcos On 17/06/2008, at 1:09 PM, Azin Moallem wrote: > Hi, > > Thanks for your usefull and descriptive response. While I think the > solution you proposed can help me I have found where the problem is. > When gridletReceive() is invoked while there are actually no events of > type GRIDLET_RETURN the program prints "Sim_system: No more future > events" and seems to even ignore all other incomming GRIDLET_RETURN > events ! > > Is this the case the famework is supposed to work or I am making a > mistake somewhere ?! > > I really appreciate your attention and descriptive responses you > provide > > Azin > > > Marcos Dias de Assuncao wrote: > > >> >> Dear Azin, >> >> Using a thread will not work as you expect. >> >> I think that what you want can be implemented in the following >> manner. You can create a GridSim entity that submits the gridlets as >> you have already created. However, this entity may submit a new >> Gridlet whenever a given internal event is received. As an example, I >> am providing some sample code wherein the entity schedules a new >> Gridlet whenever another returns. The entity's body method would look >> like this: >> >> >> public void body() { >> ... >> int numofGrls = 10; >> int grlReceived = 0; >> >> Sim_event ev = new Sim_event(); >> while (grlReceived < numofGrls) { >> super.sim_get_next(ev); >> if(ev.get_src() == super.get_id() && >> ev.get_tag() == SUBMIT_ONE_MORE_GRIDLET) { >> >> // Submit the gridlet here... >> super.gridletSubmit(gridlet, resID); >> } >> else if (ev.get_tag() == GridSimTags.GRIDLET_RETURN) { >> // add the gridlet to a list, print it or do whatever is required >> here... >> grlReceived++; >> super.sim_schedule(super.get_id(), >> GridSimTags.SCHEDULE_NOW, SUBMIT_ONE_MORE_GRIDLET); >> } >> } >> ... >> } >> >> >> where SUBMIT_ONE_MORE_GRIDLET is just a new tag defined in the class: >> >> private static final int SUBMIT_ONE_MORE_GRIDLET = 5000; >> >> I think this might give you an idea on how it works. >> >> Regards, >> >> Marcos >> >> >> On 16/06/2008, at 12:48 PM, Azin Moallem wrote: >> >>> Hi, >>> >>> I have a question simulating a scheduling algorithm using GridSim >>> and I >>> would be greatful to hear back. >>> >>> I need to simulate a dynamic job submission. I mean while >>> gridlets are >>> being submitted to the grid, completed gridlets are being sent >>> back to >>> users. To do this I am simulating submission of gridlets in the body >>> function of a GridSim entity and receiving back the gridlets is the >>> responsibilty of a Thread which I have created myself. >>> >>> The problem is gridletReceive() always returns null ! and I see this >>> output: "Sim_system: No more future events" early in my simulation. >>> >>> how can i simulate a dynamic job submission environment ? >>> >>> >>> Best, >>> Azin >>> >>> -------------------------------------------------------------------- >>> -- >>> --- >>> Check out the new SourceForge.net Marketplace. >>> It's the best place to buy or sell services for >>> just about anything Open Source. >>> http://sourceforge.net/services/buy/index.php >>> _______________________________________________ >>> Gridsim-users mailing list >>> Gri...@li... >>> https://lists.sourceforge.net/lists/listinfo/gridsim-users >> >> |
From: Gilberto C. <gil...@gm...> - 2008-06-23 13:16:16
|
Thanks Agustín, I'm looking at them carefully. On Mon, Jun 23, 2008 at 4:34 AM, Agustín Caminero Herráez < ag...@ds...> wrote: > Hi Gilberto, > > Gilberto Cunha escribió: > >> Hi! Agustín and Marcos. >> >> Thank you for your suggestions. >> >> I'm looking at them carefully. Initially i would' like use the resource >> failure functionality and set resources initially as failed as you wrote. >> >> But i' don't understand very well what's means some parameters about the >> constructor parameters of class HyperExponential. >> >> |*HyperExponential < >> http://www.gridbus.org/gridsim/doc/api/gridsim/util/HyperExponential.html#HyperExponential%28double,%20double,%20int%29>*(double >> mu, double sigma, int i)| >> >> * What's means the parameter "i" ? >> >> >> * and If I want simulate 40% (80 resources) of resources intially >> failed, and during the simulation join 2,5% (5 resources is the >> same of one machine with one PE) following the distribuition and >> fail more 2,5% of resources available. I have a set of 200 >> GridResources, where each resource is just one machine with one >> PE. For example, what's values I should use. >> >> The hyperexponential distribution is used just to make a randomized > selection of failed resources, failure length, and failure start time. So, > if u want to make such a concrete simulation (without randomized stuff > regarding the number of failed resources), what u should do is remove the > use of that distribution for that purpose, and manually set the number of > failed resources, failures times and so on. > > In order to do this, u have to modify the code of the > RegionalGISWithFailure class, so that u set 80 resources initially failed. > Then, u set the recovery of 5 of those resources (the time of the recovery > can be chosen by a distribution). Later, u must set another 5 resources as > failed (when they fail can be chosen at random). > > To do this, u have to