Archived Commentary about Serial-Line Terminal Servers and Related Equipment;
"Terminal Concentrators" is another name.

 //////////////////////////////////////////////////////////////////////////////

As of 2000-04-25, some explanation of how to use a "Digi" terminal server
to achieve serial connections over a TCP/IP network may be seen at:

    http://www.aplawrence.com/Reviews/portserver.shtml

 //////////////////////////////////////////////////////////////////////////////

Newsgroups: comp.os.linux.networking, comp.unix.solaris
References: <b2usld$edg$1@hfc.pacific.net.hk>
    <OpF4a.173329$be.157913@rwcrnsc53>
Message-ID: <slrnb56d6l.r3c.sithglan@faui00u.informatik.uni-erlangen.de>
Organization: University of Erlangen, Germany
Date: 19 Feb 2003 07:40:05 GMT
From: Thomas Glanzmann <sithglan@stud.uni-erlangen.de>
Subject: Re: console server

> Unless you are doing this for the fun of it, I'd go with the
> Cyclades units.  They'll sell you cables and adapters for
> 32pin/9pin/Netra-Cisco RJ45.

Cyclades are too expensive. But if you want to spend the money it is a
good choice. But if you go with cyclades, don't put the terminal server
in a public network ... their ssh was a long time vulnerable. So I would
put the cyclades terminal server in a mgmt network.

There is only one benefit in cyclades that other don't have: They don't
send unmeant breaks.

Greetings,
        Thomas
--
Thomas Glanzmann  ++49 (0) 9131 85-27574   Department of Computer Science III
Martensstrasse 3  D-91058 Erlangen Germany   University of Erlangen-Nuremberg
            http://www3.informatik.uni-erlangen.de/Research/UMLinux/

 //////////////////////////////////////////////////////////////////////////////

Newsgroups: comp.unix.solaris
Message-ID: <R5Ena.141$Cl7.108@newssvr16.news.prodigy.com>
References: <3E9CA6E8.9060700@qwest.no.you.dont.net>
    <rS4na.26$sy6.16454807@newssvr11.news.prodigy.com>
Date: Thu, 17 Apr 2003 20:26:25 GMT
From: Darren Dunham <ddunham@redwood.taos.com>
Subject: Re: Pin outs

Nick Allen <jnallen@swbell.net> wrote:

> The db25 jack on the box is just a standard serial port, not knowing much
> about the Nortel pinout (DCE/DTE???), I would start with a straight
> connection, and try a null-modem adapter if needed.  The port default is
> 9600 8n1, unless you've changed it...  The RSC board RJ-45 is the same
> pinout as a cisco async port.  I beleive the pinouts are covered in the
> books that come with the box.  But I think they are:
>
> 1 rts
> 2 dtr
> 3 tx
> 4 gnd
> 5 gnd
> 6 rx
> 7 dsr
> 9 cts

> Pins 1->8 right to left opposite the clip.

If this "Nortel" device is the standard old Annex terminal server, then
the pinouts are definitely not the same.  Last time I needed this I was
able to find some docs on google or something.

I'm certain that at least one of the RX/TX pins are on 4 or 5.  I was
able to make RJ-45 <-> RJ-45 cables to go between them, but the cables
were not symmetric.

-- 
Darren Dunham                                           ddunham@taos.com
Unix System Administrator                    Taos - The SysAdmin Company
Got some Dr Pepper?                           San Francisco, CA bay area
         < This line left intentionally blank to confuse you. >

 //////////////////////////////////////////////////////////////////////////////

Newsgroups: comp.unix.solaris
Message-ID: <3EA0CCCA.6070603@qwest.net>
References: <3E9CA6E8.9060700@qwest.no.you.dont.net>
    <rS4na.26$sy6.16454807@newssvr11.news.prodigy.com>
    <R5Ena.141$Cl7.108@newssvr16.news.prodigy.com>
    <anMna.407$Oc3.35200@newssvr28.news.prodigy.com>
Date: Fri, 18 Apr 2003 21:12:58 -0700
From: Karl E. Taylor <ktayloraz@qwest.net>
Subject: Re: Pin outs

Nick Allen wrote:
> "Darren Dunham" <ddunham@redwood.taos.com> wrote in message
> news:R5Ena.141$Cl7.108@newssvr16.news.prodigy.com...
> 
>>Nick Allen <jnallen@swbell.net> wrote:
>>
>>>The db25 jack on the box is just a standard serial port, not knowing much
>>>about the Nortel pinout (DCE/DTE???), I would start with a straight
>>>connection, and try a null-modem adapter if needed.  The port default is
>>>9600 8n1, unless you've changed it...  The RSC board RJ-45 is the same
>>>pinout as a cisco async port.  I beleive the pinouts are covered in the
>>>books that come with the box.  But I think the are:
>>
>>>1 rts
>>>2 dtr
>>>3 tx
>>>4 gnd
>>>5 gnd
>>>6 rx
>>>7 dsr
>>>9 cts
>>
>>>Pins 1->8 right to left opposite the clip.
>>
>>
>> If this "Nortel" device is the standard old Annex terminal server, then
>> the pinouts are definitely not the same.  Last time I needed this I was
>> able to find some docs on google or something.
>>
>> I'm certain that at least one of the RX/TX pins are on 4 or 5.  I was
>> able to make RJ-45 <-> RJ-45 cables to go between them, but the cables
>> were not symmetric.
>>
>>-- 
>>Darren Dunham                                           ddunham@taos.com
>>Unix System Administrator                    Taos - The SysAdmin Company
>>Got some Dr Pepper?                           San Francisco, CA bay area
>
>
>
>
> Hey, I remember those.  Did Nortel buy the old Annex line? 
> Anyway, check this site out:
> 
>     http://www.conserver.com/consoles/annexcons.html
> 
> If that is an old Annex, this page might help :-)
> 
> Nick


That page was just what I was looking for.

Thanks, Nick.
-- 
Karl E. Taylor
UNIX Geek

 //////////////////////////////////////////////////////////////////////////////

Newsgroups: comp.unix.solaris
References: <3E9CA6E8.9060700@qwest.no.you.dont.net>
Message-ID: <b7kgqa$1ulfe$1@ID-128882.news.dfncis.de>
Organization: San Francisco, CA
Date: 16 Apr 2003 21:09:30 GMT
From: Andrei Ivanov <iva@racoon.riga.lv>
Subject: Re: Pin outs

Karl E. Taylor <ktayloraz@qwest.no.you.dont.net> wrote:
> 
> Does any one here have, or know where I can get, the pin outs for a Sun 
> v880 db25 to rj45 connection for a Nortel Networks terminal concentrator?

Although I've never used Nortel concentrators, I'm using Cisco's
DB25(male)-to-RJ45 ("MODEM CAB-5MODCM", 29-0881-01) adapter and
rolled cable to connect Sun's serial ports to Livingston Portmaster.

    http://www.cisco.com/warp/public/793/access_dial/sun_comm_server.html
    http://www.cisco.com/warp/public/701/14.html#Rolled

-- 
andrei

 //////////////////////////////////////////////////////////////////////////////

Date: Wed, 07 May 2003 09:35:11 -0700
From: Richard Elling
Subject: Re: Question about terminal concentrator (TC)

The easiest TCs are those with Cisco/Sun standard RJ-45 pinouts. Most
vendors are using that standard today.

The Annex uses a different pinout, so you end up building cables.  :-(.

There are several good ones out there from Cisco, Cyclades, and
Lantronix. My concern about Cyclades and Lantronix is global
availability and support.  But in any case, you can find specifications
from all of the vendors.

 -- richard

 //////////////////////////////////////////////////////////////////////////////

Date: Wed, 18 Jun 2003 11:16:27 -0600
From: Chuck Cox
Subject: Re: Terminal Server aka Access Concentrator

1) Annex/MicroAnnex terminal servers.  Originally made by Xylogics, later 
bought by Bay Networks.  Discontinued a while ago, but still pretty reliable.

2) Cisco 2511 or 2600/3600-series routers with Async cards.  These are Cisco 
routers with 16 to 192 asynchronous serial ports in them.  They were meant to 
have modems attached and be used as dialup servers, but they work just as well 
"backwards" - i.e., connect servers' consoles to the aysnc ports, then 
telnet/ssh through the router to the async ports.

3) Cyclades TS-series terminal servers.  Basically a rack-mountable Linux box 
with multiport serial card built in.  We have been using the TS3000 for a year 
or so - it has 48 ports but only requires 1 rack unit.  It supports ssh 
natively.  Since it's Linux, you could install other software on it for 
authentication methods (like Radius/Safeword/SecurID) or logging/reporting. 
The TS-series have recently been replaced by the ACS series - basically the 
same box with some new features.

4) Aurora Technologies multiport serial cards and ControlTower software. 
Aurora makes 16-port serial cards (Sbus, PCI, or cPCI) that can either be 
installed directly into a Sun system or into an external chassis which would 
then be connected to a Sun system.  Maximum density with the external chassis 
would be 112 serial ports per PCI slot.  You can attach multiple expansion 
chassis to one Sun system (like an Enterprise 220R) if you need lots of ports.
The ControlTower software provides logging of any output from the console ports 
(i.e. syslog messages to /dev/console) which is something the Cisco units 
can't do.  Cyclades TS-series don't have this but the new ACS-series do.

Of course, there are other console server devices out there.  These just 
happen to be the ones that I have used/am using.

Chuck

 //////////////////////////////////////////////////////////////////////////////

Newsgroups: comp.unix.solaris
NNTP-Posting-Host: 10.162.16.177
NNTP-Posting-Date: Sun, 29 May 2005 15:21:30 EET DST
References: <3fb4qrF6stbvU3@individual.net>
Message-ID: <etime.580$Nb2.9288@news1.nokia.com>
Organization: Nokia
Date: Sun, 29 May 2005 12:21:30 GMT
From: mario <gianmarioRE.scottiMO@nokiaVE.com>
Subject: Re: Serial Console

"Cesare_72" <cesare_72.N.O.S.P.A.M.@hotmail.it> wrote in message
news:3fb4qrF6stbvU3@individual.net...
> Hy,
>
> what it's the best console server for managing lots of SunServer? I'm
> evaluating Cyclades produt (cheaper, smallest, easy to install/use) but
> there are other product over the market?
>
> Thanks.
>
> Cesare


We use Lantronix extensively and they have been working perfectly well for
several years.  These console servers have been working 24/7 for as long as
I can remember, 5 or 6 years at least, and have never caused us any problem.
If I was to decide on reliability, Lantronix would be my choice. Also, no
problem with speed or accidental break signals.

But as people have said, newer Sun boxes come with ALOM or RSC, so the need
for console servers is reduced.

-- 
mario 

 /////////////////////////////////////////////////////////////////////////////

Message-ID: <827hd5$1bo$1@triton.dnai.com>
References: <VRw14.7$NG5.8760@news>
Date: 2 Dec 1999 20:36:53 -0800
Newsgroups: comp.sys.sun.admin
From: Lon Stowell <lstowell@triton.dnai.com>
Subject: Re: Console/Terminal Server Question

 w <swfa@hotmail.com> wrote:
>Hello,
>
>We are trying to setup a Console/Terminal server for our SUN UNIX machines.
>The Black Box product "Terminal Server" promised in the advert to be the
>ideal Console/Terminal Server product. The problem is that a "break signal"
>is transmitted on the serial connection when ever we reboot the
>Console/Terminal box, shutting our SUN UNIX machines down. Black Box
>could/would not help us with a solution to solve this problem.
>
>Does anyone know how to get around this problem? Is it possible to setup the
>SUN UNIX machines to avoid the shutdown or are we forced to send back the
>product to Black Box?


   There is a patch finally available for Solaris 2.6 and Solaris 7
   that changes the stop sequence from a break.   

   The patch is a new keyboard driver that adds a new
   option "alternate" to the kbd command in addition
   to the older 'kbd -a enable' and 'kbd -a disable'. 

   Sorry, not where I can get on sunsolve at the moment
   to provide the exact patch number, it was released
   just this past summer.    

   SOME terminal servers don't really send a break, they
   just have sloppy power supplies driving their EIA
   chips.  SOMETIMES you can deal with these by adding
   a 4.7k ohm resistor between DTR and RX Data right
   at the back of the sun port [use a breakout and try 
   it.

 ..............................................................................

Newsgroups: comp.unix.solaris,comp.unix.admin
Message-ID: <xoavhf6syapt.fsf@sun.com>
References: <39DA6DA0.F7605FCA@hotmail.com> <8rdrcs$ptr$1@ncal.verio.com>
    <39DB30BB.8B600114@hotmail.com>
NNTP-Posting-Host: phorcys.east.sun.com
Date: 04 Oct 2000 09:53:02 -0400
Organization: Sun Microsystems Inc. - BDC
From: James Carlson <james.d.carlson@sun.com>
Subject: Re: Solaris crashing when keyboard unplugged.


  [ see the file  /etc/default/kbd and do "man kbd" ]


Kevin Halverson <bunnyman77@hotmail.com> writes:
>
>   If the keyboard has been disabled, how will I be able to type the
>   "kbd -a" enable command?

Check the man page.  The "-a disable" option merely disables keyboard
break.  It doesn't disable typing on the keyboard.


% man kbd
     .
     .
     .

     -a enable/disable/alternate state
           Enable,  disable,  or  alternate  the  keyboard  abort
           sequence effect. By default, a keyboard abort sequence
           (typically Stop-A or L1-A on the keyboard and BREAK on
           the serial console device) suspends the operating sys-
           tem on most systems. The default keyboard behavior can
           be  changed  using this option. The -a option can only
           be used by a superuser.

[...]
           disable
                 Disable the default/alternate effect and  ignore
                 keyboard abort sequences

-- 
James Carlson, Internet Engineering       <james.d.carlson@east.sun.com>
SUN Microsystems / 1 Network Drive         71.234W   Vox +1 781 442 2084
MS UBUR02-212 / Burlington MA 01803-2757   42.497N   Fax +1 781 442 1677
PPP 2nd Edition now available - http://people.ne.mediaone.net/carlson/ppp

 //////////////////////////////////////////////////////////////////////////////

Newsgroups: comp.os.linux.networking, comp.unix.solaris
References: <b2usld$edg$1@hfc.pacific.net.hk>
Message-ID: <OpF4a.173329$be.157913@rwcrnsc53>
Organization: AT&T Broadband
Date: Wed, 19 Feb 2003 06:23:42 GMT
From: Chris Cox <notccox@notairmail.net>
Subject: Re: console server

Brian Leung wrote:
> Hi all,
> 
> I wanna to build a console server with redhat which will be used to connect
> to Sun Enterprises and routers. I have  few questions:
> My PC has 4 serial ports:
> 1. Do i need to buy the "null modem" cables to connect the linux serial port
> to Sun servers' console port and cisco routers?

Yes.  Some Cisco equipment use "special" pinouts.
With that said, the RJ45 to RS232 adapter that came with
your Sun works with Suns with a straight through 8wire (a piece
of Cat-5 cable will do.. but less that that will do of course).

> 2. Did minicom work fine in my project?

I use minicom to directly access Sun equipment through the serial
port all of the time.  However, I also use Cyclade console servers
when possible (it runs Linux and is a dedicated console switch
appliance with SSH encryption to the ports).
http://www.cyclades.com

Unless you are doing this for the fun of it, I'd go with the
Cyclades units.  They'll sell you cables and adapters for
32pin/9pin/Netra-Cisco RJ45.


