Simulation Time vs. “Wall-Clock” Time

[Greg]

Hello:

I am trying to understand the relationship between the simulation time
and the physical (or wall-clock) time used to represent a simulation
within OMNet++/OMNEST.

I am simulation a network topology with an increasing number of
nodes. Each simulation run is limited to 600 simulation seconds by
the sim-time-limit configuration option. As expected, as the number
of nodes increases, the setup time of the topology within the
simulation also increases before the actual simulation begins (i.e.,
message passed node-to-node). When I examine simulation results I see
that the simulation was ran for 600s with an elapsed (wall-clock) time
of, for example, of 78.282s.

Could someone please provide an explanation of this please? Is there
an equation/formula that is used by OMNet++/OMNEST to relate these two
variables within the simulator?

Thank you,

Greg

[Alfonso Ariza]

http://en.wikipedia.org/wiki/Discrete_event_simulation


--------------------------------------------------
From: "Greg" <greg.a...@gmail.com>
Sent: Monday, July 27, 2009 7:17 PM
To: "omnetpp" <omn...@googlegroups.com>
Subject: [Omnetpp-l] Simulation Time vs. “Wall-Clock” Time

[Greg]

Okay. Thank you very much. That provides an explanation of the
simulation time and the physical (or wall-clock) time.

As stated previously, as the number of nodes increases, the setup time

of the topology within the simulation also increases before the actual

simulation begins (i.e., message passed node-to-node). What would be
the cause of such an increase in setup time of the topology?

Thank you,

Greg

On Jul 27, 1:20 pm, "Alfonso Ariza" <aariza...@hotmail.com> wrote:
> http://en.wikipedia.org/wiki/Discrete_event_simulation
>
> --------------------------------------------------

> From: "Greg" <greg.a.abb...@gmail.com>

> Sent: Monday, July 27, 2009 7:17 PM
> To: "omnetpp" <omn...@googlegroups.com>
> Subject: [Omnetpp-l] Simulation Time vs. “Wall-Clock” Time
>
>
>
>
>
> > Hello:
>
> > I am trying to understand the relationship between the simulation time
> > and the physical (or wall-clock) time used to represent a simulation
> > within OMNet++/OMNEST.
>
> > I am simulation a network topology with an increasing number of
> > nodes.  Each simulation run is limited to 600 simulation seconds by
> > the sim-time-limit configuration option.  As expected, as the number
> > of nodes increases, the setup time of the topology within the
> > simulation also increases before the actual simulation begins (i.e.,
> > message passed node-to-node).  When I examine simulation results I see
> > that the simulation was ran for 600s with an elapsed (wall-clock) time
> > of, for example, of 78.282s.
>
> > Could someone please provide an explanation of this please?  Is there
> > an equation/formula that is used by OMNet++/OMNEST to relate these two
> > variables within the simulator?
>
> > Thank you,
>

> > Greg- Hide quoted text -
>
> - Show quoted text -

[Alfonso Ariza]

Remember that in a discrete event simulator the computer time consume is
processing events. If the numbers of events grows the timer that the
computer needs to process the events grows also.

Explaining easily
More nodes->more modules->more source events ->more events->more computer
time needed for to process the events->simulation longer


--------------------------------------------------
From: "Greg" <greg.a...@gmail.com>
Sent: Monday, July 27, 2009 8:14 PM
To: "omnetpp" <omn...@googlegroups.com>
Subject: [Omnetpp-l] Re: Simulation Time vs. “Wall-Clock” Time