Greetings,

I have read the papers regarding iFogSim, but I would like to have a deeper explanation about how periodicity, selectivity models and migration work in the simulation:

1. In which cases would only one application edge be periodic? "Shouldn't" all the application edges be configured as periodic, since the whole environment depends on the sensors?
2. Could the selectivity model be read as a chance of losing the packet?
3. When a migration happens in an intracluster mobility scenario, does migration to the nearest fog node happen every time an edge device moves?
4. In the simulator, is it possible to change the migration policy for intercluster mobility scenarios; for example the broker of the requests to be another gateway instead of a proxy?

Thanks in advance.

I am getting the below error while running the java files using the mobility dataset

at org.fog.mobilitydata.DataParser.parseResourceData(DataParser.java:99) at org.fog.placement.LocationHandler.parseResourceInfo(LocationHandler.java:150) at org.fog.test.perfeval.CardiovascularHealthMonitoringApplication.createFogDevices(CardiovascularHealthMonitoringApplication.java:162) at org.fog.test.perfeval.CardiovascularHealthMonitoringApplication.main(CardiovascularHealthMonitoringApplication.java:109)

I am getting the below error while running TranslationServiceFog_Clustering.java, I have all the jars properly added inclusing math3.

Kindly help me out.

Starting Translation Service... Exception in thread "main" java.lang.NoClassDefFoundError: org/apache/commons/math3/util/Pair at org.fog.application.Application.addTupleMapping(Application.java:135) at org.fog.test.perfeval.TranslationServiceFog_Clustering.createApplication(TranslationServiceFog_Clustering.java:317) at org.fog.test.perfeval.TranslationServiceFog_Clustering.main(TranslationServiceFog_Clustering.java:75) Caused by: java.lang.ClassNotFoundException: org.apache.commons.math3.util.Pair at java.base/jdk.internal.loader.BuiltinClassLoader.loadClass(BuiltinClassLoader.java:641) at java.base/jdk.internal.loader.ClassLoaders$AppClassLoader.loadClass(ClassLoaders.java:188) at java.base/java.lang.ClassLoader.loadClass(ClassLoader.java:520) ... 3 more

First, thank you for your great sim. I'm working with your sim for a while for academic purposes. I have this question : can we use your sim for Context-aware Data and Task Placement as in this article?

https://ieeexplore.ieee.org/document/8767386

You can fix the dependencies that they don't have by adding the following code to the pom.xml and then right click-> Maven0->Update Project or select project and Alt+F5

<repositories>
       <repository>
       		<id>WSO2</id>
            <url>http://dist.wso2.org/maven2/</url>
        </repository>
  </repositories>
  
  <dependencies>
	  <dependency>
	    <groupId>org.apache.commons</groupId>
	    <artifactId>commons-math3</artifactId>
	    <version>3.6.1</version>
	</dependency>
		<!-- https://mvnrepository.com/artifact/org.apache.clerezza.ext/org.json.simple -->
		<dependency>
		    <groupId>org.apache.clerezza.ext</groupId>
		    <artifactId>org.json.simple</artifactId>
		    <version>0.4</version>
		</dependency>
			<!-- https://mvnrepository.com/artifact/com.google.common/google-collect -->
	<dependency>
	    <groupId>com.google.common</groupId>
	    <artifactId>google-collect</artifactId>
	    <version>1.0-rc1</version>
	</dependency>
	</dependencies>

However, the program still fails to run

When vrgameFog.java is executed, the following error message is displayed.....

Exception in thread "main" java.lang.Error: Unresolved compliation problems: Log cnanot be resolved Log cnanot be resolved The method getId() is undefined for the type FogBroker The method getId() is undefined for the type FogBroker The method getId() is undefined for the type FogBroker The method getName() is undefined for the type FogDevice The method getName() is undefined for the type FogDevice CloudSim cannot be resolved CloudSim cannot be resolved Log cnanot be resolved

Hi,

When running examples CardiovascularHealthMonitoringApplication.java and CardiovascularHealthMonitoringApplicationEdgewards.java, Mac OS seems to be getting different results to a Windows system.

I have included a zip with the outputs I receive in a Windows VM and Mac OS. Once I open the simulator in IntelliJ or Eclipse, and run those files for the first time, those are the outputs I receive.

Windows appears to be working well (I included images of partial output because the output is very large). But on Mac OS (Version 11.5.1), there are errors. How do I solve this so I don't have to use the VM?

Thanks. output.zip

When updating the other VM processing, first, it sets the new mipsshare of the VM In the following line: https://github.com/Cloudslab/iFogSim/blob/fca02da0132d155754373402e72ccaed5aca3d71/src/org/cloudbus/cloudsim/CloudletSchedulerTimeShared.java#L65

Then, it calculates the FinishedSoFar in line 70: https://github.com/Cloudslab/iFogSim/blob/fca02da0132d155754373402e72ccaed5aca3d71/src/org/cloudbus/cloudsim/CloudletSchedulerTimeShared.java#L70

While I think, First, it should calculate FinishedSoFar(which has happened based on the previous mipshare of the VM), then update the mipsshare of the VM (to calculate the processing time of the remaining length based on the new mipsshare).

Would you please check this and let me know if we should move line 70 after that?

Best, Saeedeh

Me and my team are working on a novel load balancing algorithm for fog computing. iFogSim2 has introduced a new way to bring in microservices into the application and readily coding a load balancer for it in LoadBalancer.java file which was not present in iFogSim. What does this load balancer do exactly and why is it so specific to microservices?

Also is there any difference between a microservice and a module in ifogsim other than the additional features like load balancers, service discovery etc implemented for it? Because logically when I am creating an application, I can divide it either into smaller modules or into smaller microservices, but ifogsim2 provides more fine grained control when I create microservices.

Thanks in advance!

In redme.md file of iFogSim2 mentioned that "It is fully compatible with the Latest Version of the CloudSim (i.e., CloudSim 5)" but ClodSim5 supports containers and I can't find any packages like "org.cloudbus.cloudsim.container.schedulers" and others in iFogSim2. Only the jar files of CloudSim version 3.0.3 are added to reference libraries.

Hi, I am trying to make the ifogsim and ifogsim2 running. They both have the same error messages as:

Starting DCNS... Exception in thread "main" java.lang.Error: Unresolved compilation problems: The constructor AppModule(int, String, String, int, double, int, long, long, String, CloudletScheduler, Map<Pair<String,String>,SelectivityModel>) refers to the missing type Pair Pair cannot be resolved to a type

at ifogsim_1/org.fog.application.Application.addAppModule(Application.java:65)
at ifogsim_1/org.fog.test.perfeval.DCNSFog.createApplication(DCNSFog.java:233)
at ifogsim_1/org.fog.test.perfeval.DCNSFog.main(DCNSFog.java:68)

I am using JDK13.0.2.

Updates made with reference to Issue #5. Corrects the typo in MicroservicesController.java and enables static clustering for dynamic scheduling scenarios thereafter.

Hello,

In the org.fog.placement.MicroservicesController class, the two init() methods operate differently on the Config.ENABLE_STATIC_CLUSTERING value.

In particular:

org.fog.placement.MicroservicesController::init() checks for: if (Config.ENABLE_STATIC_CLUSTERING)

where

org.fog.placement.MicroservicesController::init(Map<Integer, List> monitored) checks the opposite:

if (!Config.ENABLE_STATIC_CLUSTERING)

Looking at the rest of the implementation, it seems likely a typo in one of them. In this case, which one is the "correct" one? On the other hand, if this is intentional, could you elaborate the reasoning behind this difference between the two methods?

Thanks, Emmanouil Maroudas