 //////////////////////////////////////////////////////////////////////////////

Newsgroups: comp.unix.solaris
NNTP-Posting-Host: host-65-121-79-248.skisolitude.com [65.121.79.248]
NNTP-Posting-Date: Mon, 30 May 2005 17:37:27 +0000 (UTC)
References: <3fb4qrF6stbvU3@individual.net>  
    <Y12ke.43$jU5.146080@twister.nyc.rr.com>   <3fc4qmF73nmpU1@individual.net> 
     <1116866258.853325.207410@g44g2000cwa.googlegroups.com>  
    <42982831.9090000@gmail.com>
Message-ID: <1117474642.128165.133980@g44g2000cwa.googlegroups.com>
Date: 30 May 2005 10:37:22 -0700
From: Peter Hunt <peter.hunt@opengear.com>
Subject: Re: Serial Console

Cesare <cesare.tensi@gmail.com> wrote:
>
> The main criteria is manageability.
>
> I mean: easy to install and configure, security of connections estabilished
> from the apparatus to remote user, no-send break command in serial line
> (prevent to halt the machine, solution also is to key-lock the machine).


Most of the console servers I have looked at provide web interfaces
(which vary widely in usability) for simplifying configuration of a
single unit, and some even have (expensive) management software for
controlling multiple units globally. We are currently writing
documentation for better integration with conserver...

    http://conserver.com/

...which is a promising open-source solution which can be used for
managing multiple arbitrary console/terminal servers.

Just FYI there is a page at the conserver site which lists servers that
are "accidental break" safe:

    http://conserver.com/consoles/BREAK-off/breakinfo.html

I am currently trying to contact the Conserver people to donate some
hardware their way.

I myself have worked on the web UI for our console server line, and a
large proponent of ease of use and setup. Like most other Console
Servers we provide a command line accessible via SSH or Telnet, which
essentially gives you a Linux bash prompt with access to some custom
tools for configuration / user management.

I would like to know what your preference is  for setting up  the
console servers you have used so far?


> And also the price. I discovered that Cyclades is cheaper and very easy to
> install/deploy on datacenter. It'm my opinion.
>
> And you?

One of the reasons my company was founded is that they don't agree with
the current pricing of the serial console market. The leaders such as
Cyclades / Avocent-Equinox and Raritan are building products from
technologies that have been around for a long time and in my opinion
are over-pricing what is essentially a simple product. One of our hopes
is to lower the bar to enter server management with console servers.

-- 
Peter.Hunt@opengear.com


 //////////////////////////////////////////////////////////////////////////////

Newsgroups: comp.os.linux.networking, comp.unix.solaris
References: <b2usld$edg$1@hfc.pacific.net.hk>
Message-ID: <slrnb56cqk.r3c.sithglan@faui00u.informatik.uni-erlangen.de>
Organization: University of Erlangen, Germany
Date: 19 Feb 2003 07:33:41 GMT
From: Thomas Glanzmann <sithglan@stud.uni-erlangen.de>
Subject: Re: console server

> 1. Do i need to buy the "null modem" cables to connect the linux serial port
> to Sun servers' console port and cisco routers?

Try google and you will find the pin assignment.

> 2. Did minicom work fine in my project?

Minicom would not be your first choice. Use screen. Benefits are:
        one programm to handle all consoles
        multi user can work on the same console or others
        you can give some user read only acess
        it logs (minicom can do this also)

In order to send a break via screen you have to use a special (long enough)
break. I attached our screenrc [1] with that break on CTRL A SHIFT B.

Have a look at the manpage of screen for further informations.

We have two console servers. One for cisco routers and one for our sun / linux
server consoles. One console Server is a Ultra 1 running Sol 9 with two MAGMA
cards (expensive!) and one is a DUAL XEON 2.0 GHZ running Linux with a real
cheap serial CARD (no cyclades ... they are much too expensive). I don't
remeber
the manifacture ... but if you're interested I have a look.

Greetings,
        Thomas
--
Thomas Glanzmann  ++49 (0) 9131 85-27574   Department of Computer Science III
Martensstrasse 3  D-91058 Erlangen Germany   University of Erlangen-Nuremberg
            http://www3.informatik.uni-erlangen.de/Research/UMLinux/

[1] our console screenrc

# don't mess up my xterm title
hardstatus alwaysignore

shell -/bin/false

screen -t faui01  -L 1  /dev/ttyS4  9600
screen -t faui02  -L 2  /dev/ttyS5  9600
screen -t faui03  -L 3  /dev/ttyS6  9600
screen -t faui04a -L 4  /dev/ttyS7  9600
screen -t faui05  -L 5  /dev/ttyS8  9600
screen -t faui07  -L 7  /dev/ttyS9  9600
screen -t openbsd -L 10 /dev/ttyS10 9600

multiuser on

acladd simigern
acladd sithglan
acladd sialgern
acladd sichkoll
acladd jnreinwa
acladd hrschulz

umask +a
defscrollback 5000

# extra long break for sun.
bind B break 3
autodetach on
logfile /local/console/log/screenlog.%t
defwritelock off

startup_message off
caption always '%{-b}%{= wb}%-Lw%{= wr}%{+b}%n%f %t%{= wb}%{-b}%+Lw %{= wr}
%=%c %Y-%m-%d'

 //////////////////////////////////////////////////////////////////////////////

Newsgroups: comp.unix.solaris
NNTP-Posting-Host: 63.204.7.5
NNTP-Posting-Date: Tue, 16 Aug 2005 16:31:13 EDT
References: <1124216779.135401.48520@z14g2000cwz.googlegroups.com>
Message-ID: <l2sMe.42$GV7.37@newssvr25.news.prodigy.net>
Date: Tue, 16 Aug 2005 20:31:13 GMT
From: Darren Dunham <ddunham@redwood.taos.com>
Subject: Re: console server choice

Vahid Moghaddasi <vahid.moghaddasi@gmail.com> wrote:
> Hi all,
> We have so many WTI's (terrible thing) and some Lantronix as console
> servers for our Sun machine.
> I am not too happy with either of them. I guess WTI was never made to
> be a console server and Lantronix is too slow to perform any
> administrative task on it since it has an embedded Linux and slow CPU.


Can you give a few more details of what makes them unsuitable?  I
wouldn't believe a machine could be too slow to be a console server.  

Lots of folks use Cyclades.  ssh support, history, etc.  I've personally
used cisco, xylogics/nortel/bay, digi, IOlan, lightwave, and cyclades.
Only the xylogics (Annex) was one that I found fairly annoying to
administer, but it was ancient.

Console history/logging and the ability to have multiple connections to
a port are wonderful.  Nowadays I'd demand ssh connectivity.

There's no real standard for RJ-45 serial connectivity, so getting
something that's compatible with the Cisco/Sun pinout might be nice.
I've made lots of custom, directional cables for console servers.

Don't trust *any* device to not send a break.  Sure lots of them (now)
are designed not to, but you can always glitch a line to the point that
it looks like a break.  Set up your hosts to use the alternate break
sequence instead.

-- 
Darren Dunham                                           ddunham@taos.com
Senior Technical Consultant         TAOS            http://www.taos.com/
Got some Dr Pepper?                           San Francisco, CA bay area
         < This line left intentionally blank to confuse you. >

 \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
 ===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#==
 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
 ===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#==
 //////////////////////////////////////////////////////////////////////////////

Newsgroups: comp.terminals,comp.sys.dec,vmsnet.networks.misc,comp.os.linux.networking
Organization: University of Tennessee, Knoxville--Dept. of Computer Science
Message-ID: <4kj6b3INN67v@duncan.cs.utk.edu>
References: <4jq536$egp@stratus.skypoint.net> <4kc3c4INNft5@duncan.cs.utk.edu>
            <1996Apr10.115740.24404@indyvax.iupui.edu>
Summary: DEC's terminal servers
Keywords: DEC, terminal server, LAT, I/O, VAX, MOP
Expires: 4 May 1996 22:33:11 EDT
Date: 11 Apr 1996 10:52:19 -0400
From: shuford@cs.utk.edu (Richard Shuford)
Subject: Terminal Servers (was Re: How do you hook a dumb terminal to a PC?)

In article <1996Apr10.115740.24404@indyvax.iupui.edu>,
   mwood@indyvax.iupui.edu (Mark H. Wood) writes:
>
> In article <4kc3c4INNft5@duncan.cs.utk.edu>,
>  shuford@cs.utk.edu (Richard Shuford) writes:
>>
>>  [omitted discussion of how to connect serial terminals to a Linux PC]
>>
>> ...Terminal Server.  There are many vendors
>> for these, including Cisco, Xyplex, Livingston, Xylogics, Lantronix, and
>> even DEC.
>
> ("even DEC"???  If they didn't have the first commercial terminal server,
>  they weren't far behind.  Our 9-year-old DECserver 200s still run and run
>  and run.)


OK, to clarify about DEC:

Yes, as far as I know, Digital Equipment Corporation was the first vendor
to market Ethernet terminal servers.

As I understand the matter, DEC historically had employed I/O "front end"
processors for its large-scale mainframe computers (DECsystem-10 and -20).
I think that small PDP-11 processors were typically used.

When the first VAX "super-minicomputers" were being developed (circa 1977),
the traditional I/O front end was not employed, but the designers saw that
it was still valuable to relieve the main processor from the burden of
managing the serial I/O from user terminals.  So the "DECserver" terminal
servers were developed to buffer up input data from many terminals in use
at once; then, using the LAT protocol, pump a big chunk of input data into
the VAX very quickly through the high-speed Ethernet interface.  (Or vice
versa for output.)

(It was still possible to use a VAX without terminal servers.  I have seen
a low-end VAX-11/750 with 16 terminals connected via DZ-11--which is to say,
primitive--serial interfaces.  The processor in that system spent about half
its time servicing serial interrupts!)

    [And the computational throughput of that 11/750 was 
     noticeably increased when the sysadmin, at my suggestion,
     reduced the serial data rate of the 16 terminals from
     19,200 bps to 4800 bps.]

As I understand it, the DEC-invented LAT (Local Area Transport) protocol
was optimized for just this sort of use.  (The lightweight functionality
which makes it good for local terminals means LAT is not routable and is
prone to timeout restrictions over bridged distant networks.)

To be fair to DEC, LAT terminal servers were invented before TCP/IP became
important to the world.  And while Telnet over TCP/IP is routable and
flexible (and nowadays ubiquitous), Telnet does take up more network
bandwidth for a given setup of a few host computers and many remote terminals.

Now, my justification for saying "even DEC" is that DEC rode a horse that
wore only the proprietary LAT livery for a long time, and other vendors
hurdled into the marketplace with terminal servers that used the open
standard of Telnet over TCP/IP to achieve the same purpose.  (While DEC's
lawyers tried to keep anybody else's horse from wearing LAT colors....
Was it Penril/Datability that first reverse-engineered LAT and took a few
blows from the legal riding crop?)  And, for a long time, you had to have a
VAX/VMS system on the Ethernet so that the DECserver could load its software
using the MOP protocol.  [Very inconvenient for Ultrix shops...]

Yes, now [1996] DEC does have a competitive, interoperable line of terminal
servers.  Most of them can speak both LAT and TCP/IP, and many seem to also
do IPX and AppleTalk.

To find out more about them, you can check the World Wide Web at this URL:

    http://www.networks.digital.com/npb/html/products_guide/cat_toc.html#chap3a

    [as of A.D. 2003, the above URL is long stale]

While we're pointing out Web things, another helpful collection of reference
information on subjects of Data Communications is maintained by "Black Box"
Corporation:

 *  http://www.blackbox.com/bb/refer.html/tigc641

and I modestly mention

    http://www.cs.utk.edu/~shuford/terminal_index.html

-- 
 ...Richard S. Shuford  |"He who oppresses the poor to increase his wealth and
 ...shuford@cs.utk.edu  | he who gives gifts to the rich--both come to poverty.
 ...Info-Stratus contact| Proverbs 22:16


  [* As of 1998, alas, Black Box seems to have discontinued the glossary.
     You could try
          http://www.InstantWeb.com/foldoc/
  ...RSS]


 //////////////////////////////////////////////////////////////////////////////
 ===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#==
 \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\


 [1991 survey]

Newsgroups: comp.dcom.lans,vmsnet.networks.misc,comp.protocols.tcp-ip
Path: utkcs2!emory!swrinde!elroy.jpl.nasa.gov!usc!orion.oac.uci.edu
      !cerritos.edu!arizona.edu!leonard
Message-ID: <1991Oct2.113915.1674@arizona.edu>
Date: 2 Oct 1991 18:39:15 GMT
Organization: University of Arizona
From: leonard@arizona.edu (Aaron Leonard)
Subject: FYI: Results of terminal server evaluation


Greetings fellow networkers,

The University of Arizona has just completed an excruciatingly
protracted Request for Proposal for terminal servers.  (Actually,
we haven't *completed* it in the sense that there are still a
dozen or two bureaucrats' desks across which it still must pass,
but at any rate it's left *my* desk.)

Anyhow, we were looking for a comprehensive terminal server
solution, one that would offer good TCP/IP and LAT access, that
offered flexible wiring options, that was manageable and cost 
effective.  The evaluation process took several months and quite
a bit of personnel time.

So after doing all this work to evaluate these terminal
servers, I figured that it would be a nice gesture to offer our
results to the Net.  Thus I am posting a DRAFT version of our 
selection document.  (The actual dollar amounts have been censored,
but information about relative pricing is included.)

Disclaimers:

1. The closing date for responses on this RFP was in Jan. 1991, so
   some of the product info may be out of date by now.
2. This posting is just me, NOT official University of Arizona or 
   State of Arizona blah blah woof woof.  No applicability or
   suitability is expressed or implied, etc.
3. If you are a losing vendor, I don't want to see any whining mail
   or hear any pleading phone calls.
4. If you are an interested user and have questions about the terminal
   servers, please contact the vendors and not me.

Aaron

Aaron Leonard (AL104), <Leonard@Arizona.EDU>
University of Arizona Telecommunications, Tucson AZ 85721


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

Vendor phone #'s (Arizona-centric):

Xylogics - Sales: Bret Bergman (408)354-4122
Xyplex - Sales: Teresa Lamoureaux (303)277-9229
cisco Systems - Sales: Rick Balli (303)721-0408
Anixter/Synoptics+Racal Interlan - Sales: Tony Peredes (602)624-9945/968-7901
Hughes LAN Systems - Sales: James Marzola (602)820-9155
Synoptics - Sales: Pam Wolters-Mills (602)844-7332
BFA/Datability - Sales: Dick Leger (602)994-5400
Lantronix - Sales: Anthony M. Swanson (800)422-7022
Four Corners/Equinox+Emulex - Sales: Kim Patberg (602)998-4440

 --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --
DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT 

                                  RFP 927-90
                            University of Arizona
                             Summary of Proposals
                               Terminal Servers

                                Aaron Leonard

                              September 24, 1991

1. The Selection.

CCIT-Telecommunications has evaluated a portion of the RFP 927-90.
The results of that evaluation are listed below.  This evaluation
is limited to terminal servers.

The RFP had specified two different categories of terminal servers: 
"Terminal Servers for use with dumb terminals" and "Terminal Servers
with modem leads for use with intelligent devices."  As the
capabilities of the latter are a superset of the former, we have
selected one terminal server, the Xylogics Annex 3, to fulfill both
categories.  Additionally, we have selected another terminal server,
the Xyplex, as an alternative, in order to meet requirements of
specialized applications

The recommended terminal server is the Xylogics Annex 3, as bid by
R-Squared. It met all of the requirements of both categories. 
Additionally,  in the maximal configuration, it has the lowest price
per port of any terminal server bid.  Our evaluation unit performed
with no hardware or software problems.