modify processGridResource_Failure function: > > 1- Set 80 resources initially as failed: Step 4 in Fig 3 in the first > document mentioned below. Send a GRIDRESOIURCE_FAILURE event to those 80 > resources. The getNextFailureLengthSample() function should return a finite > duration for 5 of those resources (since they will recover during sims), so > u may probably need to modify this function. For the other 75 resources that > will not recover, the duration of the failure should be infinite. > > 2- For the 5 resources initially working that fail during sims, u must > manually set the number of resources that fail, and choose the moment and > length of the failure. > > U will find an explanation of the way how the failure functionality works > here: > > > http://www.dsi.uclm.es/descargas/thecnicalreports/DIAB-07-01-1/DIAB-07-01-1.pdf > http://www.gridbus.org/papers/gridsim-failure-detection-icpads2007.pdf > > Take a look into the processGridResource_Failure function and tell em if u > keep having doubts. > > All the best, > > Agustin > > Thanks and Regards, >> >> Gilberto Cunha >> >> >> >> >> >> On Wed, Jun 11, 2008 at 7:16 AM, Marcos Dias de Assuncao < >> ma...@cs... <mailto:ma...@cs...>> wrote: >> >> >> Dear Gilberto, >> >> In theory you could create a GridResource dynamically during the >> simulation in the same way that you create GridResources before >> the simulation starts. One entity could, for example, create a >> resource with the required characteristics. The constructors of >> the Sim_entity class call Sim_system.add(this) to include the just >> created entity to the list of entities created. The method add of >> sim_system is as follows: >> >> /** >> * Add a new entity to the simulation. This is present for >> compatibility with existing >> * simulations since entities are automatically added to the >> simulation upon instantiation. >> * @param e The new entity >> */ >> public static void add(Sim_entity e) { >> Sim_event evt; >> if (running) { >> // Post an event to make this entity >> evt = new >> Sim_event(Sim_event.CREATE,clock,current_ent().get_id(),0,0, e); >> future.add_event(evt); >> } else { >> if (e.get_id()==-1) { // Only add once! >> e.set_id(entities.size()); >> entities.add(e); >> } >> } >> } >> >> Basically, if the simulation is running, it will schedule one >> event, which is handled by Sim_system itself in a private method. >> >> Everything looks fine. However, in practice SimJava has a bug >> because future.add_event(evt); may raise an exception as it is not >> synchronised and the list called future (containing the future >> simulation events) may be changed simultaneously by two entities. >> I had concurrent modification problems with the auction framework >> when adding entities at runtime. It is actually simple to solve >> it, but it requires one to change SimJava. >> >> Regards, >> >> Marcos >> >> >> >> On 11/06/2008, at 5:28 PM, Agustín Caminero Herráez wrote: >> >> Hello, >> >> I don't think u can create gridResources dynamically during >> simulations. >> >> What u can do instead is use the resource failure >> functionality and set resources initially as failed, then make >> them available as sim progresses. >> >> This way u would simulate the dynamic creation of GridResources. >> >> Regards, >> >> Agustin >> >> >> Gilberto Cunha escribió: >> >> >> Hello, >> >> I have a question regarding GridResource. >> My doubt is as follows: >> >> How can i create GridResources during the simulation, in a >> dynamic way? >> >> Regards, >> -- Gilberto Cunha Filho >> gil...@gm... <mailto:gil...@gm...> >> <mailto:gil...@gm... >> <mailto:gil...@gm...>> >> SISMO - Laboratório de Sistemas e Mobilidade >> UFMA >> >> ------------------------------------------------------------------------ >> >> >> ------------------------------------------------------------------------- >> Check out the new SourceForge.net Marketplace. >> It's the best place to buy or sell services for >> just about anything Open Source. >> http://sourceforge.net/services/buy/index.php >> >> ------------------------------------------------------------------------ >> >> _______________________________________________ >> Gridsim-users mailing list >> Gri...@li... >> <mailto:Gri...@li...> >> https://lists.sourceforge.net/lists/listinfo/gridsim-users >> >> >> -- =============================================== >> Agustin Caminero >> PhD Student >> Computing Systems Department >> The University of Castilla La Mancha, Albacete. Spain. >> Phone: +34 967 599 200 Ext. 2693. Fax: +34 967 599 343 >> http://www.i3a.uclm.es/ >> =============================================== >> >> >> <agustin.vcf>------------------------------------------------------------------------- >> >> >> Check out the new SourceForge.net Marketplace. >> It's the best place to buy or sell services for >> just about anything Open Source. >> >> http://sourceforge.net/services/buy/index.php_______________________________________________ >> Gridsim-users mailing list >> Gri...@li... >> <mailto:Gri...@li...> >> https://lists.sourceforge.net/lists/listinfo/gridsim-users >> >> >> >> >> >> -- >> Gilberto Cunha Filho >> gil...@gm... <mailto:gil...@gm...> >> SISMO - Laboratório de Sistemas e Mobilidade >> UFMA >> > > -- > =============================================== > Agustin Caminero > PhD Student > Computing Systems Department > The University of Castilla La Mancha, Albacete. Spain. > Phone: +34 967 599 200 Ext. 2693. Fax: +34 967 599 343 > http://www.i3a.uclm.es/ > =============================================== > > -- Gilberto Cunha Filho gil...@gm... SISMO - Laboratório de Sistemas e Mobilidade UFMA |
