Re: [rtl] Networking With RT-Linux

[Jens Michaelsen]

Hi members

did someone use / does someone know about:

SCSI Networking??
For example 2 linux boxes using one harddisk via SCSI.
Is it possible to use this bus to talk from linux box to the other??

Jens Michaelsen
(!!my compuserve account is no longer valid!!)
( only [EMAIL PROTECTED] will work )
 

Peter Lindener schrieb:

 Hi
     RT Networking folks -

     Yes the people involve in this dialogue are for the most part fairly knowledgeable about networking topics
relating to real time performance issues.   But then I am also finding this discussion to be rather long winded
with respect to getting any place in particular.  ????

    While I am watching this dialogue go by, I am also laying the road map for Hard Real-time on IEEE-1394.
I'm wondering if any one out there is interested in pointing this dialogue towards building the open real-time networking infrastructure that
we are all wishing for ???

    See correspondence below for our immediate plans,
We would be happy to carry on dialog with respect to additional open project planning.
 

     Best Regards
          -Peter Lindener    Home Office Ph# : (650) 694-7352

--------------

Subject: 
               RTLinux support ?
         Date: 
               Tue, 23 Nov 1999 10:24:45 -0800
        From: 
               Peter Lindener <[EMAIL PROTECTED]>
 Organization: 
               Lindener & Control Associates
           To: 
               [EMAIL PROTECTED]




Hi
   Folks at FSM_Labs --

   We will be doing open source development work with RT-Linux and IEEE-1394's hard real-time IsoSynchronous network communications,
using the TI LV23 OHCI compatible device.

    See : http://www.ti.com/sc/docs/products/analog/tsb12lv23.html

  Our first task will be to make the 1394 hardware device directly accessible at the hard real time level, in addition to
still providing the 1394 services to the non real-time Linux side of the world.

    IEEE-1394 FireWire on Linux   See : http://eclipt.uni-klu.ac.at/ieee1394/main.html

   Then our next goal will be to slave the time bases of the machines on our labs 1394 network, to the Iso synchronous clock time stamp is
shared
between these machines.  After that our plans are to implement a rudimentary publish / subscribe (NDDS like) network for the communication
of low
latency real-time servo related variables.

   For info on NDDS see : http://www.rti.com/products/ndds/NDDS.html

    We are also looking at what it will take to develop RT-Linux drivers for the DSCC4 high performance serial device
from Infineon ( previously Siemens )

    See : http://www.infineon.com/products/commics/dscc4/dsc4_b.pdf

    Please indicate the potential role(s) your firm could play in supporting such efforts.

       Sincerely Yours
            --Peter Lindener

--------------

Guilherme Nelson F De Souza wrote:

>From: "Dresner, Norman A." <[EMAIL PROTECTED]>
>Subject: RE: [rtl] Networking With RT-Linux
>Date: Tue, 23 Nov 1999 11:56:54 -0500
>
>       There are a large class of networks that are designed to have
>"errors", networks like Ethernet that work by collision-detection.
>
>       But there are a larger class of networks that have "no errors"
>during normal operation, for example networks that work on master-slave
>protocols and token-ring networks.  {Also MIL-1553x MUXBUS networks which
>I'm heavily involved with}

   That's definitely NOT true. Any communication protocol carries an
intrinsic non-zero error rate (as was well said before). Token-ring
networks can be used where to allow a deterministic behavior, but that
has nothing to do with the error rate.
   The collision detection in CSMA/CD (ethernet for one) networks just
makes it non-deterministic, but the error rate can be smaller (or greater)
than a token-ring. That's a complete different issue.