We award R-Squared the selection for the following Xylogics Annex 3
configurations.  All ports support modem control.

# ports   List price   Discount     Price/Per port   Price/Per port w/ LAT
-------   ----------   --------    ---------------   ---------------------
    8        
   16       
   32      
   64      

The Xyplex MAXserver 1000 terminal server, as bid by High-Tech
Associates, met all requirements of the RFP.  Our evaluation unit
performed with no hardware or software problems, except that it was
not able to support a binary file transfer from a PC to a Unix host.
(See sec. 5 below for details.)  We recommend the Xyplex MAXserver
1000 series for the following applications:

* Standalone units in support of low port density.  At the 16-port
  concentration level, the MAXserver 1000 is somewhat less expensive per 
  port than the Annex 3.
* Applications requiring reverse LAT or LAT application ports.  The Annex 3 
  only supports "forward" LAT (terminal connections).  The Annex cannot
  offer LAT services.
* Applications where a Unix host is not available to perform terminal
  server load services.  The Annex 3 requires that a Unix host be used
  to configure and load the terminal server.  The Xyplex MAXserver 1000
  may be loaded from Unix or VMS hosts, or from local floppy disk.

We select the following Xyplex MAXserver 1000 models as bid by
High-Tech Associates, for use in the above applicatons.  All units 
support limited modem control and have 16 ports.

Model #         LAT  TCP/IP  Memory  Floppy   List price    Price/Per port
----------      ---  ------  ------  ------   ----------    ---------------
MAX1820-E        Y     Y      2MB      Y      
MAX1800-E        Y     Y      1MB      Y      
MAX1520-E        Y     Y      2MB      N      
MAX1500-E        Y     Y      1MB      N      
MAX1120T-E       N     Y      2MB      N      
MAX1100T-E       N     Y      1MB      N      
MAX1120L-E       Y     N      2MB      N      
MAX1100L-E       Y     N      1MB      N          

MX-420-0177 - upgrade MAXserver 11x0 to 15x0      


 --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --
DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT 

2. The selection process.

In response to RFP 927-90, we received thirteen different bids
(vendor/hardware combinations.)  After evaluating the price and
technical specifications of the products bid (which are summarized
in section 3 below), we selected four finalists based upon the
criteria given in section 4 below.  

For each of the four finalists, we asked for and received an
evaluation unit.  Each evaluation unit was installed and configured,
and then tested by Telecom staff, as well as by interested users
in other UofA departments.  The results of the evaluation period
are given in section 5 below.

 --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --
DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT 

3. Summary of bid responses.

Following are the criteria established for the terminal servers
in the RFP:

The following criteria applied to both portions of the terminal 
server request for proposal ("dumb" and "smart")
SNMP                    "should"
TELNET                  "must"
rlogin                  "should"
LAT                     "must"
rackmount               "should"
accounting              "should"
TN3270                  "should"
warranty                the longer the better
price                   the lower the better
delivery schedule       the shorter the better

The following criteria applied to the "smart" portion only:
modem control           "must"
reverse connections     "should"
SLIP/PPP                "should"


The scorecard for each terminal server is given.  Pricepoints
are given for each terminal server (or family of terminal servers)
across the range from the smallest number of ports offered in a
server to the largest.  Where pricing differs for modem control
ports and non-modem control ports, both prices are given.

Vendor #1 - R Squared/Xylogics
SNMP                    yes
TELNET                  yes
rlogin                  yes
LAT                     yes
rackmount               yes
accounting              yes
TN3270                  no
warranty                1 year
delivery schedule       4 weeks
modem control           yes (standard)
reverse connections     yes (TCP/IP only)
SLIP/PPP                yes (SLIP)

Pricing:
Model #         MC?     # ports          Price          Per port (*)
------------    ---     -------         ------          --------
Annex 3         yes         8         
Annex 3         yes        16         
Annex 3         yes        32        
Annex 3         yes        64        
(*) $xx extra per port for LAT

Vendor #2 - High-Tech/Xyplex
SNMP                    yes
TELNET                  yes
rlogin                  yes
LAT                     yes
rackmount               yes
accounting              no
TN3270                  yes (option)
warranty                software: 1 year; hardware: 5 years
delivery schedule       3 weeks
modem control           yes (optional)
reverse connections     yes
SLIP/PPP                yes (SLIP)

Pricing (*):
Model #         MC?     # ports          Price          Per port
------------    ---     -------         ------          --------
MAXserv 1500    yes        16       
  (w/ TN3270)                       
MAX4500         yes        40       
  (w/ TN3270)                       
MAX4500         no         80       
  (w/ TN3270)                       
MAX5000         yes        96       
  (w/ TN3270)                       
                no        192       
  (w/ TN3270)                       
                yes       128       
  (w/ TN3270)                       
                no        256       
  (w/ TN3270)                       

(*) TN3270 option is $xxxxxx/8 ports and $xxx/16 ports.
    MAX4500 (5-slot chassis) is $xxxx.
    MAX5000 (16-slot chassis) is $xxxx.
    TSRVM-J8 (8-port MC board) is $xxxx.
    TSERV-J16 (16-port no MC board) is xxxx.

Vendor #3 - cisco Systems
SNMP                    yes
TELNET                  yes
rlogin                  yes
LAT                     yes (option)
rackmount               yes (ASM & MSM)
accounting              yes
TN3270                  future
warranty                software: 90 day; hardware: 1 year
delivery schedule       60 days
modem control           yes (standard)
reverse connections     yes
SLIP/PPP                yes (SLIP)

Pricing:
Model #         MC?     # ports          Price          Per port(*)
------------    ---     -------         ------          --------
STS-10          yes        10        
MSM             yes        16        
                           32        
ASM             yes        48        
                           96        
(*) $xx extra per port for LAT

Vendor #4 - Xyplex/Xyplex
- Xyplex bid the same product line as High-Tech did, but with 
  higher prices across the board (xx% discount vs. xx%.)

Vendor #5 Anixter/Synoptics
- Anixter bid a SynOptics part #3395 16-port MC line card
for $xxxx.  This part is the same as what SynOptics bid for $xxxxxxx.
Anixter did NOT bid a price for SynOptics hubs, so based upon this 
bid we could not configure a complete terminal server.

Vendor #6 - Anixter/Racal Interlan
- the capabilities of the product could not be discerned from the bid -

Pricing:
Model #         MC?     # ports          Price          Per port
------------    ---     -------         ------          --------
INX 5000-3C     ???        16       
                           48       
INX 5000-12C    ???        96       
                          192       

Vendor #7 - Hughes LAN Systems/HLS
SNMP                    yes
TELNET                  yes
rlogin                  yes
LAT                     yes
rackmount               yes (4296)
accounting              ???
TN3270                  no
warranty                software: 90 days; hardware: 1 year
delivery schedule       4 weeks
modem control           yes (standard)    
reverse connections     LAT only (?)
SLIP/PPP                yes (SLIP)

Pricing:
Model #         MC?     # ports          Price          Per port
------------    ---     -------         ------          --------
4208            yes         8          
4296            yes        32          
                           96          

Vendor #8: SynOptics/SynOptics
SNMP                    yes (extra cost)
TELNET                  yes
rlogin                  yes
LAT                     yes
rackmount               yes
accounting              ???
TN3270                  future (extra cost)
warranty                1 year
delivery schedule       ???
modem control           yes (standard)
reverse connections     yes
SLIP/PPP                yes (SLIP)

Pricing (*):
Model #         MC?     # ports          Price          Per port
------------    ---     -------         ------          --------
3030            yes        16        
                           48        
3000            yes        96        
                          176        
(note: SNMP and TN3270 are extra cost options and were not quoted.)

(*) 3030 (4-slot chassis) is $xxxxxx.
    3000 (12-slot chassis) is $xxxx.
    3333 (Ethernet AUI card) is $xxxxxx.
    3395 (16-port MC board) is $xxxxxxx.

Vendor #9: Digital/Digital
SNMP                    future
TELNET                  yes
rlogin                  no
LAT                     yes
rackmount               yes
accounting              no
TN3270                  no
warranty                1 year
delivery schedule       ???
modem control           no
reverse connections     yes
SLIP/PPP                future

Pricing:
Model #         MC?     # ports          Price          Per port
------------    ---     -------         ------          --------
DECserver 300   no         16           $ xxxx          $ xxxxxx

Vendor #10: BFA/Datability
SNMP                    yes
TELNET                  yes
rlogin                  yes
LAT                     yes
rackmount               yes (VCP-1000 - $xx extra)
accounting              no
TN3270                  no
warranty                1 year
delivery schedule       ???
modem control           yes (VCP/LC-32-RS423: partial [6 leads])
reverse connections     yes
SLIP/PPP                yes (SLIP)

Pricing:
Model #         MC?     # ports          Price          Per port
------------    ---     -------         ------          --------
VCP-200         full        8       
VCP-300         full       16       
VCP-1000        full       32       
                some       32       
                some       64       
                some      128       

Vendor #11: Lantronix/Lantronix
SNMP                    yes
TELNET                  yes
rlogin                  yes
LAT                     yes
rackmount               no
accounting              no
TN3270                  no
warranty                hardware: 2 years; software: 1 year update service
delivery schedule       ???
modem control           yes (6 leads)
reverse connections     yes
SLIP/PPP                no

Pricing:
Model #         MC?     # ports          Price          Per port
------------    ---     -------         ------          --------
EPS-4           yes     4s + 1p        
ETS-8           yes         8          
ETS-16          yes        16          

Vendor #12: Four Corners/Equinox
SNMP                    no
TELNET                  future
rlogin                  future
LAT                     yes
rackmount               option
accounting              no
TN3270                  no
warranty                hardware: 3 years
delivery schedule       ???
modem control           partial (4 leads)
reverse connections     yes (LAT)
SLIP/PPP                no

Pricing:
Model #         MC?     # ports          Price          Per port
------------    ---     -------         ------          --------
ELS-48          some       24        
                           48        

Vendor #13: Four Corners/Emulex
SNMP                    yes
TELNET                  yes
rlogin                  yes 
LAT                     yes
rackmount               yes
accounting              no
TN3270                  no
warranty                hardware: 5 year
modem control           P4000 - none (4 leads)
                        P8000 - partial (5 leads) or full
reverse connections     yes
SLIP/PPP                yes (SLIP)

Pricing: 
Model #         MC?     # ports          Price          Per port
------------    ---     -------         ------          --------
P4000           none        8        
                           16        
                           32        
P8000           some       64        
                          128        
                full       32        
                           64        


 --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --
DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT 

4. First cut analysis of the bid responses.

The first cut of the evaluation process was to select "finalists"
from the thirteen bids.  Following is an explanation, for each of
the bids, of why it was or was not selected as a finalist.

Vendor #1: R Squared/Xylogics: FINALIST
Why:
* Met all required criteria
* Met all selective criteria but TN3270 
* Modem control standard
* Best price per port at maximum port density ($xxxxxx)

Vendor #2: High-Tech/Xyplex: FINALIST
Why:
* Met all required criteria
* Met all selective criteria but accounting
* Only vendor offering TN3270
* Good price per port at 16-port density ($xxxxxxxx)

Vendor #3: cisco Systems: NOT FINALIST
Why:
* Price per port not competitive at either maximum port density
  ($xxxxxxxx) nor at 10-port density ($xxxxxxxx)

Vendor #4: Xyplex/Xyplex: NOT FINALIST
Why:
* Prices for the same product line xx% higher across the board than
  Vendor #2

Vendor #5: Anixter/Synoptics: NOT FINALIST
Why:
* Unable to configure a complete terminal server based upon the bid

Vendor #6: Anixterm/Racal Interlan: NOT FINALIST
Why:
* Vendor did not provide sufficient information with the bid to enable
  us to figure out whether the product met our requirements.  (Besides,
  the prices were noncompetitive.)

Vendor #7: Hughes LAN Systems/HLS: NOT FINALIST
Why:
* Prices per port not competitive at significant port densities 
  ($xxx - $xxx).

Vendor #8: SynOptics/SynOptics: NOT FINALIST
Why:
* Vendor bid Xyplex cards in SynOptics chassis, but prices per port
  were worse at almost all price points than Xyplex MAXserver 1000s
  as bid by High-Tech
* SynOptics' implementation of Xyplex software lags behind Xyplex'
  own (e.g. no TN3270 available from SynOptics as of time of bid)

Vendor #9: Digital/Digital: NOT FINALIST
Why:
* No modem control available from vendor
* Product lacks most of the selective criteria: no SNMP, no rlogin,
  no accounting, no TN3270, no SLIP.

Vendor #10: BFA/Datability: FINALIST
Why:
* Met all mandatory criteria and all selective criteria but TN3270
  and accounting
* Good price per port at 16 ($xxx) and 128 ($xxx)

Vendor #11: Lantronix/Lantronix: FINALIST
Why:
* Met all mandatory criteria and some selective criteria 
* Best price per port at 16 ($xxx)

Vendor #12: Four Corners/Equinox: NOT FINALIST
Why:
* Did not meet mandatory criteria (no TELNET)

Vendor #13: Four Corners/Emulex: NOT FINALIST
Why:
* Low-end unit unit lacks sufficient leads for our modem
  control applications (4 not enough, we need 5)
* High-end unit not price competitive at under 128 ports


 --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --- --
DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT 

5. Results of terminal server evaluation period.

From the thirteen bids, four were selected to be evaluated
on site: Vendors #1 (R Squared/Xylogics), #2 (High-Tech/Xyplex),
#10 (BFA/Datability) and #11 (Lantronix/Lantronix).  The following
evaluation units were sent to us:

    Xylogics Annex 3 
    Xyplex MAXserver 1820 
    Datability VCP-1000 with 8-port MC board 
    Lantronix ETS-16

We installed each of these units and cabled them to our IDX-3000
terminal switch for testing purposes.  Each was configured to run
both TCP/IP and LAT, and each was configured to support both 
"dialin" and "dialout" connections.  Once connected, we let
interested users on campus know how to access these units for testing.

The Datability VCP-1000 did not work properly.  When initially 
cabled and powered up, it responded to terminal activity, but 
soon after that it ceased functioning.  A week of phone calls 
to the vendor produced no satisfactory results, and so the product
was stricken from the list of finalists and was shipped back to
the vendor.

We managed to get the other three terminal servers functioning.

The Lantronix proved to be the least capable of the three.  While it
did provide TELNET, rlogin and LAT terminal connections, users 
reported some dissatisfaction with it.  

One item was that, in rlogin to a Unix host, some host applications
will set the terminal into "raw" mode.  The Xylogics and the Xyplex
honored the mode change, so that the application would have access to
characters (such as XON and XOFF) that would otherwise be reserved for
use by the terminal server.  However, the Lantronix would not toggle
into "raw" mode, so that a user running an application that requires
use of the full character set is required to escape back to the server
and manually change the port settings.

Another complaint about the Lantronix was that there did not appear 
to be a way to customize the local line-editing characters.  Again, 
both the Xyplex and the Xylogics provided such a capability.

Another, more minor problem with the Lantronix was that it did not
display LAT services in alphabetical order.

Once during our evaluation period, the Lantronix seized up and would
not respond to commands.  We were required to power cycle the unit in
order to restore it to service.