From: Agustín C. H. <ag...@ds...> - 2008-06-23 07:23:33
|
Hi Gilberto, Gilberto Cunha escribió: > Hi! Agustín and Marcos. > > Thank you for your suggestions. > > I'm looking at them carefully. Initially i would' like use the > resource failure functionality and set resources initially as failed > as you wrote. > > But i' don't understand very well what's means some parameters about > the constructor parameters of class HyperExponential. > > |*HyperExponential > <http://www.gridbus.org/gridsim/doc/api/gridsim/util/HyperExponential.html#HyperExponential%28double,%20double,%20int%29>*(double mu, > double sigma, int i)| > > * What's means the parameter "i" ? > > > * and If I want simulate 40% (80 resources) of resources intially > failed, and during the simulation join 2,5% (5 resources is the > same of one machine with one PE) following the distribuition and > fail more 2,5% of resources available. I have a set of 200 > GridResources, where each resource is just one machine with one > PE. For example, what's values I should use. > The hyperexponential distribution is used just to make a randomized selection of failed resources, failure length, and failure start time. So, if u want to make such a concrete simulation (without randomized stuff regarding the number of failed resources), what u should do is remove the use of that distribution for that purpose, and manually set the number of failed resources, failures times and so on. In order to do this, u have to modify the code of the RegionalGISWithFailure class, so that u set 80 resources initially failed. Then, u set the recovery of 5 of those resources (the time of the recovery can be chosen by a distribution). Later, u must set another 5 resources as failed (when they fail can be chosen at random). To do this, u have to modify processGridResource_Failure function: 1- Set 80 resources initially as failed: Step 4 in Fig 3 in the first document mentioned below. Send a GRIDRESOIURCE_FAILURE event to those 80 resources. The getNextFailureLengthSample() function should return a finite duration for 5 of those resources (since they will recover during sims), so u may probably need to modify this function. For the other 75 resources that will not recover, the duration of the failure should be infinite. 2- For the 5 resources initially working that fail during sims, u must manually set the number of resources that fail, and choose the moment and length of the failure. U will find an explanation of the way how the failure functionality works here: http://www.dsi.uclm.es/descargas/thecnicalreports/DIAB-07-01-1/DIAB-07-01-1.pdf http://www.gridbus.org/papers/gridsim-failure-detection-icpads2007.pdf Take a look into the processGridResource_Failure function and tell em if u keep having doubts. All the best, Agustin > Thanks and Regards, > > Gilberto Cunha > > > > > > On Wed, Jun 11, 2008 at 7:16 AM, Marcos Dias de Assuncao > <ma...@cs... <mailto:ma...@cs...>> wrote: > > > Dear Gilberto, > > In theory you could create a GridResource dynamically during the > simulation in the same way that you create GridResources before > the simulation starts. One entity could, for example, create a > resource with the required characteristics. The constructors of > the Sim_entity class call Sim_system.add(this) to include the just > created entity to the list of entities created. The method add of > sim_system is as follows: > > /** > * Add a new entity to the simulation. This is present for > compatibility with existing > * simulations since entities are automatically added to the > simulation upon instantiation. > * @param e The new entity > */ > public static void add(Sim_entity e) { > Sim_event evt; > if (running) { > // Post an event to make this entity > evt = new > Sim_event(Sim_event.CREATE,clock,current_ent().get_id(),0,0, e); > future.add_event(evt); > } else { > if (e.get_id()==-1) { // Only add once! > e.set_id(entities.size()); > entities.add(e); > } > } > } > > Basically, if the simulation is running, it will schedule one > event, which is handled by Sim_system itself in a private method. > > Everything looks fine. However, in practice SimJava has a bug > because future.add_event(evt); may raise an exception as it is not > synchronised and the list called future (containing the future > simulation events) may be changed simultaneously by two entities. > I had concurrent modification problems with the auction framework > when adding entities at runtime. It is actually simple to solve > it, but it requires one to change SimJava. > > Regards, > > Marcos > > > > On 11/06/2008, at 5:28 PM, Agustín Caminero Herráez wrote: > > Hello, > > I don't think u can create gridResources dynamically during > simulations. > > What u can do instead is use the resource failure > functionality and set resources initially as failed, then make > them available as sim progresses. > > This way u would simulate the dynamic creation of GridResources. > > Regards, > > Agustin > > > Gilberto Cunha escribió: > > > Hello, > > I have a question regarding GridResource. > My doubt is as follows: > > How can i create GridResources during the simulation, in a > dynamic way? > > Regards, > -- > Gilberto Cunha Filho > gil...