> matrix_df hotmail wrote:
> > Hello Jan
> >
> > Ok, the overlong round trip times (rtt) from slave to
> > master on my system were introduced by the switch (Longshine
> > LCS-FS6108).
>
> Oh yeah. And there are people that still think real-time
> problems can be solved by attaching arbitrary switches.
>
> > For normal TCP/IP traffic the Longshine LCS-FS6108 switch
> > works. I have 2 of them. So the risk that
> > both are out of order is relatively small.
> > Do you know switches witch do not work?
>
> I can only say that using unmanaged and preferably only 8-port
> switches are known to introduce small bounded delays (we once
> measured ~10 us). They are simply too dumb to cause more harm
> ;) (more precisely: they are realised completely in hardware).
> The more intelligence they show, the more likely are rare but
> unexpectedly high latencies.
It is too bad only bridges can be found nowadays.
It is very hard to find real class 2 repeaters (which do no
store-and-forward), or they are overpriced ("industrial" hubs),
and anyway repeaters for gigabit Ethernet seem to never have
existed.
My solution is to buy motherboards with as many PCI slots as
possible, and to put many network cards, and to connect PCs
using cross cables. I think that this is even cheaper than
buying good bridges, and the latency is unbeatable (at least on
the network between interfaces).
--
Romain LENGLET
-------------------------------------------------------
This SF.Net email is sponsored by xPML, a groundbreaking scripting language
that extends applications into web and mobile media. Attend the live webcast
and join the prime developer group breaking into this new coding territory!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=121642
_______________________________________________
RTnet-users mailing list
RTnet-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/rtnet-users