Due to these problems, plus the fact that it met the smallest set of
our selective criteria, we decided that the Lantronix was
unacceptable and returned the evaluation unit to the vendor.

We then focused our evaluation efforts on the Xylogics and the Xyplex.
Users found both products generally acceptable.  Both systems worked
flawlessly from a hardware standpoint.

Significant differences were seen in the following areas:

* The Xylogics was seen as being considerably more responsive to
  flow-control commands when running rlogin
* The Xylogics' documentation set was easier to follow than Xyplex'
* The Xylogics supports queued access to outbound rotaries, while
  the Xyplex does not
* The Xylogics supports the capability of defining different
  IP addresses for different outbound rotary sets, while the
  Xyplex does not.
* The Xyplex supports reverse LAT, while the Xylogics does not
* The Xylogics offers the ability to define macros and menus for
  the user interface, while the Xyplex does not
* The Xylogics can log connects to a host, for accounting and security
  purposes, while the Xyplex cannot
* With Xyplex' implementation of TCP, an attempted connect to an
  unsupported port on the Xyplex was met with a TCP-level timeout, 
  rather than an immediate "connection refused" as was the case with
  unsuccessful connect attempts to the Xylogics.

We did encounter one unresolved problem with the Xyplex.  The
application consisted of using ZMODEM to upload a binary file from a
PC running MS-DOS to a Sun, via an rlogin session on the terminal
server.  The transfer occurred flawlessly when using the Xylogics
Annex, but aborted when using the Xyplex in an otherwise identical
configuration.  Xyplex technical support was contacted but was unable
to solve the problem.

The Xyplex supports TN3270, while the Xylogics does not.  A member of
our IBM systems staff evaluated the TN3270 feature.  While she found
that it worked, she found the feature generally unsatisfactory, due to
its slow performance and the fact that it supports neither Kermit file
transfers nor non-ANSI terminals. In her judgement, the behavior of
the Xyplex TN3270 would be considered unsuitable by a user accustomed
to an IBM 7171 terminal protocol converter.  When we factor in the
additional $53 per port cost of the Xyplex TN3270 option, we cannot
recommend this feature as a solution for connecting terminals to
IBM mainframes.

On the basis of the Xylogics' superior manageability and
configurability, and its more featureful implementation of TCP/IP,
and also its better price at high port concentrations, we therefore
selected the Annex 3 as our terminal server of choice.  However,
recognizing that there are niches that it cannot fill, we selected the
Xyplex MAXserver 1000 series as an alternate.

DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT DRAFT 

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

 ===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#==

Path: utkcs2!darwin.sura.net!wupost!gumby!destroyer!uunet!munnari.oz.au!metro!grivel!loki!mark
Newsgroups: comp.os.vms
Message-ID: <2911@loki.une.edu.au>
References: <1992Jun30.030026.4362@ecl.psu.edu>
Organization: University of New England - Northern Rivers, (Lismore)
Lines: 14
Date: 4 Jul 1992 11:24:01 GMT
From: mark@loki.une.edu.au (Mark Garrett)
Subject: Re: SLIP is it available for VAX??

From article <1992Jun30.030026.4362@ecl.psu.edu>, by asm@ecl.psu.edu:
>
> SLIP (Serial Line Internet Protocol)  Is it available??. for VAX?

The most reliable easy (probably not the cheapest) solution is to run
TCP on VMS and run SLIP from a terminal server. If you have any server 300
or server 700 from DEC get V2.1 software for the 300 and you will have slip,
the 700 has it by default.


	Cheers
		Mark :)
-- 
Mark Garrett	Internet:  mark@loki.une.edu.au	Phone: 	+61 66 20 3859
   University of New England, Northern Rivers, Lismore NSW Australia.


    [contact Process Software for commercial SLIP or PPP implementations]


 ===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#==

Newsgroups: vmsnet.internals
Path: cs.utk.edu!gatech!paladin.american.edu!howland.reston.ans.net!wupost
      !sdd.hp.com!swrinde!elroy.jpl.nasa.gov!ames!network.ucsd.edu
      !mvb.saic.com!macro32
Message-ID: <20A00202_000774D8.00966D7727CED2C0$137_1@UK.AC.KCL.PH.IPG>
Date: Tue, 19 JAN 93 11:41:45 BST
Organization: Macro32<==>Vmsnet.Internals Gateway
From: SYSMGR@IPG.PH.KCL.AC.UK
Subject: Re: Need faster async board...dlv11-J? Something else?


>> Personally, at this point I'd feel a considerable  urge to bypass TTDRIVER
>> with all its unwanted baggage, and write my own UUCP-packet driver to access
>> the DHV hardware efficiently.
>
> One of the problems you'd run into real fast with this approach is that
> it would require you to use your driver for _all_ the ports on the DHV.
> Because of the hardware design, there isn't a good way for multiple
> drivers to use different lines on a multiplexed device like the DHV.
...
> Bruce C. Wright

I was thinking in view of the earlier discussion of the hardware thruput
limitations of the DHV, that allowing users onto the other ports was
most undesirable anyway. If that's untrue, then of course the best
compromise is an alternative class driver.

As far as terminal servers are concerned, I'd suggest contacting Lantronix.
Their prices are far lower than anything DEC make, and in particular they
do a tiny box called the EPS-1 which has one parallel port and one serial
port. You'd want to check up carefully on thruput and modem support on the
serial (RJ) line. Note that Lantronix servers have very full IP support,
including the definition of SLIP lines. I've been inpressed by Lantronix
hardware and their tech support, although I have no experience of the EPS1.

Out of interest, how practical is it to attempt to solve this problem using
LAT and software at the VAX end? You certainly get a high byte rate to or
from the server, but isn't there a problem with a long turn-around
time between the arrival of a packet at the terminal server and the dispatch
of any necessary acknowledgement message from the terminal server? I haven't
got recent experience: it was some time ago that I discovered that a
primitive KERMIT's thruput was drastically reduced when it was moved off
a DZV line onto a DS200.

        Yours,
                Nigel Arnot

                NRA%ipg.ph.kcl.ac.uk@nsfnet-relay.ac.uk   (internet)
                NRA%uk.ac.kcl.ph.ipg@ukacrl.bitnet        (bitnet)

 ===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#==

Newsgroups: vmsnet.misc,vmsnet.internals,comp.sys.dec,comp.org.decus
Path: cs.utk.edu!gatech!concert!rutgers!spcvxb!terry
Message-ID: <1993Jan23.235519.4987@spcvxb.spc.edu>
Date: 24 Jan 93 04:55:19 GMT
References: <1993Jan22.151835.1265@cmkrnl.com> <14012022@zl2tnm.gen.nz>
            <1993Jan23.110940.1270@cmkrnl.com>
Followup-To: vmsnet.internals
Organization: St. Peter's College, US
From: terry@spcvxb.spc.edu (Terry Kennedy, Operations Mgr.)
Subject: Re: Need faster async board...dlv11-J?  Something else?


In article <1993Jan23.110940.1270@cmkrnl.com>, jeh@cmkrnl.com writes:

> Newsgroups: vmsnet.misc,vmsnet.internals,comp.sys.dec,comp.org.decus

  Can we keep this to a smaller subset of the newsgroups, please? I've set
the followup header on this posting to vmsnet.internals only.

> There was a VT1xx terminal that worked in block mode... I want to say VT105,
> but that was a half-assed attempt at a graphics terminal.  

  VT131.

> That's never stopped them before.  Presumably the folks who designed the
> terminal server firmware had access to the same spec, but they came up with a
> different interpretation.  Several different interpretations, in fact.

  The physical serial communication chips in the DS200 won't send if CTS isn't
asserted. Period. It can't be "fixed" in the firmware/load image. However, DEC
is free to do it wrong in future server products.

> They also put out some release note on, I believe, DS200 3.1 code, talking
> about the server "asserting RI [ring indicator]", when RI is of course not
> sourced by the terminal server! 

  What they mean is that when you use the crossover cable (BC22R) to do re-
verse LAT, the signal that appears on the far end of the BC22R as RI is assert-
ed by the DS200. And they're correct.

	Terry Kennedy		Operations Manager, Academic Computing
	terry@spcvxa.bitnet	St. Peter's College, Jersey City, NJ USA
	terry@spcvxa.spc.edu	+1 201 915 9381



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

 ===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#==

Newsgroups: comp.os.vms
Path: cs.utk.edu!gatech!howland.reston.ans.net!agate!dog.ee.lbl.gov!network.ucsd.edu!mvb.saic.com!carpenter
From: Carpenter@Fwva.Saic.Com (Apprentice Wizard)
Subject: Re: DATABILITY (lat/telnet) servers... they ok??
Message-ID: <11795294@MVB.SAIC.COM>
Date: Tue, 06 Apr 1993 15:14:36 GMT
References: <20202B63_00186158.0096A95FF41E1D80$755_2@UK.AC.KCL.PH.IPG>
Organization: Science Applications Int'l, Corp. - Computer systems operation
Lines: 40

SYSMGR@IPG.PH.KCL.AC.UK writes:
> > Despite the problems I still like them. I just wish I had cheap (read free)
> > firmware upgrades (or at least fixes for current features that dont work).
> >
> > Stephan
> > jansen@madraf.astro.wisc.edu
>
>I can't stand this any more. You have to PAY to keep them working and you
>STILL LIKE THEM???
>
>Purely as someone else's satisfied customer, find out about Lantronix
>products. Their servers are smaller, cheaper, quieter, and better-featured
>than anything else I've ever heard of. Their latest software is available by
>anonymous ftp from ftp.lantronix.com (as are the manuals), and the guys on
>support@lantronix.com answer questions by return even on Labor day.
>
 
Purely as an *unsatisfied* customer, be sure to check out the Lantronix servers
before you use them. They do have a very nice menu system built in, but if
you try to segregate your LAT groups properly on a WAN, the Lantronix does not
work properly. For example, I want all of my LAN users to be able to connect tp
LAT groups 115 and 200. Within these groups, there are 15 vaxen/pc's. All other
connections should be made through TCP/IP. In order to cut down on the memory
usage (my two Lantronix kept crashing because there are so many LAT nodes on my
WAN) I set the node limit to 20. Both the server and all ports were set to
recognize just groups 115 and 200. Unfortunately, the server still recognizes
services from outside those groups and the group 115 and 200 services cycle in
and out of the services offered. Lantronix technical support acknowledged that
this is indeed a problem (three months ago), but also said that they would not
be fixing it anytime soon.
On the positive side, I have had no problems with the TCP/IP side of things and
I like the fact that you can have multiple virtual telnet and lat sessions.
 
=-=-=-=-=-=-==-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-==-=-=
Scott Carpenter                              
VAX Systems Manager                  Ya dina' tell 'im how long it'd really
SAIC Falls Church, VA                take ta fix it did ya'?
CARPENTER@FWVA.SAIC.COM                              M. Scott, CAPT, SUFP

 ===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#==

Newsgroups: comp.os.vms
Path: cs.utk.edu!darwin.sura.net!haven.umd.edu!uunet!nih-csl!postman
Message-ID: <1993Apr7.205934.29150@alw.nih.gov>
Sender: postman@alw.nih.gov (AMDS Postmaster)
Organization: Computer Systems Lab., NIH
References: <20202B63_00186158.0096A95FF41E1D80$755_2@UK.AC.KCL.PH.IPG>
            <11795294@MVB.SAIC.COM>
Date: Wed, 7 Apr 1993 20:59:34 GMT
From: tate@faxcsl.dcrt.nih.gov (Ramon Tate)
Subject: Re: DATABILITY (lat/telnet) servers... they ok??

In article <11795294@MVB.SAIC.COM> Carpenter@Fwva.Saic.Com (Apprentice Wizard)
writes:
>SYSMGR@IPG.PH.KCL.AC.UK writes:
>>> Despite the problems I still like them.  I just wish I had cheap (read
free)
>>> firmware upgrades (or at least fixes for current features that dont work).
>>>
>>> Stephan
>>> jansen@madraf.astro.wisc.edu
>>
>>I can't stand this any more. You have to PAY to keep them working and you
>>STILL LIKE THEM???
>>
>>Purely as someone else's satisfied customer, find out about Lantronix
>>products. Their servers are smaller, cheaper, quieter, and better-featured
than
>>anything else I've ever heard of. Their latest software is available by
>>anonymous ftp from ftp.lantronix.com (as are the manuals), and the guys on
>>support@lantronix.com answer questions by return even on Labor day.
>>
> 
>Purely as an *unsatisfied* customer, be sure to check out the Lantronix
servers
>before you use them. 

[gory details ommitted]

>-=-=-=-=-=-==-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-==-=-=

>Scott Carpenter                              
>VAX Systems Manager                  Ya dina' tell 'im how long it'd really
>SAIC Falls Church, VA                take ta fix it did ya'?
>CARPENTER@FWVA.SAIC.COM                              M. Scott, CAPT, SUFP
>-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=


Strange coincidence here - just had a call from a user about a Lantronix
terminal server problem. Seems they powered the unit down after doing some sort
of maintenance, and when it came up, it got caught up in a deadly embrace with
a hitherto unknown BOOTP server on the LAN. Turns out that the unit first asks
for a BOOTP load, and if there is no response, it then makes a MOP request
(which is what the VAX load host expects). The BOOTP server was locked in a
closet and is administered by someone else in another building, who obviously
wasn't aware that (1) there was a problem and (2) that a problem might even be
possible. When Lantronix tech support was asked about how to circumvent the
BOOTP request, the answer was that it was a feature and couldn't be changed.
The short term fix was to temporarily disconnect the BOOTP server and let MOP
do its thing; in the long term, it MAY be possible to screen out the terminal
server's Ethernet address. Maybe.

Sounds like a design flaw to me....
-------------------------------------------------------------------------------
Ramon L. Tate                   "Eschew equimagnitudinal contraincrementation."
Division of Computer                                  - Richard Shrager
  Research and Technology
National Institutes of Health
Bethesda, MD 20892    USA                   tate@faxcsl.dcrt.nih.gov

 ===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#==

Newsgroups: comp.sys.dec
Path: cs.utk.edu!gatech!howland.reston.ans.net!agate!headwall.Stanford.EDU!rutgers!spcvxb!hsh!lee
From: lee@hsh.com (Lee Havemann)
Subject: Re: DEC server 200/mc serial line driver IC
Message-ID: <1993May19.094902.1677@hsh.com>
Date: 19 May 93 13:49:02 GMT
References: <lyeo.1.737646823@UPEI.CA>
Organization: HSH Associates
Lines: 33

In article <lyeo.1.737646823@UPEI.CA>, lyeo@UPEI.CA (Larry Yeo) writes:
>
> We have a Dec server 200/mc (Enet to 8 port serial) with one bad serial port
> and we think that the problem is a bad "line driver?" TS1080 IC.....
> The problem is we have no data on this IC, and would like to find a source 
> or a replacement chip. This chip can be found on the I/O board which has the 
> 8 DB25 connectors, and they are 10 of these chips on this board. If anyone 
> has found a replacement IC or supplier ( Dec only replaces boards..Pricey!) 
> we would appreciate any information you can give.
> 
> 				Larry Yeo
> 				Technical Services
> 				University of Prince Edward Island
> 				Canada


You can try Newark Electronics, who has offices in the US and Canada.  
The Ontario # is (519) 685-4280. Email me if you need any other phone #'s.
They have always been helpful in cross-referencing custom chip numbers.


