Understanding the UtilizationModelFull

[Manoel Campos]

Hello, I want to understand clearly the UtilizationModelFull class.

Does the getUtilization method indicate that

at the specified time, the cloudlet

will use 100% of the required resource?

Thank you.

[Mario Henrique Souza Pardo]

Hello Manoel Campos,

Yeah buddy, that's the deal.

UtilizationModelFull instructs the simulator to pick up 100% from CPUs requested by the task. For instance, if a cloudlet have 10.000 MIPS (length) and request 2 CPUs (PEs) of 1.000 each one, the task will perform in 5 seconds, because it's using full utilization of those 2 CPUs.

Regards!

Mário Pardo
ICMC-USP / São Carlos / Brazil 
mhp...@icmc.usp.br

--
 
---
You received this message because you are subscribed to the Google Groups "cloudsim" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cloudsim+u...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[cloud research]

Hi all,

In Cloudsim API, along with UtilizationModelFull we also find

UtilizationModelPlanetLabInMemory = Used when we use PlanetLab Workload

UtilizationModelStochastic => it utilizes the cpu randomly

1) Is this true. / What exactly do they mean. Can anyone explain with an example please.

2) And what is the meaning of UtilizationModelNull. It says, "the cloudlet expects zero capacity"

Always a cloudlet needs some CPU resources for it's work. Then why will a cloudlet expect zero capacity.

3) How to decide what is the best option for our simulation.( As far as I observed, most of the examples in cloudsim go with UtilizationModelFull.

    can anyone mention where the other models are used.)

Can anyone explain please.

Thanks and Regards.

VGR.

[Mario Henrique Souza Pardo]

Hello VGR,

   I will try to explain your doubts with a simple approach.

   The UtilizationModel classes are done to create the desired behavior of tasks (cloudlets) into the Cloud Computing System environment (i.e., it makes different result effects on CloudSim output report). 

1.   UtilizationModelFull - emulate the behavior of tasks that uses 100% of CPU while its execution.

   UtilizationModelPlanetLabInMemory - emulate the behavior of tasks performed in big computing system (PlanetLab) based on its workload log.

   UtilizationModelStochastic - uses a stochastic model to create tasks that perform on CPU varying the CPU % Usage while its execution.

2. UtilizationModelNull - emulate a task that uses ZERO % of CPU Usage. It's often used to simulate NOP (No Operation Process) or other kinds of situation where you need a tasks to fill space in VM Queue and don't use CPU and don't consumes energy.

3. The best model to use depends of application that you considers in your work. For example, for HPC applications UtilizationModelFull is frequently used. 

Regards,

Mário Henrique de Souza Pardo / PhD Candidate

University of São Paulo / Institute of Mathematical and Computer Science (ICMC)

Laboratory of Distributed Systems and Concurrent Programming (LaSDPC)

São Carlos / SP / Brazil

mhp...@icmc.usp.br

Research Gate Profile [LINK]

 

--

---
You received this message because you are subscribed to the Google Groups "cloudsim" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cloudsim+u...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

[Hoda Faregh]

Hi,
I m looking for web app and Hpc app workload for a long time ,but I could find any thing .
web App has bursty behavior in nature in cpu utilization .
do you have an idea to generate such cpu utilization model ?or can you introduce me a site for web app trace?