Jump to content

Server Idle Time Experiment


g1h2j3
 Share

Recommended Posts

As a response value for my experiments i want the total idle time of all servers during the experiment. But i have around 100 servers and don't want to type for every server a expression like "Server6.ResourceState.TotalTime( 0 )" and sum them up. Is there an easier way? I would like to use only one expression which gives my the idle time of all servers. Something like All_servers.ResourceState.TotalTime(0).

Link to comment
Share on other sites

Hi willem,


thank you very much again for your answer. I am not getting it,sorry :( . So i implemented an ObjectList. So far so good.

Then i should create a process, which is initialized in the beginning right? This process has first a search step which searches through my object list. I leave Match and Search condition empty? I am not so familiar with the search step. I can search for the idle servers but then i need to enter the expression again -.-. I fear you need to explain it to me in more detail. I don't see the solution. Sorry for that.

Link to comment
Share on other sites

Hi willem,


thanks a lot!

Can i ask you why the process is OnRunEnding? Or is that necessary to record it during the simulation? Would it work when i initialize it in the beginning?

I did the experiments and i get slightly different values. But they should be the same right? Or can there be a deviation?

Thank you very much again! I appreciate your time!

Unbenannt.JPG.27c39fed3e441b8f80908c2bcc720560.JPG

Unbenannt1.JPG.7446220f1a4e60f0e2811fc15118e4e0.JPG

Link to comment
Share on other sites

Hi,


The idle time is tallied over the whole run, it should be written out at the end of the run as it takes total time into consideration (better stat to work with). In other words it really does record during the run.


I really can't see how it works at the beginning of the run, but the random streams might be responsible for the deviation.


Regards,

Willem

Link to comment
Share on other sites

Hi willem,


thanks a lot. So I tried now several runs and models and there is always a small deviation but really not big, so its ok. But still kind of strange.

Your method is much easier to implement so i will use it for sure. Thanks!!

Link to comment
Share on other sites

×
×
  • Create New...