@gm... <mailto:gil...@gm...> > <mailto:gil...@gm... > <mailto:gil...@gm...>> > SISMO - Laboratório de Sistemas e Mobilidade > UFMA > ------------------------------------------------------------------------ > > ------------------------------------------------------------------------- > Check out the new SourceForge.net Marketplace. > It's the best place to buy or sell services for > just about anything Open Source. > http://sourceforge.net/services/buy/index.php > ------------------------------------------------------------------------ > > _______________________________________________ > Gridsim-users mailing list > Gri...@li... > <mailto:Gri...@li...> > https://lists.sourceforge.net/lists/listinfo/gridsim-users > > > -- > =============================================== > Agustin Caminero > PhD Student > Computing Systems Department > The University of Castilla La Mancha, Albacete. Spain. > Phone: +34 967 599 200 Ext. 2693. Fax: +34 967 599 343 > http://www.i3a.uclm.es/ > =============================================== > > <agustin.vcf>------------------------------------------------------------------------- > > > Check out the new SourceForge.net Marketplace. > It's the best place to buy or sell services for > just about anything Open Source. > http://sourceforge.net/services/buy/index.php_______________________________________________ > Gridsim-users mailing list > Gri...@li... > <mailto:Gri...@li...> > https://lists.sourceforge.net/lists/listinfo/gridsim-users > > > > > > -- > Gilberto Cunha Filho > gil...@gm... <mailto:gil...@gm...> > SISMO - Laboratório de Sistemas e Mobilidade > UFMA -- =============================================== Agustin Caminero PhD Student Computing Systems Department The University of Castilla La Mancha, Albacete. Spain. Phone: +34 967 599 200 Ext. 2693. Fax: +34 967 599 343 http://www.i3a.uclm.es/ =============================================== |
From: Gilberto C. <gil...@gm...> - 2008-06-22 01:00:08
|
Hi! Agustín and Marcos. Thank you for your suggestions. I'm looking at them carefully. Initially i would' like use the resource failure functionality and set resources initially as failed as you wrote. But i' don't understand very well what's means some parameters about the constructor parameters of class HyperExponential. *HyperExponential<http://www.gridbus.org/gridsim/doc/api/gridsim/util/HyperExponential.html#HyperExponential%28double,%20double,%20int%29> *(double mu, double sigma, int i) - What's means the parameter "i" ? - and If I want simulate 40% (80 resources) of resources intially failed, and during the simulation join 2,5% (5 resources is the same of one machine with one PE) following the distribuition and fail more 2,5% of resources available. I have a set of 200 GridResources, where each resource is just one machine with one PE. For example, what's values I should use. Thanks and Regards, Gilberto Cunha On Wed, Jun 11, 2008 at 7:16 AM, Marcos Dias de Assuncao < ma...@cs...> wrote: > > Dear Gilberto, > > In theory you could create a GridResource dynamically during the simulation > in the same way that you create GridResources before the simulation starts. > One entity could, for example, create a resource with the required > characteristics. The constructors of the Sim_entity class call > Sim_system.add(this) to include the just created entity to the list of > entities created. The method add of sim_system is as follows: > > /** > * Add a new entity to the simulation. This is present for compatibility > with existing > * simulations since entities are automatically added to the simulation > upon instantiation. > * @param e The new entity > */ > public static void add(Sim_entity e) { > Sim_event evt; > if (running) { > // Post an event to make this entity > evt = new Sim_event(Sim_event.CREATE,clock,current_ent().get_id(),0,0, > e); > future.add_event(evt); > } else { > if (e.get_id()==-1) { // Only add once! > e.set_id(entities.size()); > entities.add(e); > } > } > } > > Basically, if the simulation is running, it will schedule one event, which > is handled by Sim_system itself in a private method. > > Everything looks fine. However, in practice SimJava has a bug because > future.add_event(evt); may raise an exception as it is not synchronised and > the list called future (containing the future simulation events) may be > changed simultaneously by two entities. I had concurrent modification > problems with the auction framework when adding entities at runtime. It is > actually simple to solve it, but it requires one to change SimJava. > > Regards, > > Marcos > > > > On 11/06/2008, at 5:28 PM, Agustín Caminero Herráez wrote: > > Hello, >> >> I don't think u can create gridResources dynamically during simulations. >> >> What u can do instead is use the resource failure functionality and set >> resources initially as failed, then make them available as sim progresses. >> >> This way u would simulate the dynamic creation of GridResources. >> >> Regards, >> >> Agustin >> >> >> Gilberto Cunha escribió: >> >>> >>> Hello, >>> >>> I have a question regarding GridResource. >>> My doubt is as follows: >>> >>> How can i create GridResources during the simulation, in a dynamic way? >>> >>> Regards, >>> -- >>> Gilberto Cunha Filho >>> gil...