I have not looked inside a 200 /mc, but I would think that it would use the 
standard 8 pin line drivers that DEC has used in all its other products,
which cross refs to a Texas Instruments number - which is either a 
UA9639CP or a UA9636CP (or ACP). One is a receive buffer, the other is a 
line driver (I always get confused as to which is which :-)



 Lee Havemann, Comp Ops Dir.    HSH Associates      (201) 838-3330 
 Internet: lee@hsh.com    Compuserve: 70410,3507    AOL: HSH Assoc
 "Any opinions expressed are not necessarily those of anyone else, 
  including myself."                     

 ===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#==

Newsgroups: comp.sys.dec
Path: cs.utk.edu!gatech!howland.reston.ans.net!zaphod.mps.ohio-state.edu!cs.utexas.edu!rutgers!spcvxb!terry
From: terry@spcvxb.spc.edu (Terry Kennedy, Operations Mgr.)
Subject: Re: DEC server 200/mc serial line driver IC
Message-ID: <1993May19.155226.5925@spcvxb.spc.edu>
Date: 19 May 93 19:52:26 GMT
References: <lyeo.1.737646823@UPEI.CA> <1993May19.094902.1677@hsh.com>
Organization: St. Peter's College, US
Lines: 15

In article <1993May19.094902.1677@hsh.com>, lee@hsh.com (Lee Havemann) writes:
>
> I have not looked inside a 200 /mc, but I would think that it would use the 
> standard 8 pin line drivers that DEC has used in all its other products,
> which cross refs to a Texas Instruments number - which is either a 
> UA9639CP or a UA9636CP (or ACP). One is a receive buffer, the other is a 
> line driver (I always get confused as to which is which :-)

  Nope, DEC discovered that users knew where to get those parts and changed
to something even more bizarre 8-(. It's a large chip, about the size of a
2708 EPROM as I recall, and it may have built-in voltage converters like some
of the Maxon (or is that Maxxon?) parts.

	Terry Kennedy		Operations Manager, Academic Computing
	terry@spcvxa.bitnet	St. Peter's College, Jersey City, NJ USA
	terry@spcvxa.spc.edu	+1 201 915 9381

 ===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#==

Newsgroups: comp.dcom.servers,comp.terminals
Path: cs.utk.edu!gatech!howland.reston.ans.net!math.ohio-state.edu!uwm.edu
      !linac!att!cbnewsk!wmg
Organization: AT&T
Distribution: na
Lines: 121
Message-ID: <CAJrxL.845@cbnewsk.cb.att.com>
Followup-To: comp.dcom.servers
Keywords: terminal, modem, servers
Date: Thu, 22 Jul 1993 03:48:56 GMT
From: wmg@cbnewsk.cb.att.com (william.m.gilroy)
Subject: SUMMARY terminal servers


Here is my original request to the net for terminal servers.  I had
said that I would provide a summary of the results that I received.
Sorry for the delay but things have been crazy here at work.

> I am in the market for a terminal server.  The use of this product
> is going to be as a modem pool.  On the one side will be a bunch
> of modems (about 8 to 16).  I figure that the terminal server will
> be connected to the ethernet.  I will have some Unix machines Suns, 
> SVR4, HP's, etc talking to the terminal server to dial out.  I also
> want the terminal sever be able to route call to the correct host 
> based on either the port that the call is received or on some type
> of login and routing script.  This Unix and TPC/IP side should be
> fairly easy to find and setup.
> 
> The next thing is a bit more difficult.  On the ethernet I have a 
> bunch of PC's running MS Windows 3.1 and Lan Manager 2.0 
> (starlan).  I would also like the PC's to be able to connect to the
> terminal server via ethernet for out bound only.  It would be really
> nice if the PC users could use all of their windows based software
> for communications.
> 
> Does anyone know of a product that can do all of this.  What should I 
> be considering, what should I look out for.  Any information would be
> appreciated.  Also if anyone is doing this type of stuff I would really
> like to hear from them.  
> 
> Also I was not quite sure if these were the correct news group to post 
> this message, if there is a better choice please let me know.  I will 
> summarize the responses that I get and post that summary.
> 
> Thanks,
> 
> Bill Gilroy
> AT&T Image Solutions
> e-mail: wmg@pixels.att.com
> voice:  (908) 981-8000 x239
> 

The one thing I have learned is that there are many manufacturers of terminal
servers and they all are sort of close in features and capacity.  There are
many things you want to keep in mind when buying a terminal server.  Some
are listed below in no particular order.

	1) Does the terminal server have full modem control on all ports?
	2) What is the maximum speed (bps) per port?  Can all ports run
	at that speed?
	3) What protocols does it support? TCP/IP, LAT, telnet, etc.
	4) Does the vendor supply you with drivers for the host that will
	make the terminal servers ports look local?
	5) Maximum number of host sessions.
	5) What type of reverse protocols are supported?
	7) How is the software loaded on the terminal server (ROM, host 
	download, floppy, etc.)?

These questions may or may not be an important issue to every person.  The
best thing that I can recommend is decide what features you need and
then see what vendors support those features.  
The first that that I found after I posted my request for information 
was an article about terminal servers in the May 1993 issue of 'Unix
Review.'  They had about 4 pages of text along with a chart that compared
50 terminal servers.  It had some good information about terminal
servers, plus many manufacturers of terminal servers.  I recommend that 
you obtain a copy of this article.

If anyone has any specific questions you and mail them to me and I will
be glad to try to answer them.  But like I said if this is your 
first trip into the world terminal servers I highly recommend the 
Unix Review article.

I myself am leaning towards a Xylogics Annex because it has drivers
to make the ports appear to be on host.  This will allow most of 
the software on my server to run without any modifications.  The other
possibility is the Equinox ELS-48, while that does not have any drivers
to create local ports it does support reverse TCP/IP connections.  They
also should be releasing support for SLIP this month.  I do not know
if PPP is in the works.  Anyway I still have to decide which terminal
server would be the best for my application.

It would also be nice if the terminal server could be monitored/controlled
by SNMP.

Thanks to the people who responded to my request and sorry for the 
delay in posting a summary.

Eric Walters <ericw@sequent.com>
Alexandre Khalil <0758CS11@DT3.DT.UH.EDU>
Chris Chase <chase@aphill.jhuapl.edu>
Alexandre Khalil <0758CS11@DT3.DT.UH.EDU>
L.G. "Ted" Stern <stern@amath.washington.edu>
vic@cd.com (Vic Serbe x237)
root@foobar.hanse.de (Jens Stark)
rrc@attics.rn.com (Randy Cable - Contractor)


===== Summary below =====


I received one recommendation that Xylogics makes a good, high quality
product.  They did not included about how it works, just that 
the user was satisfied.

Someone else also recommend a Central Data SCSI terminal server.  While
this product appears well suited for adding more ports to a host computer
it lacks the ability for all hosts and PC's to access the terminal 
server.  While this is good for many situations it does not meet my 
needs.  I also believe that the price is cheaper per port than an 
ethernet terminal server.

There was also a recommendation for a 3COM 2100 or 3100 terminal server.

And also a recommendation for a LINC/TERM Multiple Protocol Terminal Server
from Hughes LAN Systems.

I am not sure if this came from my post or someone else's but it is
applicable.  They were requesting a terminal server that could 
make the ports look like local tty's on a Sun.  They had obtained
these recommendations.  The following vendor provide such a product,
Xylogics (Annex 3, Micro Annex), Livingston (Portmaster), and 
Lantronix (ETS-8 and others) all support this kind of functionality.

 //////////////////////////////////////////////////////////////////////////////
 ===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#==
 \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\

Newsgroups: comp.sys.dec,comp.os.vms,vmsnet.networks.tcp-ip.misc
Path: cs.utk.edu!darwin.sura.net!howland.reston.ans.net!vixen.cso.uiuc.edu
      !usenet.ucs.indiana.edu!indyvax.iupui.edu!imhw400
Message-ID: <1993Jul26.095331.1537@indyvax.iupui.edu>
References: <22kq5sINN4ck@gap.caltech.edu> <22ksnq$935@sol.TIS.COM>
            <1993Jul24.141501.3076@nntpd.lkg.dec.com>
Date: 26 Jul 1993 09:53:31 -0500
From: imhw400@indyvax.iupui.edu
Subject: Re: TCP/IP instead of DECNET?


In article <1993Jul24.141501.3076@nntpd.lkg.dec.com>,
 thomas@lkg.dec.com (Matt Thomas) writes:
> 
> In article <1993Jul23.102714.6649@infodev.cam.ac.uk>,
> rf@cl.cam.ac.uk (Robin Fairbairns) writes:
>>
>>In article <1993Jul22.104512.1518@indyvax.iupui.edu>,
>> imhw400@indyvax.iupui.edu writes:
>>>
>>> Can anybody comment on just what ISO VTP might do for us?
>>
>>I doubt it'll do very much for you.  I know of few implementations of
>>it, not even one by DEC despite a DEC man being one of the leading
>>lights of the development of the standard.
> 
> VTP comes standard as part of DECnet-OSI for ULTRIX or OSF/1 or OpenVMS.
> That being said, VTP is just another way of remote teminals.  It's doesn't
> really differ much in functionality from the other remote terminal protocols.


Yeah, but is it routable?  Non-routability is the only real strike
against LAT. Is VTP engineered to take advantage of opportunities for
multiplexing?  Does it encompass most of what people actually want to do
*today* without resorting to a dozen afterthoughts^Jaddenda to describe
all the options?  Does it do a good job of moving typical terminal-driver
operations out close to the user? 

Inquiring minds want to know.

-- 
Mark H. Wood, Lead Systems Programmer    +1 317 274 0749   [@disclaimer@]
Internet:  IMHW400@INDYVAX.IUPUI.EDU     BITNET:  IMHW400@INDYVAX


 ===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#==

Newsgroups: comp.sys.dec,comp.os.vms,vmsnet.networks.tcp-ip.misc
Path: cs.utk.edu!gatech!howland.reston.ans.net!noc.near.net!uunet
     !elroy.jpl.nasa.gov!decwrl!pa.dec.com!nntpd2.cxo.dec.com!nntpd.lkg.dec.com
     !lkg.dec.com!thomas
Message-ID: <1993Jul27.112105.15561@nntpd.lkg.dec.com>
Sender: thomas@lkg.dec.com (Matt Thomas)
Organization: Digital Equipment Corporation
References: <22kq5sINN4ck@gap.caltech.edu> <22ksnq$935@sol.TIS.COM>
            <1993Jul24.141501.3076@nntpd.lkg.dec.com>
            <1993Jul26.095331.1537@indyvax.iupui.edu>
Date: Tue, 27 Jul 1993 11:21:05 GMT
From: thomas@lkg.dec.com (Matt Thomas)
Subject: Re: TCP/IP instead of DECNET?


In article <1993Jul26.095331.1537@indyvax.iupui.edu>,
| imhw400@indyvax.iupui.edu writes:
|>> VTP comes standard as part of DECnet-OSI for ULTRIX or OSF/1 or OpenVMS.
|>> That being said, VTP is just another way of remote teminals.  It's doesn't
|>> really differ much in functionality from the other remote terminal protocols.
|>
|>Yeah, but is it routable?  Non-routability is the only real strike against
|>LAT. Is VTP
|>engineered to take advantage of opportunities for multiplexing?  Does it
|>encompass most of what people actually want to do *today* without resorting
|>to a dozen afterthoughts^Jaddenda to describe all the options?  Does it do a
|>good job of moving typical terminal-driver operations out close to the user?
|>Inquiring minds want to know.

1) It is routable.  It runs over OSI Transport (which runs over either over
   CONS/X.25 or CLNS/PhaseIV.

2) VTP is an ISO defined protocol.  Of course not.

3) An ISO spec without addendums and implementor agreements?  I don't think
   that is allowed. 

3) VTP supports a number of profiles, among these are Forms (which we don't
   implement) and Generalized Telnet (which we do).  As I said, VTP is no
   better nor worse than other Telnet-like protocols.

--
Matt Thomas                            Internet:   thomas@lkg.dec.com
U*X Networking                         UUCP:       ...!decwrl!thomas
Digital Equipment Corporation          Disclaimer: This message reflects my own
Littleton, MA                                      warped views, etc.

 ===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#==

Article 19572 of comp.sys.dec:
Newsgroups: comp.sys.dec
Path: cs.utk.edu!emory!europa.eng.gtefsd.com!uunet!spcuna!spcvxb!terry
From: terry@spcvxb.spc.edu (Terry Kennedy, Operations Mgr.)
Subject: Re: DS200 reprogramming
Nntp-Posting-Host: spcvxa.spc.edu
References: <1993Oct26.091307.121@opo.van.wa.us>
Sender: news@spcuna.spc.edu (Network News)
Organization: St. Peter's College, US
Date: Wed, 27 Oct 1993 02:00:54 GMT
Message-ID: <1993Oct26.220054.1@spcvxb.spc.edu>
Lines: 25

In article <1993Oct26.091307.121@opo.van.wa.us>, evm@opo.van.wa.us writes:
> 	How hard is it to write your own code to down load to a decserver? 
> With a number of DS200 on the used market cheap this 68000 platform may make 
> sense for other control applications or just some fun. An ethernet port and 8 
> RS232 lines are ok.
> 
> 	To do this one would need to know the hardware and the details of the 
> software. Are schemetics out there somewhere? how about down loading 
> development support tools. DEC of course must have a full set of these. How do 
> I get a copy?


  You need EK-DECS2-TM, "DECserver 200 Technical Manual" at $42.00 from DEC.
This gives the port address and bit assignments, etc. The download image is
formatted as an RSX-11 task image, except that it's 68000 code 8-). That is
all described in the MOP Protocol Specification, which DEC has freely made
available (anonymous FTP to ftp.spc.edu in [.mop-server] and get mopv50.ps.

  I don't believe prints are available, but they shouldn't be necessary.

  Similar books are available for the DS100 (EK-DSRVA-TM, $42.00) and DS300
EK-A0367-TM, $63.00) although I can't vouch for the contents.

	Terry Kennedy		Operations Manager, Academic Computing
	terry@spcvxa.bitnet	St. Peter's College, Jersey City, NJ USA
	terry@spcvxa.spc.edu	+1 201 915 9381


 ===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#==

Newsgroups: comp.os.vms,comp.sys.dec,vmsnet.misc
Path: cs.utk.edu!emory!europa.eng.gtefsd.com!uunet!decwrl!decwrl!waikato!comp.vuw.ac.nz!zl2tnm!toyunix!don
Subject: Re: Why do DECSERVER 100s Recover Slowly After Power Failure?
Message-ID: <9751307@zl2tnm.gen.nz>
From: don@zl2tnm.gen.nz (Don Stokes)
Date: 4 Nov 93 10:34:49 GMT
Sender: news@zl2tnm.gen.nz (GNEWS Version 2.0 news poster.)
References: <1993Nov3.143737.1@lib.haifa.ac.il>
Distribution: world
Organization: The Wolery
Lines: 25
Xref: cs.utk.edu comp.os.vms:76405 comp.sys.dec:19850 vmsnet.misc:2230

jody@lib.haifa.ac.il (Yosef Branse) writes:
> We recently installed some second-hand DECSERVER 100 terminal servers. They
> work fine, but we encountered a problem when our system re-boots after a
> power failure. Our other servers (DECSERVER 300s) resume operation as soon as
> the computer is back up, but the 100s aren't online until about 15 minutes
> later.

