Re: [ns] timer granularity

[Di-Fa Chang <difac@usc.edu>]

On Fri, 30 Nov 2001, Michael Savoric wrote:

> Hi,
> am I right that the new default minimum timeout timer
> (minrto_) in tcp is one second?
>
> I have changed the tcp_tick_ variable to 0.01 to reach a
> finer granularity for my timer measurements. But there can
> be a problem if the minrto_ variable is set to 0 (the
> earlier default value).
> Then the minimum timeout value is 2*tcp_tick_ (=0.02 s
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
How you calculate this? read from the minrto_ variable?


> in my simulation). Shouldn't be the minimum timeout
> value independent from tcp_tick_?
> I think that this is a small bug in ns.
>
> Best regards,
> Michael Savoric
>
> ======================================================
> E-Mail:		savoric@ee.tu-berlin.de
>
> Phone:	 	(030)314-23840
>
> Postal address:	Technical University Berlin
> 		Telecommunication Networks Group (TKN)
> 		Einsteinufer 25, 10587 Berlin
> ======================================================
>