@gm... <mailto:gil...@gm...> >>> SISMO - Laboratório de Sistemas e Mobilidade >>> UFMA >>> ------------------------------------------------------------------------ >>> >>> ------------------------------------------------------------------------- >>> Check out the new SourceForge.net Marketplace. >>> It's the best place to buy or sell services for >>> just about anything Open Source. >>> http://sourceforge.net/services/buy/index.php >>> ------------------------------------------------------------------------ >>> >>> _______________________________________________ >>> Gridsim-users mailing list >>> Gri...@li... >>> https://lists.sourceforge.net/lists/listinfo/gridsim-users >>> >>> >> -- >> =============================================== >> Agustin Caminero >> PhD Student >> Computing Systems Department >> The University of Castilla La Mancha, Albacete. Spain. >> Phone: +34 967 599 200 Ext. 2693. Fax: +34 967 599 343 >> http://www.i3a.uclm.es/ >> =============================================== >> >> >> <agustin.vcf>------------------------------------------------------------------------- >> Check out the new SourceForge.net Marketplace. >> It's the best place to buy or sell services for >> just about anything Open Source. >> http://sourceforge.net/services/buy/ >> index.php_______________________________________________ >> Gridsim-users mailing list >> Gri...@li... >> https://lists.sourceforge.net/lists/listinfo/gridsim-users >> > > -- Gilberto Cunha Filho gil...@gm... SISMO - Laboratório de Sistemas e Mobilidade UFMA |
From: Azin M. <az...@ma...> - 2008-06-17 03:09:50
|
Hi, Thanks for your usefull and descriptive response. While I think the solution you proposed can help me I have found where the problem is. When gridletReceive() is invoked while there are actually no events of type GRIDLET_RETURN the program prints "Sim_system: No more future events" and seems to even ignore all other incomming GRIDLET_RETURN events ! Is this the case the famework is supposed to work or I am making a mistake somewhere ?! I really appreciate your attention and descriptive responses you provide Azin Marcos Dias de Assuncao wrote: > >Dear Azin, > >Using a thread will not work as you expect. > >I think that what you want can be implemented in the following >manner. You can create a GridSim entity that submits the gridlets as >you have already created. However, this entity may submit a new >Gridlet whenever a given internal event is received. As an example, I >am providing some sample code wherein the entity schedules a new >Gridlet whenever another returns. The entity's body method would look >like this: > > >public void body() { >... > int numofGrls = 10; > int grlReceived = 0; > > Sim_event ev = new Sim_event(); > while (grlReceived < numofGrls) { > super.sim_get_next(ev); > if(ev.get_src() == super.get_id() && > ev.get_tag() == SUBMIT_ONE_MORE_GRIDLET) { > > // Submit the gridlet here... > super.gridletSubmit(gridlet, resID); > } > else if (ev.get_tag() == GridSimTags.GRIDLET_RETURN) { > // add the gridlet to a list, print it or do whatever is required >here... > grlReceived++; > super.sim_schedule(super.get_id(), > GridSimTags.SCHEDULE_NOW, SUBMIT_ONE_MORE_GRIDLET); > } > } >... >} > > >where SUBMIT_ONE_MORE_GRIDLET is just a new tag defined in the class: > >private static final int SUBMIT_ONE_MORE_GRIDLET = 5000; > >I think this might give you an idea on how it works. > >Regards, > >Marcos > > >On 16/06/2008, at 12:48 PM, Azin Moallem wrote: > >> Hi, >> >> I have a question simulating a scheduling algorithm using GridSim >> and I >> would be greatful to hear back. >> >> I need to simulate a dynamic job submission. I mean while gridlets are >> being submitted to the grid, completed gridlets are being sent back to >> users. To do this I am simulating submission of gridlets in the body >> function of a GridSim entity and receiving back the gridlets is the >> responsibilty of a Thread which I have created myself. >> >> The problem is gridletReceive() always returns null ! and I see this >> output: "Sim_system: No more future events" early in my simulation. >> >> how can i simulate a dynamic job submission environment ? >> >> >> Best, >> Azin >> >> ---------------------------------------------------------------------- >> --- >> Check out the new SourceForge.net Marketplace. >> It's the best place to buy or sell services for >> just about anything Open Source. >> http://sourceforge.net/services/buy/index.php >> _______________________________________________ >> Gridsim-users mailing list >> Gri...@li... >> https://lists.sourceforge.net/lists/listinfo/gridsim-users > > |
From: Marcos D. de A. <ma...@cs...> - 2008-06-16 07:49:51
|