Sounds like you've run into the exponential backoff algorithm the DS100s 
(and 200s) use to determine their load request retries.

Basically, the DECservers try to load as soon as they power up.  If that
fails, they retry almost imediately, then wait a minute before the next 
retry, then two minutes for the next, then four minutes, then eight and
so-on.  (The numbers may not be right, but you get the general idea.)

The problem is that the VAXes won't load the terminal servers until DECnet
is started, and they may be well into the eight or sixteen minute backoff
by the time the VAXes are ready.  

Basically, leave the DECservers powered off until DECnet has started on a
VAX with the load images on it.  Then power up the DECservers, and they
should all load almost immediately.

--
Don Stokes, Network Manager, Victoria University of Wellington, New Zealand.
don@zl2tnm.gen.nz(home) don@vuw.ac.nz(work) +64 4 495-5052 Fax+64 4 471-5386


 ===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#==

Newsgroups: comp.os.vms,comp.sys.dec,vmsnet.misc
Path: cs.utk.edu!emory!swrinde!cs.utexas.edu!uwm.edu!vixen.cso.uiuc.edu!moe.ksu.ksu.edu!mccall!cmkrnl!jeh
Subject: Re: Why do DECSERVER 100s Recover Slowly After Power Failure?
Message-ID: <1993Nov5.235230.2975@cmkrnl.com>
From: jeh@cmkrnl.com
Date: 5 Nov 93 23:52:30 PST
Followup-To: comp.os.vms,comp.sys.dec,vmsnet.misc
References: <1993Nov3.143737.1@lib.haifa.ac.il>,<9751307@zl2tnm.gen.nz> 
<1993Nov4.090036.264@buckie.hsc.colorado.edu> <CG1p7M.1v3@wpdis11.wpafb.af.mil>
Organization: Kernel Mode Systems, San Diego, CA
Lines: 63
Xref: cs.utk.edu comp.os.vms:76563 comp.sys.dec:19910 vmsnet.misc:2244

In article <CG1p7M.1v3@wpdis11.wpafb.af.mil>, akee@wpdis11.wpafb.af.mil ((HCST) Earle F. Ake) writes:
> Dan Wing (dwing@uh01.Colorado.EDU) wrote:
> : In article <9751307@zl2tnm.gen.nz>, don@zl2tnm.gen.nz (Don Stokes) writes:
>
> : >Basically, leave the DECservers powered off until DECnet has started on a
> : >VAX with the load images on it.  Then power up the DECservers, and they
> : >should all load almost immediately.
>
> : Or, have an InfoServer do all your DECserver loads, and put a little UPS on
> : the InfoServer if necessary.
>
> 	Actually I thought about something else someone might try.  I don't
> have access to VMS anymore or I would try it myself.


So that explains why we've heard nothing from you lately... hope you're doing
well!

> 	First determine if the terminal server is up or not.  I think you
> can do this with NCP.

You can, interactively.  NCP> CONNECT NODE server-name ,  If it suceeeds,
the server is up.   If it doesn't succeed, though, it may take an unreasonably
long time to decide that fact.

Doing it non-interactively (and getting a result that you can test via DCL) is
more difficult.  The best approach would be to simply write some code that
tries to speak MOP to the terminal server.

Failing that, I tried this (don't laugh!) (well, maybe just a little!):

	$ ncp :== $ncp
	$ define/user sys$command nl:
	$ ncp connect node server-name

...if the connection works you get SS$_ILLIOFUNC, as NCP tries to perform
some illegal function (maybe a read with prompt?) on the null device.  If it
doesn't work you get some other status depending on just how it failed.

> Then using NCP I believe you can do a 'load node x'
> where 'x' is the name of your terminal server.  This might work but you
> best make sure it is not already loaded and running.

The command is NCP> TRIGGER NODE node-name (assuming that the node name for
the server has been set up in your local DECnet database)

> 	Using NCP I think you can check the return status of a NCP loop node
> command to see if it is up and running.

You could, if the terminal server was a DECnet node.  But it isn't (it merely
has a DECnet node name and address defined in a VMS system's DECnet database;
and in fact, it doesn't even need that), so it can't respond to NCP LOOP NODE
commands.

Hmmm, you might be able to loop your Ethernet circuit, specifying the terminal
server as the loop assistant?  Hmmm, I just tried it, no good.  Odd, I thought
that was all done via MOP.  Maybe the DECserver 200MC's MOP implementation
doesn't support loop assistance.

	--- Jamie Hanrahan, Kernel Mode Systems, San Diego CA
drivers, internals, networks, applications, and training for VMS and Windows NT
uucp protocol weenie and release coordinator, VMSnet (DECUS uucp) Working Group
Chair, Programming and Internals Working Group, U.S. DECUS VMS Systems SIG
Internet:  jeh@cmkrnl.com (JH645)  Uucp: uunet!cmkrnl!jeh  CIS: 74140,2055

 //////////////////////////////////////////////////////////////////////////////

Newsgroups: comp.os.vms,comp.sys.dec,vmsnet.misc
Path: cs.utk.edu!gatech!udel!news.sprintlink.net!crash!cmkrnl!jeh
From: jeh@cmkrnl.com
Subject: Re: Why do DECSERVER 100s Recover Slowly After Power Failure?
Message-ID: <1993Nov6.220728.2978@cmkrnl.com>
Date: 6 Nov 93 22:07:28 PST
References: <1993Nov3.143737.1@lib.haifa.ac.il>,<9751307@zl2tnm.gen.nz>
           <1993Nov5.235230.2975@cmkrnl.com> <1993Nov6.182819.1@spcvxb.spc.edu>
Followup-To: comp.os.vms,comp.sys.dec,vmsnet.misc
Organization: Kernel Mode Systems, San Diego, CA
Lines: 31

In article <1993Nov6.182819.1@spcvxb.spc.edu>,
 terry@spcvxb.spc.edu (Terry Kennedy, Operations Mgr.) writes:
>
> In article <1993Nov5.235230.2975@cmkrnl.com>, jeh@cmkrnl.com writes:
>> Doing it non-interactively (and getting a result that you can test via DCL)
>> is more difficult.  The best approach would be to simply write some code
>> that tries to speak MOP to the terminal server.
> 
>   That code won't work on a system with DECnet/MOP running, since the system
> will reject the attempt to register a duplicate protocol when you talk to the
> Ethernet device.You might be able to piggyback onto the existing MOP handler,
> but I don't believe it has any external hooks.

Hmmm.  Exactly when does the MOP protocol get registered?  I've always assumed
that it happened when you enabled the circuit for "service".  

>> Hmmm, you might be able to loop your Ethernet circuit, specifying the
>> terminal server as the loop assistant?  Hmmm, I just tried it, no good. 
>> Odd, I thought
>> that was all done via MOP.  Maybe the DECserver 200MC's MOP implementation
>> doesn't support loop assistance.
> 
>   I don't think that stuff lives in MOP, since (for example) you can loop an
> Ethernet circuit through a PDP-11 running RSTS/E, which doesn't support MOP

MOP does define a loopback function at the circuit level (at least according to
the old, circa-1983 MOP document, haven't checked the new one), but this is
optional.  DECnet Phase IV specs for Ethernet nodes do require that the
Ethernet controller support loopback -- but, of course, a terminal server isn't
a DECnet node!  So I guess the omission is defensible from the p.o.v. of the
published standards (which is not to say that the omission was a good thing). 

	--- Jamie Hanrahan, Kernel Mode Systems, San Diego CA
Internet:  jeh@cmkrnl.com (JH645)  Uucp: uunet!cmkrnl!jeh  CIS: 74140,2055

 //////////////////////////////////////////////////////////////////////////////

Newsgroups: comp.os.vms
Path: cs.utk.edu!nntp.memst.edu!ukma!news2.uunet.ca!math.ohio-state.edu
      !howland.reston.ans.net!paladin.american.edu!darwin.sura.net
      !news.lsu.edu!sysmgr
Message-ID: <2e7gnf$46js@te6000.otc.lsu.edu>
References: <1993Dec7.231426.26196@cs.rit.edu>
Organization: Louisiana State University - Mech. Engr. Dept.
NNTP-Posting-Host: imr00.me.lsu.edu
In-Reply-To: whd0675@cs.rit.edu's message of Tue, 7 Dec 1993 23:14:26 GMT
Date: 9 Dec 1993 15:37:19 GMT
From: SYSTEM@imr00.me.lsu.edu (Alaric S. Haag)
Subject: Re: DECservers...

In <1993Dec7.231426.26196@cs.rit.edu> whd0675@cs.rit.edu writes:
>
> I have a DECserver that has no documentation, and I'm trying to get it to
> work.  Could someone please point me to where I can get documentation on
> DECservers?  Without any documentation I have been able to get it to remotely
> boot off of the
> VAX and I can get the local prompt.  But it does not see any services...  
>
> In addition, when I connect to it via NCP, I get a # and that's all.  
> It is a DECserver 100, but I would imagine that most DECservers have a core
> command set, no?  If anyone could tell me where I can get a copy of the 
> DECserver manuals or anything, I would greatly appreciate it.
> 
> Walter Dick
> whd0675@cs.rit.edu


I don't know if DEC still sells them, but the part number for the
DS100  manual is AA-Z085B-TK. If I recall correctly, you can hold the
"red square" (!) button down on the back _while_you_power_it_on_ and
you will set the server up in its factory configuration. At that point
the SET PRIV password is "system" and the login password (#) is
"access".

Hope this gets you going!

Ric
%^{)

----
[ Alaric S. Haag, Research Associate                 mehaag@sn01.sncc.lsu.edu ]
[ Louisiana State University, Mech. Engr. Dept.       Facts:   (504) 388-5990 ]
[ Baton Rouge, LA 70803                            Opinions:   (504) 388-5897 ]


 ===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#==

Newsgroups: comp.os.vms
Path: cs.utk.edu!gatech!europa.eng.gtefsd.com!uunet!spcuna!spcvxb!terry
From: terry@spcvxb.spc.edu (Terry Kennedy, Operations Mgr.)
Subject: Re: DECservers...
Nntp-Posting-Host: spcvxa.spc.edu
References: <1993Dec7.231426.26196@cs.rit.edu> <patrickr-101293072503@137.244.224.101> <2ea7shINN2ra@oasys.dt.navy.mil>
Followup-To: comp.os.vms
Sender: news@spcuna.spc.edu (Network News)
Organization: St. Peter's College, US
Date: Fri, 10 Dec 1993 23:23:50 GMT
Message-ID: <1993Dec10.182350.1@spcvxb.spc.edu>
Lines: 32

In article <2ea7shINN2ra@oasys.dt.navy.mil>,
 conrad@oasys.dt.navy.mil (Johnson) writes:
>
> I dread any non-privileged user getting any passwords that they are not
> entitled.  They can get into places where they are not allowed and
> really muck up the system and all hard work we as system managers
> have done.  Pleas send passwords via e-mail and not via posts. thnks 


  Well, hopefully anyone that's that security-conscious will not be using the
default DEC-supplied password for the MOP Remote Console function. Of course,
that password doesn't get you anything beyond the same level of access that a
local user would have.

  Next, if the assumption is that a "bad guy" may be asking the question,
why hide the exchange in e-mail? Also, do you give your users access to the
documentation (either hard-copy or Bookreader)? It's in there, too...

  Lastly, you should know that the LAT protocol isn't secure. There are a num-
ber of packages around that will decode LAT sessions. Also, I have written
code in the past that will dump the entire setup of a DECserver (including all
of the passwords in cleartext), without needing anything other than access to
the Ethernet segment the server is on. [No, I'm not giving these out!].

  Terminal-server-based passwords for the entire server provide a limited
amount of security (since more than one person knows them, the entire set of
server users can deny giving out the passwords - you can't prove that any one
user disclosed the passwords). The servers provide additional tools to help
with security, such as the LIMITED VIEW and SECURE settings. However, these
are no substitute for adequate host-based security, and users shouldn't be

lulled into thinking that they are.

	Terry Kennedy		Operations Manager, Academic Computing
	terry@spcvxa.bitnet	St. Peter's College, Jersey City, NJ USA
	terry@spcvxa.spc.edu	+1 201 915 9381

 //////////////////////////////////////////////////////////////////////////////

Newsgroups: comp.os.vms,comp.protocols.tcp-ip,comp.sys.dec,comp.sys.m68k
Path: cs.utk.edu!gatech!howland.reston.ans.net!pipex!uknet!EU.net!news.eunet.fi!dkuug!eskimo.CPH.CMC.COM!lars
From: lars@Eskimo.CPH.CMC.COM (Lars Poulsen)
Subject: Re: KA9Q ported to 68000?
Message-ID: <1993Dec17.121541.18689@Eskimo.CPH.CMC.COM>
Organization: CMC Network Products, Copenhagen DENMARK
References: <1993Dec13.155812.138@opo.van.wa.us> <2erpo1$n3g@mips.arb-phys.uni-dortmund.de>
Date: Fri, 17 Dec 93 12:15:41 GMT
Lines: 30

In article <1993Dec13.155812.138@opo.van.wa.us>,  <evm@opo.van.wa.us> wrote:
>>I've got a few Dec DS200/mc terminal servers floating around ...
>>	They have 1 AUI connection for ethernet and 8 serial lines 
>>with a max speed of 38Kbps. Internally is a 10 MHz 68000, 768 Kb Ram 
>>and 128 Kb Eprom.

In article <2erpo1$n3g@mips.arb-phys.uni-dortmund.de>
   wb@arb-phys.uni-dortmund.de (Wilhelm B. Kloke) writes:
>If you manage to program it for support of V32bis/V42bis modem slip or
>ppp support (at 38kBd), I would like to hear from you. The 68k should be
>sufficient to support more than 1 line at full speed. I think it should
>even do all lines, if programmed well.
>
>The DS200 (as used by DEC) does only support 9600 Bd. Where did you get
>the info about the max speed?

Based on my work with similar hardware, I assert that a 10MHz 68000
with ordinary double- or triple-buffered UARTs will just barely drive
8 full-duplex 9600 bps ports at full speed. Most of the common UART
components of the vintage in question will support speeds up to 38400
bps. You could probably get away with driving two V.32bis/V.42bis
modems at the same time with only occasional overruns.

Considering the effort required to program this (and reverse-engineer
the programming spec for the board) I don't think it is worth it.
-- 
/ Lars Poulsen			Internet E-mail: lars@CMC.COM
  CMC Network Products		Phone: (011-) +45-31 49 81 08
  Hvidovre Strandvej 72 B	Telefax:      +45-31 49 83 08
  DK-2650 Hvidovre, DENMARK	Internets: designed and built while you wait

 ===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#==

Newsgroups: comp.os.vms
Path: cs.utk.edu!emory!europa.eng.gtefsd.com!howland.reston.ans.net!usc!news.service.uci.edu!unogate!mvb.saic.com!buckie.ucha!buckie.nntp!DWING
Subject: Re: Rebooting DEC servers 200 and 550??
Message-ID: <1993Dec15.161607.527@buckie.hsc.colorado.edu>
From: dwing@uh01.Colorado.EDU (Dan Wing)
Date: 15 Dec 93 16:16:07 MDT
Reply-To: dwing@uh01.Colorado.EDU
References: <1993Dec15.160654.7628@mnemosyne.cs.du.edu>,<1993Dec15.165444.1@spcvxb.spc.edu>
Organization: Ski Bum Wanna-be, Incorporated
Nntp-Posting-Host: buckie.hsc.colorado.edu
Lines: 11

In article <1993Dec15.165444.1@spcvxb.spc.edu>, terry@spcvxb.spc.edu (Terry Kennedy, Operations Mgr.) writes:
>Then say "LOGOUT PORT x", *twice*. The twice-in-a-row part is the secret.
>The first one clears the hang condition (XOFF, flow control, etc.) and the
>second one does the actual logout.