Dear Azin, Using a thread will not work as you expect. I think that what you want can be implemented in the following manner. You can create a GridSim entity that submits the gridlets as you have already created. However, this entity may submit a new Gridlet whenever a given internal event is received. As an example, I am providing some sample code wherein the entity schedules a new Gridlet whenever another returns. The entity's body method would look like this: public void body() { ... int numofGrls = 10; int grlReceived = 0; Sim_event ev = new Sim_event(); while (grlReceived < numofGrls) { super.sim_get_next(ev); if(ev.get_src() == super.get_id() && ev.get_tag() == SUBMIT_ONE_MORE_GRIDLET) { // Submit the gridlet here... super.gridletSubmit(gridlet, resID); } else if (ev.get_tag() == GridSimTags.GRIDLET_RETURN) { // add the gridlet to a list, print it or do whatever is required here... grlReceived++; super.sim_schedule(super.get_id(), GridSimTags.SCHEDULE_NOW, SUBMIT_ONE_MORE_GRIDLET); } } ... } where SUBMIT_ONE_MORE_GRIDLET is just a new tag defined in the class: private static final int SUBMIT_ONE_MORE_GRIDLET = 5000; I think this might give you an idea on how it works. Regards, Marcos On 16/06/2008, at 12:48 PM, Azin Moallem wrote: > Hi, > > I have a question simulating a scheduling algorithm using GridSim > and I > would be greatful to hear back. > > I need to simulate a dynamic job submission. I mean while gridlets are > being submitted to the grid, completed gridlets are being sent back to > users. To do this I am simulating submission of gridlets in the body > function of a GridSim entity and receiving back the gridlets is the > responsibilty of a Thread which I have created myself. > > The problem is gridletReceive() always returns null ! and I see this > output: "Sim_system: No more future events" early in my simulation. > > how can i simulate a dynamic job submission environment ? > > > Best, > Azin > > ---------------------------------------------------------------------- > --- > Check out the new SourceForge.net Marketplace. > It's the best place to buy or sell services for > just about anything Open Source. > http://sourceforge.net/services/buy/index.php > _______________________________________________ > Gridsim-users mailing list > Gri...@li... > https://lists.sourceforge.net/lists/listinfo/gridsim-users |
From: Azin M. <az...@ma...> - 2008-06-16 02:48:03
|
Hi, I have a question simulating a scheduling algorithm using GridSim and I would be greatful to hear back. I need to simulate a dynamic job submission. I mean while gridlets are being submitted to the grid, completed gridlets are being sent back to users. To do this I am simulating submission of gridlets in the body function of a GridSim entity and receiving back the gridlets is the responsibilty of a Thread which I have created myself. The problem is gridletReceive() always returns null ! and I see this output: "Sim_system: No more future events" early in my simulation. how can i simulate a dynamic job submission environment ? Best, Azin |
From: Marcos D. de A. <ma...@cs...> - 2008-06-11 10:16:47
|
Dear Gilberto, In theory you could create a GridResource dynamically during the simulation in the same way that you create GridResources before the simulation starts. One entity could, for example, create a resource with the required characteristics. The constructors of the Sim_entity class call Sim_system.add(this) to include the just created entity to the list of entities created. The method add of sim_system is as follows: /** * Add a new entity to the simulation. This is present for compatibility with existing * simulations since entities are automatically added to the simulation upon instantiation. * @param e The new entity */ public static void add(Sim_entity e) { Sim_event evt; if (running) { // Post an event to make this entity evt = new Sim_event(Sim_event.CREATE,clock,current_ent().get_id (),0,0, e); future.add_event(evt); } else { if (e.get_id()==-1) { // Only add once! e.set_id(entities.size()); entities.add(e); } } } Basically, if the simulation is running, it will schedule one event, which is handled by Sim_system itself in a private method. Everything looks fine. However, in practice SimJava has a bug because future.add_event(evt); may raise an exception as it is not synchronised and the list called future (containing the future simulation events) may be changed simultaneously by two entities. I had concurrent modification problems with the auction framework when adding entities at runtime. It is actually simple to solve it, but it requires one to change SimJava. Regards, Marcos On 11/06/2008, at 5:28 PM, Agustín Caminero Herráez wrote: > Hello, > > I don't think u can create gridResources dynamically during > simulations. > > What u can do instead is use the resource failure functionality and > set resources initially as failed, then make them available as sim > progresses. > > This way u would simulate the dynamic creation of GridResources. > > Regards, > > Agustin > > > Gilberto Cunha escribió: >> >> Hello, >> >> I have a question regarding GridResource. >> My doubt is as follows: >> >> How can i create GridResources during the simulation, in a dynamic >> way? >> >> Regards, >> -- >> Gilberto Cunha Filho >> gil...@gm... <mailto:gil...@gm...