I always thought the need for two of the LOGOUT PORTs was due to a bug 
in the software (I had forgotten that you do need two LOGOUT PORTs when
I wrote my post an hour or two ago).  Thanks for the information, Terry!

-Dan Wing, Systems Administrator, University Hospital, Denver
 dwing@uh01.colorado.edu or wing@eisner.decus.org

 //////////////////////////////////////////////////////////////////////////////

Newsgroups: comp.dcom.modems
Path: cs.utk.edu!gatech!howland.reston.ans.net!xlink.net!zib-berlin.de!news.belwue.de!noc!georgi
Organization: Computer Center, University of Stuttgart
Lines: 23
Distribution: world
Message-ID: <2fuo6c$bip@news.belwue.de>
References: <9312291728.AA14316@im9>
Reply-To: georgi@BelWue.DE
NNTP-Posting-Host: noc.belwue.de
Date: 30 Dec 1993 14:22:04 GMT
From: georgi@BelWue.DE (Juergen Georgi)
Subject: Re: Zmodem in unix..."rz"


In article AA14316@im9, dmiller@im.lcs.mit.edu (Dick and Jill Miller) writes:
>
> Thanks for the several suggestions, to get the Unix 'rz' utility to 
> receive my MSDOS dial-up upload using Qmodem's Z (Zmodem-batch) upload 
> transfer option.  The Unix folks at MIT do think the 'rz' utility is 
> working, but my various tests have only transferred 0 or 64 bytes of a 
> 100K file.  When I DON'T invoke 'rz' at Unix before uploading with 
> Qmodem's Z option, rz IS invoked but doesn't proceed with the transfer.  
> As I've reported, when I download using 'sz', Qmodem receives the file(s) 
> automatically and accurately.
> ... 

I missed the beginning of this thread, but the problem sounds familiar.
I also could use sz but rz failed. I finally realized that our cisco
terminal server offered a user command to set the ability to act as a
transparent pipe. If you pass through a cisco box, enter "terminal download" 
before connecting to the unix host.

--Juergen Georgi


 ===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#==

Newsgroups: vmsnet.sources.d
Path: cs.utk.edu!avdms8.msfc.nasa.gov!europa.eng.gtefsd.com!howland.reston.ans.net!cs.utexas.edu!rutgers!noao!CS.Arizona.EDU!buckie.ucha!buckie.nntp!DWING
Message-ID: <1994Jan19.133942.782@buckie.hsc.colorado.edu>
References: <2hk09o$s8a@antares.Austin.Lockheed.COM>
Reply-To: dwing@uh01.Colorado.EDU
Organization: Ski Bum Wanna-be, Incorporated
Nntp-Posting-Host: buckie.hsc.colorado.edu
Lines: 19
Date: 19 Jan 1994 19:39:39 GMT
From: dwing@uh01.Colorado.EDU (Dan Wing)
Subject: Re: getting lat port number

In article <2hk09o$s8a@antares.Austin.Lockheed.COM>,
 pr09027@austin.lockheed.com (pr09027) writes:
>
>   Does anyone have some source code to get the LAT port number that
>somebody is logging in from?  For example our terminal server has 16
>ports, and we have a program that we want to restrict functionality of it
>to certain ports.  When you do a "$show terminal" command it says
>DEC301/PORT_14 for my particular station, and we need a function, written
>in VAX/VMS Ada preferably, that would return an integer 14 in this case. 
>Thanks in advance.

From DCL, F$GETJPI(0,"TT_ACCPORNAM") returns this information; $GETJPI is 
callable from a HLL, so you can easily get that information.  The 
information is returned as character, and includes the DECserver name (very
important unless you want to put this restriction on port 1-6 of all your 
DECservers or something!!), a slash, and the port name, usually of the form 
"PORT_nn", but it can be changed by anyone that knows the privileged 
password to the DECserver.

-Dan Wing, Systems Administrator, University Hospital, Denver
 dwing@uh01.colorado.edu or wing@eisner.decus.org

 ===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#==

Newsgroups: comp.dcom.modems
Path: cs.utk.edu!darwin.sura.net!howland.reston.ans.net!vixen.cso.uiuc.edu!sdd.hp.com!decwrl!decwrl!netcomsv!netcom.com!wolfgang
From: wolfgang@netcom.com (Wolfgang Henke)
Subject: Re: Suggestions for Dialup routers/Terminal servers?
Message-ID: <wolfgangCKMurB.7p8@netcom.com>
Organization: Netcom
X-Newsreader: TIN [version 1.2 PL1]
References: <2ieum4$459@panix.com>
Date: Thu, 3 Feb 1994 05:18:47 GMT
Lines: 30

Robert Fellenz (bobf@panix.com) wrote:
: I am looking for some dialup routers and terminal servers to purchase.
: I would appreciate any suggestions that someone could give me.

Robert,

quite a few sites I know prefer Livingston Portmaster Terminal Servers
and routers. The Portmaster is an expandable terminal server with 10, 20
or 30 serial ports and supports SLIP/CSLIP/PPP and host device
emulation. You can have a look at the manual and software in the
ftp.netcom.com /pub/livingston directory. For pricing have a look in
ftp.netcom.com /pub/wolfgang/Modem.prices. 

Or more elegantly, use gopher metronet.com, select option #1, then 
option #7 for "Digicom Modem Information", then "Modem Prices available 
from WH Networks" and then "Modem Prices". Let me know if I can help 
with any other questions you may have. These units are backlogged about 
3-4 weeks currently since they are quite popular.

Cheers,

Wolfgang

-- 
     WH Networks 2672 Bayshore Pkwy Suite 503  Mountain View CA 94043  
     415.390.9316 fax: 415.964.2027      ftp.netcom.com /pub/wolfgang
     gopher metronet.com #1 #7      feenix.metronet.com /pub/wolfgang 
     wish to see me? lynx ftp://ftp.netcom.com /pub/wolfgang/wolfgang


 ===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#==

2001-01-08: Unofficial Information on Livingston Portmaster Terminal Servers
            (now a Lucent product)

    http://www.bess.net/livingston/

 ===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#==

Newsgroups: vmsnet.pdp-11,curtin.decus
Path: cs.utk.edu!news.msfc.nasa.gov!europa.eng.gtefsd.com!howland.reston.ans.net!cs.utexas.edu!uunet!world!mbg
From: mbg@world.std.com (Megan)
Subject: Re: DECSERVER-100 information
Message-ID: <CL0E9u.Mo9@world.std.com>
Organization: The World Public Access UNIX, Brookline, MA
References: <1994Feb10.192119.1@cc.curtin.edu.au>
Date: Thu, 10 Feb 1994 12:49:06 GMT
Lines: 22

zcookbruc@cc.curtin.edu.au (Bruce Cook) writes:

>The explicit problem I'm trying to deal with right now, is
>how to reset the thing.

>When I power up the RACK, the DECSERVER sends out its MOP
>request that the 11 doesn't see because it's not yet started.
>The Server then sens out another 2 requests and gives up.

>Later the 11 boots, and is ready.

Then wait until the 11 is booted before applying power to the
DS100.

>I thought the button the back reset the thing, but that has no
>effect, and I have to pull the power cord to get it to try again.

It does, but only when you press and hold it *as* you unplug it.

					Megan Gentry
					Former RT-11 Developer


 ===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#==

Newsgroups: comp.dcom.modems,comp.dcom.servers
Path: cs.utk.edu!gatech!howland.reston.ans.net!agate!library.ucla.edu!csulb.edu!paris.ics.uci.edu!news.claremont.edu!kaiwan.com!not-for-mail
From: luke@kaiwan.com (The Sky Walker)
Subject: Re: Hardware Flow control between modem & terminal server
Date: 3 Mar 1994 19:00:42 -0800
Organization: KAIWAN Internet (310/527-4279,818/756-0180,714/741-2920)
Lines: 62
Message-ID: <2l688q$9r3@kaiwan.kaiwan.com>
References: <2l3naoINN9vf@duncan.cs.utk.edu>
NNTP-Posting-Host: kaiwan.kaiwan.com
Keywords: terminal_servers flow_control

In article <2l3naoINN9vf@duncan.cs.utk.edu>,
JOHN FARMER <jfarmer@cs.utk.edu> wrote:
=
=I'm in the process of setting up a Sun Sparc with a bank of dial-in
=modem lines.  I'm planning to use a terminal server or two for the
=modem connections.  At least for the time being, I will be using
=14.4 modems (planning on AT&T Dataports).  The main question that
=I have is what kind of flow control should I/must I use between the
=terminal servers and the modems?  I suspect that hardware DTS/CTS 
=would be the best, but what would be the effect of software flow
=control?  that is the infamous XON/XOFF?  Most of the incoming
=lines will be PC's using the comm program of their choice with
=modems ranging from 2400 to 14.4k.  Some users will be making
=SLIP/CSLIP/PPP connections.
=
=-- 
=John T. Farmer, Jr.					jfarmer@cs.utk.edu
=Full-time Daddy, Some-time Grad Student,		Ask me about Internet
=Full-time wage slave, Wood butcher when I can.		Access in East Tenn.
=
=My son's hero's are Capt. Jean Luc Picard, Norm Abrams, and Mr. Rogers!


Use Hardware Flow control as you can. A terminal server without hardware
flow control is for low speed Terminals only. High speed modems can not
be connected on no-hardware flow control Termina Server.

I have used 2 low price(less than $2000.00) Terminal servers Chase and
Computone.

Their features:
Hardware flow control, 8-bit clean connection, easy menu driven
configuration, DB25 connectors(ZyXEL modems need at least 9 pins to work),
DTE speed up to 115200, support SLIP...

Chase comes with life time support.

Computone comes with 800 number support.(Talk on the phone for couple
hours to help me setup PPP/SLIP server).

Computone Terminal server also support PPP/SLIP/CSLIP and can be a
dial in PPP/SLIP/CSLIP server.

They can send you an unit for evaluation for ~$1500.
Call 1-800-241-3946 and ask for Matt Johnson. You can mention Luke
from KAIWAN CORP to him.

Last time I ask about Zylogics' price is way over $2000.00.
No chance to get an evaluation unit from them.

PS. Below is my AT&T setup for Terminal server. You might find it userful.
atl0m0e0q1v0x0&c1&r0\d1\g1\x1\t5s0=1s79=1
at&w0

-- 

 ===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#==

Newsgroups: comp.dcom.modems
Path: cs.utk.edu!gatech!howland.reston.ans.net!math.ohio-state.edu!cs.utexas.edu!convex!news.duke.edu!news-feed-1.peachnet.edu!paperboy.wellfleet.com!noc.near.net!xap!Ralphs-PC!rbecker
From: rbecker@sup.xyplex.com (Ralph Becker)
Subject: ZModem on Xyplex servers
Date: Fri, 6 May 1994 17:33:19 EDT
Organization: Xyplex Customer Support
Lines: 68
Message-ID: <rbecker.89.06CB501A@sup.xyplex.com>
NNTP-Posting-Host: rgb.xyplex.com
Keywords: xyplex zmodem
X-Newsreader: Trumpet for Windows [Version 1.0 Rev B]

I cannot take credit for the following information, but I post it here for
the convenience of everyone who may have encountered this question.
Happy connecting!

===========================================================

I have been investigating how to get ZMODEM to work through Xyplex terminal
servers without changing the session mode from interactive. After some
work, I've got it working with a couple of terminal emulation/ZMODEM
PC programs.

Set up the terminal server port as follows:

SET PORT TYPEAHEAD SIZE 1024 (or larger)
SET PORT INTERNET TCP WINDOW SIZE 1024 (or larger)
SET PORT TELNET CSI ESCAPE ENABLED
SET PORT TELNET NEWLINE FILTERING STANDARD

Here's why you need to do all this...

ZMODEM uses 1024 byte packets by default at speeds of 4800 bps or
greater, so the typeahead buffer and TCP window size need to support
that.

CSI escape enabled is necessary to prevent telnet in the server from
converting certain escape sequences from one format to another.

Newline filtering standard causes CR/NUL received from the network
to be sent to the serial line (the PC) as CR, but leaves CR/LF as is.
Apparently some ZMODEM implementations, like the one in DCA's Crosstalk
cannot deal with CR/NUL.

In my investigation, I was doing ZMODEM transfers between a Unix work
station using the sz and rz commands, and a PC using DCA's Crosstalk
for Windows versions 2.0 and 2.1 or the shareware utility dsz.

Notes for each package:

sz/rz on Unix:

To move a binary file from Unix to a PC using sz, specify the b and e
options or you will get a failure. The b options says to not convert
any characters, and the e option says to escape all control characters.

dsz on PC:

This worked great with no problems at all. Keep in mind that dsz does
ZMODEM transfers only and does not provide terminal emulation.

Crosstalk for Windows:

I had a lot of trouble getting this to work, but it was setting telnet
newline filtering properly that got it to work at all in interactive
mode. Using version 2.0 to send files from the PC to Unix automatically
invokes rz on the Unix side to receive the file. Version 2.1 does not
do that, so you have to start rz on the Unix side first. Also with 2.1,
when you start rz or sz on the Unix side, it causes the Crosstalk side
to start up automatically, i.e., for rz it pops up the window you use
to select files for sending, and for sz it actually starts the transfer
into the default receive directory on the PC side. With 2.0 or 2.1, I
would often get CRC errors on a transfer, but it did end up transferring
files ok. Finally, with 2.1, Crosstalk would sometimes crash during a
transfer. Looks like there could be some problems with Crosstalk 2.1.


Ralph Becker
Xyplex Customer Support [Tech. Support hotline 800-435-7997]
rbecker@sup.xyplex.com or 71174.1262@compuserve.com

 ===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#==

Newsgroups: comp.protocols.kermit.misc
Path: cs.utk.edu!gatech!bloom-beacon.mit.edu!uhog.mit.edu!news.mathworks.com!news.alpha.net!uwm.edu!lll-winken.llnl.gov!sol.ctr.columbia.edu!news.columbia.edu!watsun.cc.columbia.edu!fdc
From: fdc@watsun.cc.columbia.edu (Frank da Cruz)
Subject: Re: Help: file transfer problem
Date: 22 Nov 1994 14:31:55 GMT
Organization: Columbia University
Lines: 30
Message-ID: <3asvcr$qfg@apakabar.cc.columbia.edu>
References: <CzMLA2.K01@liverpool.ac.uk> <3aqlhu$7q6@apakabar.cc.columbia.edu> <CHANG.94Nov21182356@theta.math.wsu.edu>
NNTP-Posting-Host: watsun.cc.columbia.edu

In article <CHANG.94Nov21182356@theta.math.wsu.edu>,
 Ching Mo Chang <chang@theta.math.wsu.edu> wrote:
>
>I also have the file upload problem with MS-Kermit (Also C-Kermit in OS/2),
>I can download file with packets length at 5012, but when I use the same
>setting to upload file I got lots of timeout and eventually end up too many
>retries error and fail to upload file. If I really need to upload file, I
>need to increase the retry limit and set the packets length to 92, let the
>file crawl from my PC to the host or I will just paste the file into an editor
>in my host side (only work in ASCII file).