> >> SISMO - Laboratório de Sistemas e Mobilidade >> UFMA >> --------------------------------------------------------------------- >> --- >> >> --------------------------------------------------------------------- >> ---- >> Check out the new SourceForge.net Marketplace. >> It's the best place to buy or sell services for >> just about anything Open Source. >> http://sourceforge.net/services/buy/index.php >> --------------------------------------------------------------------- >> --- >> >> _______________________________________________ >> Gridsim-users mailing list >> Gri...@li... >> https://lists.sourceforge.net/lists/listinfo/gridsim-users >> > > -- > =============================================== > Agustin Caminero > PhD Student > Computing Systems Department > The University of Castilla La Mancha, Albacete. Spain. > Phone: +34 967 599 200 Ext. 2693. Fax: +34 967 599 343 > http://www.i3a.uclm.es/ > =============================================== > > <agustin.vcf>--------------------------------------------------------- > ---------------- > Check out the new SourceForge.net Marketplace. > It's the best place to buy or sell services for > just about anything Open Source. > http://sourceforge.net/services/buy/ > index.php_______________________________________________ > Gridsim-users mailing list > Gri...@li... > https://lists.sourceforge.net/lists/listinfo/gridsim-users |
From: Anthony S. <an...@cs...> - 2008-06-11 07:32:32
|
Hello Pablo, Thank you for your suggestions. I will look at them carefully. cheers, anthony ---------------------------------------- Anthony Sulistio PhD student Room 5.18a, ICT Building 111 Barry St Department of Computer Science and Software Engineering The University of Melbourne Victoria 3053, Australia Website: http://www.csse.unimelb.edu.au/~anthony ---------------------------------------- On Tue, 3 Jun 2008, pchacin wrote: > Hi all > > I'm just starting to use gridsim. Looking at the code of > ResouceCalendar class, I've seen that the background load > is calculated using the "regularload" array, which is defined > internally in the class's construct methods. Therefore it seams > that I can't change it. Is this correct? > > If so, I would suggest to modify the code to allow this array to me > customized. Some options are: > 1. make it a protected field and initialize it in the declaration > so that subclasses could redefine it. > 2. Pass it as a parameter to the constructor (a better option) > > Regards > > -- > ======================================================================== > 0 0 0 Pablo Chacin | Dept. d'Arquitectura de Computadors > 0 0 0 e-mail: pc...@ac... | UPC-Campus Nord. Modul D6-212 > 0 0 0 phone: +34 +93 405 4059 | Jordi Girona, 1-3 > U P C fax: +34 +93 401 7055 | 08034 Barcelona - SPAIN > www: http://personals.ac.upc.es/pchacin > ======================================================================== > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2008. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ > Gridsim-users mailing list > Gri...@li... > https://lists.sourceforge.net/lists/listinfo/gridsim-users > |
From: Agustín C. H. <ag...@ds...> - 2008-06-11 07:17:45
|
Hello, I don't think u can create gridResources dynamically during simulations. What u can do instead is use the resource failure functionality and set resources initially as failed, then make them available as sim progresses. This way u would simulate the dynamic creation of GridResources. Regards, Agustin Gilberto Cunha escribió: > > Hello, > > I have a question regarding GridResource. > My doubt is as follows: > > How can i create GridResources during the simulation, in a dynamic way? > > Regards, > -- > Gilberto Cunha Filho > gil...@gm... <mailto:gil...@gm...> > SISMO - Laboratório de Sistemas e Mobilidade > UFMA > ------------------------------------------------------------------------ > > ------------------------------------------------------------------------- > Check out the new SourceForge.net Marketplace. > It's the best place to buy or sell services for > just about anything Open Source. > http://sourceforge.net/services/buy/index.php > ------------------------------------------------------------------------ > > _______________________________________________ > Gridsim-users mailing list > Gri...@li... > https://lists.sourceforge.net/lists/listinfo/gridsim-users > -- =============================================== Agustin Caminero PhD Student Computing Systems Department The University of Castilla La Mancha, Albacete. Spain. Phone: +34 967 599 200 Ext. 2693. Fax: +34 967 599 343 http://www.i3a.uclm.es/ =============================================== |
From: 黄. <yua...@js...> - 2008-06-11 03:07:44
|
Hi,everyone: i meet some difficulty when i use gridsim to implement my thoughts. I plan to create a architecture that node A has two sub-nodes B,C, and B,C have two sub-nodes D,E and F,G respectively, which like this: F------| |-------C----| G------| | |-------A D------| | |-------B----| E------| At the first beginning, D E F G will send messages to B C, and B C then send new messages to A. After processing in A, messages will be back from A to D E F G through B C, then the next cycle begin again. But i found the amount of messages send by D E F G will not be very large. 200~500 is ok, but when it exceed 1000, i will get the message "Sim_system: No more future events". I try to solve it by myself, but found no reason. So i hope anyone here can help me. Thanks a lot!! hyqgod 2008-06-11 |
From: Gilberto C. <gil...@gm...> - 2008-06-10 19:25:00
|
Hello, I have a question regarding GridResource. My doubt is as follows: How can i create GridResources during the simulation, in a dynamic way? Regards, -- Gilberto Cunha Filho gil...@gm... SISMO - Laboratório de Sistemas e Mobilidade UFMA |
From: Agustin C. Herra¡ez <sup...@gm...> - 2008-06-09 16:07:33
|
Hello, I have a question regarding RateControlledScheduler. My problem is as follows: I have 2 network classes, with different baud rates associated to each one. If I understood the explanation of RateControlledScheduler, each traffic class has its own baud rate in each link. This is, no matter how much traffic from one class exists, the other class will use the same baud rate at all times, am I right? But I found that ping packets (from one of the classes) took different time depending on how much gridlets traffic (from the other class) exists. My experiments are as follows, my gridlets belong to one of the classes, and I do pings which belong to the other class. Nothing else flow through the network apart from pings and gridlets. What do u think is the reason for this behavior? Can anything else affect the latencies of pings packets? Perhaps the load of computing resources affects latencies of pings? Thanks in advance for ur help! All the best, Agustin |
From: Agustín C. H. <ag...@ds...> - 2008-06-09 08:05:55
|
Hello, I have a question regarding RateControlledScheduler. My problem is as follows: I have 2 network classes, with different baud rates associated to each one. If I understood the explanation of RateControlledScheduler, each traffic class has its own baud rate in each link. This is, no matter how much traffic from one class exists, the other class will use the same baud rate at all times, am I right? But I found that ping packets (from one of the classes) took different time depending on how much gridlets traffic (from the other class) exists. My experiments are as follows, my gridlets belong to one of the classes, and I do pings which belong to the other class. Nothing else flow through the network apart from pings and gridlets. What do u think is the reason for this behavior? Can anything else affect the latencies of pings packets? Perhaps the load of computing resources affects latencies of pings? Thanks in advance for ur help! All the best, Agustin -- =============================================== Agustin Caminero PhD Student Computing Systems Department The University of Castilla La Mancha, Albacete. Spain. Phone: +34 967 599 200 Ext. 2693. Fax: +34 967 599 343 http://www.i3a.uclm.es/ =============================================== |
From: Agustín C. H. <ag...@ds...> - 2008-06-08 11:20:17
|
Hello, I have a question regarding RateControlledScheduler. My problem is as follows: I have 2 network classes, with different baud rates associated to each one. If I understood the explanation of RateControlledScheduler, each traffic class has its own baud rate in each link. This is, no matter how much traffic from one class exists, the other class will use the same baud rate at all times, am I right? But I found that ping packets (from one of the classes) took different time depending on how much gridlets traffic (from the other class) exists. My experiments are as follows, my gridlets belong to one of the classes, and I do pings which belong to the other class. Nothing else flow through the network apart from pings and gridlets. What do u think is the reason for this behavior? Can anything else affect the latencies of pings packets? Perhaps the load of computing resources affects latencies of pings? Thanks in advance for ur help! All the best, Agustin -- =============================================== Agustin Caminero PhD Student Computing Systems Department The University of Castilla La Mancha, Albacete. Spain. Phone: +34 967 599 200 Ext. 2693. Fax: +34 967 599 343 http://www.i3a.uclm.es/ =============================================== |
From: pchacin <pc...@ac...> - 2008-06-03 11:26:28
|
Hi all I'm just starting to use gridsim. Looking at the code of ResouceCalendar class, I've seen that the background load is calculated using the "regularload" array, which is defined internally in the class's construct methods. Therefore it seams that I can't change it. Is this correct? If so, I would suggest to modify the code to allow this array to me customized. Some options are: 1. make it a protected field and initialize it in the declaration so that subclasses could redefine it. 2. Pass it as a parameter to the constructor (a better option) Regards -- ======================================================================== 0 0 0 Pablo Chacin | Dept. d'Arquitectura de Computadors 0 0 0 e-mail: pc...@ac... | UPC-Campus Nord. Modul D6-212 0 0 0 phone: +34 +93 405 4059 | Jordi Girona, 1-3 U P C fax: +34 +93 401 7055 | 08034 Barcelona - SPAIN www: http://personals.ac.upc.es/pchacin ======================================================================== |