This is not Kermit's fault.  You should be thankful that Kermit lets you
adjust these parameters to make the file transfer work at all.  Why is this
happening?  It's hard to say without more information, but the most likely
culprit is a lack of buffer capacity in the "upstream" direction, coupled
with a lack of effective flow control -- a fatal combination, and a common
one.

Many communications processors (terminal servers, front ends, host console
drivers) are designed on the assumption that traffic *to* the host consists
of nothing but keystrokes -- which hardly anybody can produce at more than
about ten per second (= 100 bps) -- whereas traffic in the downstream
direction is voluminous -- file listings, etc.  So they have big output
buffers and tiny input buffers.

To compound the problem, some communications processors take this assumption
one step further and do not even provide flow control in the upstream
direction, because they figure they will never need it.  A well-known example
is the Cisco ASM series of terminal servers.

- Frank

 //////////////////////////////////////////////////////////////////////////////
 ===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#==

Newsgroups: comp.os.vms
Path: cs.utk.edu!gatech!psuvax1!uwm.edu!news.alpha.net!mvb.saic.com!info-vax
From: John Nebel <nebel@athena.csdco.com>
Subject: Re: *VERY BAD* experience with DATABILITY terminal server
Message-ID: <Pine.ULT.3.91.950313072353.13522A-100000@athena.csdco.com>
Date: Mon, 13 Mar 1995 07:34:42 -0700 (MST)
Organization: Info-Vax<==>Comp.Os.Vms Gateway
X-Gateway-Source-Info: Mailing List
Lines: 25

On Sun, 12 Mar 1995, it was written:

> I have had a Lantronix EPS12 server for a couple of years. It freezes once a
> few months.  I thought that was bad.  So this time I bought a DATABILITY
> VCP200 terminal server with both LAT and TCP support.  Since the day the
> DATABILITY arrived, I regret my poor judgment every day now.
> 
cut
> 
> Worse of all, it freezes at least once every day.  When it freezes, the only
> command that works is probably HELP and CRASH which allows you to reboot.
> I'm getting tired of going down to the lab and reboot it every day.
> 

I've got a number of Datability VCP-2000s and they work fine for LAT
(never a crash) - so the things are good for something if you can't take
it back. 

For tcp/ip, the DECserver 700 is great.  The 16 port model will work fine 
with dial-in lines and it supports SLIP and PPP (assuming DNAS software, 
v1.5).  PC dial-in users (Mac, Windows, and Linux) think it's fine.

John Nebel 

 ===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#===#==
 //////////////////////////////////////////////////////////////////////////////

Newsgroups: comp.dcom.lans,comp.unix.questions,comp.protocols.tcp-ip
Path: cs.utk.edu!stc06.ctd.ornl.gov!fnnews.fnal.gov!nntp-server.caltech.edu
     !netline-fddi.jpl.nasa.gov!elroy.jpl.nasa.gov!swrinde!gatech
     !news.mathworks.com!transfer.stratus.com!xylogics.com!Xylogics.COM!carlson
From: carlson@Xylogics.COM (James Carlson)
Subject: Re: terminal Servers: identify port?
Date: 12 May 1995 11:19:15 GMT
Organization: Xylogics Incorporated
Lines: 38
Distribution: world
Message-ID: <3ovg7j$cqh@newhub.xylogics.com>
References: <3okgo9$1ji@raffles.technet.sg>
Reply-To: carlson@xylogics.com
NNTP-Posting-Host: newhub.xylogics.com

In article <3okgo9$1ji@raffles.technet.sg>,
 mathias@unicorn.swi.com.sg (Mathias Koerber) writes:
|> Hi,
|> 
|> I would like to know whether there are terminal servers that allow me to
|> identify (from a UNIX app) which port a terminal is connected to? I 
|> need this for accounting and security information, but so far the
|> nearest I can get is identifying the IP address of the Terminal server.

Usually, this is a failing of the host, rather than of the terminal
server.  In our terminal servers, we provide a means to discover the
serial port number if you're able to find the remote TCP port number (we
choose the numbers according to a simple published algorithm), but most
systems will not transmit this information to the application through
the telnetd.  We also provide TELNET SNDLOC support to send an arbitrary
"location" string up to the telnet daemon, but most hosts do not support
this, either.  And, finally, we have tap/ident and finger protocols
which can both identify users if you have the right information, but,
again, most implementations of host daemons do not give you access to
the information you need.

(We also supply a host tool called rtelnet which could be used to fix a
serial port to a predetermined pseudo device on the host, but I
certainly would not recommend this.)

The only reliable means of doing this I've seen -- and it certainly
isn't good for security in some applications (although it does work in
some restricted cases) -- is to set the terminal type on the serial
port with a special string.  Most (but not all) hosts correctly query
the terminal type string from the terminal server and will make it
available to applications.  If you set the terminal type string on the
terminal server to include the location of the port, you can then parse
this information out into the port number and actual terminal type,
perhaps in a .login script.

---
James Carlson <carlson@xylogics.com>            Tel:  +1 617 272 8140
Annex Software Support / Xylogics, Inc.               +1 800 225 3317
53 Third Avenue / Burlington MA  01803-4491     Fax:  +1 617 272 2618

 //////////////////////////////////////////////////////////////////////////////

Newsgroups: comp.dcom.modems,comp.dcom.servers
Path: cs.utk.edu!nntp.memphis.edu!nntp.msstate.edu!cssun.mathcs.emory.edu!swrinde!howland.reston.ans.net!ix.netcom.com!netcom.com!megazone
Message-ID: <nntpuserDHCGBA.EC@netcom.com>
Date: Wed, 1 Nov 1995 02:59:34 GMT
Sender: netnews@mork.netcom.com
References: <intercomDGv48D.Bn0@netcom.com> <46ftu4$j3d@newhub.xylogics.com>
 <46s3sc$1aj@news.vector.net> <BILLW.95Oct31010748@puli.cisco.com>
From: megazone@livingston.com (Brian 'MegaZone' Bikowicz)
Subject: Re: Rackmounts/ Xylogics vs. Multitech


billw@puli.cisco.com (William ) shaped the electrons to say:
>
>I'm curious.  Just what does Livingston mean when they say they can run all
>ports at 115.2kbps?  Does that mean they can accept VJ Header Compressed

I phrased that poorly, I admit.  I already answered this in email for
someone else.

---
We can handle all ports set at 115200 with no loss via buffering.  Depending
on overhead we can pass 100,000-150,000 bytes at a sustained rate from the
ether to the serial ports.  (PPP takes more overhead that SLIP for example.)
1-1.5 Megabits, not 6.9 Megabits.  Sorry if I wasn't clear.
---

Technically we can do better, but we say 1-1.5 to be conservative.

>Ask me sometime about the !@#!%$$ testers at cisco who bombarded a cisco-500
>with 14000 pps on the Ethernet, all destined for a single async line, and
>wondered why things weren't working very well. (Yes, the processor ended up
>spending nearly all it's time throwing away packets...)

I think most of us have seen examples like that.  If not internally, then
users.

"Why doesn't this work?"  "It's not supposed to work."

BTW, the ISDN units were announced at Internet World yesterday, so there 
are some new things on our web for the curious.

Don't blame the markup on me, I haven't had time to gut the pages and make
them decent HTML yet... ;-)  Someone else marked them up.

-MZ
-- 
Livingston Enterprises Technical Support
Phone: 800-458-9966      FAX: 510-426-8951
support@livingston.com  <http://www.livingston.com/> 
6920 Koll Center Parkway  #220, Pleasanton, CA 94566

 //////////////////////////////////////////////////////////////////////////////

Newsgroups: comp.dcom.servers
Path: cs.utk.edu!cssun.mathcs.emory.edu!swrinde!sdd.hp.com!usc!chi-news.cic.net
      !newsfeed.internetmci.com!news.exodus.net!news.wni.com!noc.netcom.net
      !noc.netcom.net!usenet
From: "T.C." <chan@pointcom.com>
Subject: Top Servers Related Sites
Date: 8 Nov 1995 15:39:03 GMT
Organization: Point Communications
Message-ID: <47qiun$o4i@news2.noc.netcom.net>
NNTP-Posting-Host: 205.186.243.53
X-URL: newsrc://nntp.noc.netcom.net/


Dear Friends:

We'd like to invite you to check out our reviews of Servers-related 
sites.  Point reviews and rates thousands of the very best sites on the 
Web, including sites in this area of interest.

There's no charge and no catch.  We hope our lively, independent reviews 
will prove valuable to members of newsgroups like this one.  We hope 
you'll stop by at http://www.pointcom.com.  Thanks for your time.

The sites are rated from 0-50, in: Content, Presentation and 
Experience(sorted)

Unisys	36	41	38
Sun Microsystems	42	45	38
Digital Equipment Corporation	37	41	38
Cray Research Inc.	34	40	38
ANDATACO	32	40	38
Auspex Systems, Inc.	34	40	35
Northeast Parallel Architectures Center	36	31	27
Advanced Logic Research (ALR)	28	29	25


 //////////////////////////////////////////////////////////////////////////////

Newsgroups: comp.terminals
Path: utkcs2!stc06.ctd.ornl.gov!news.er.usgs.gov!jobone!news2.acs.oakland.edu
      !newsxfer.itd.umich.edu!newsxfer3.itd.umich.edu!news-peer.sprintlink.net
      !news-peer.sprintlink.net!news.sprintlink.net!sprint!howland.erols.net
      !europa.clark.net!smartdna!news.io.com!not-for-mail
Date: 9 Apr 1997 11:49:40 GMT
Organization: Illuminati Online
Message-ID: <01bc44a9$f20ea8a0$5e59aac7@mcclure3>
References: <01bc4275$f9e12520$5c59aac7@mcclure3>
From: "Rory McClure" <rory@io.com>
Subject: Re: MICOM to MICOM connection

I found my answer, so I'm posting an answer to my own post...

Rory McClure <rory@io.com> wrote in article
<01bc4275$f9e12520$5c59aac7@mcclure3>...
>
>    I'm looking for the RS232 pin outs for cabling a MICOM 8000 series
> modem-multiplexer directly to another MICOM 8000 without the intervening
> telco or modem-telco-modem.  I have tried a PC null-modem
> (2->3,3->2,4->5,5->4,6->20,20->6), and variations, but that doesn't seem to
> work.

>    These units are rs232 analog modem multiplexers and at least 5 years
> old.  One unit would take analog (tty) lines from a host system, multiplex
> them together and send them out through an rs232 port to either a
> dedicated-line telco (4 wire connection) or to a modem.  The other unit
> takes input through a similiar rs232 port, dedicated-line telco or modem,
> and de-mulitplexes it into several terminals.
>
> The VAR who sold them to us is chapter-11 -- so I'm looking for a paddle,
> or even the creek...
>
> Rory McClure, I/S Manager
> Capitol Bolt & Supply, Inc.
> rory@capbolt.com


To hook up the MICOM rs232 DB25 to the 4-wire telco you use:
        DB25            Telco
         6      -->     RXB  (the black wire)
        14      -->     TXB  (the green wire)
        16      -->     TXA  (the red wire)
        22      -->     RXA  (the yellow wire)

To hook the MICOM up to another MICOM, you use:
        DB25            DB25
         6 RXB  -->     14 TXB
        14 TXB  -->      6 RXB
        16 TXA  -->     22 RXA
        22 RXA  -->     16 TXA

   HOWEVER, you need to be careful in the back-to-back arrangement because
the MICOM units usually transmit the signals "hot" (at 0dbm for US phones)
but expects the signals at -8dbm to -12dbm.  If this arrangement causes
Link Alarms (LA), it is a good sign that this is causing problems.

   To fix this, there are jumpers on the on-board modem inside the unit
that will 'cool down' the transmit power level.

Rory McClure, I/S Manager
Capitol Bolt & Supply, Inc.
rory@capbolt.com

 //////////////////////////////////////////////////////////////////////////////

Newsgroups: comp.unix.sco.misc,comp.unix.admin
Path: utkcs2!stc06.ctd.ornl.gov!fnnews.fnal.gov!uwm.edu!vixen.cso.uiuc.edu
      !howland.erols.net!infeed1.internetmci.com!newsfeed.internetmci.com
      !news.infi.net!news.infi.net!not-for-mail
Organization: Lake Taylor Hospital
Message-ID: <33D602DE.167E@yy.laketaylor.org>
References: <33D5009B.2781@yy.laketaylor.org> <33d5345c.0@samba>
Date: Wed, 23 Jul 1997 09:10:54 -0400
From: "Mark A. Davis" <mark@yy.laketaylor.org>
Subject: Re: Sh tricks/holding ports open
X-Archive-Subject: [printing via Digi Portserver]

Robert Lipe wrote:
>
> In <33D5009B.2781@yy.laketaylor.org> "Mark A. Davis" <mark@yy.laketaylor.org>
writes:

Yo Robert :)  I knew the word "Digi" on this group would bring you to life :)


> >I have need to hold printer ports open with ditty/stty settings.  This
> >is needed for non-lp printing and for keeping ports straight (no resets-
> >Digi Portservers) even for lp printing.
>
> Does that RealPort driver support the 'ditty printer' option?

No.  It is my understanding that the "printer" option is unsupported,
and I don't know exactly why- to quote the (recent) docs "printer- not
supported by this device driver".

> If so, you shouldn't have to do any of this stuff.

Rub it in!!!

>   It essentially inhibits
> final close/initial open tty action from happening.
>
> >Anyway, I am putting the following in inittab for each port that needs
> >to be held open:
>
> >Nf01:23:respawn:/usr/taylorbin/lpgetty ttya01 50 fastbaud dsrpace
> >-fastcook
>
> >lpgetty is just a script as follow:
>
> >port=$1
> >shift
> >(ditty $* ; while : ; do sleep 7200 ; done ) < /dev/$port
>
> >Well, this appears to work ok.  But the problem is this- when that
> >process is running, it just looks like a "sleep".... I can't tell which
> >port it is running on!  Why???
>
> >shell>ps -e | grep sleep
> > 12979  ?        0:00 sleep
>
> Don't look at the sleep.   Look for the lpgetty process.

shell>ps -ef | grep lpgetty
    root  4095     1  0 11:49:32  ?        0:00 INITSH /etc/initscript
Na01 2 re
spawn /usr/taylorbin/lpgetty ttya01 50 fastbaud
    root  4096  4095  0 11:49:32  ?        0:00 INITSH /etc/initscript
Na01 2 re
spawn /usr/taylorbin/lpgetty ttya01 50 fastbaud


EEK!  Bingo, I never thought of looking for lpgetty, mostly because it
does not show up in a ps -e.  Silly me.  But, it CAN be tricky- "ps -e"
and "ps -f" do not show it, ONLY "ps -ef" together.  I simply HAVE to
remember that.

> { Which is
> a little misnamed becusae it's not really going to "get a tty" ready
> for login, but that's not he point. }

Do you have a better name?????????  :)  :) :)
Really- I was using to use "lpset", but that is already taken for the
moment.  Then I was going to use "printer", or "porthold" or something.
Then I just gave up and picked something that at the time (sort-of)
seemed catchy.

> >Gettys show which port IT is running on, and it is run from inittab the
> >same way:
>
> >shell>ps -e | grep f01
> > 28480  f01      0:01 getty
>
> This is because getty sets up a controlling tty and your script doesn't.

Oh!  Ok, now I get it.

> See if 'ditty printer' is there.  I can't find the source for it
> right now, or I'd look for you.

shell>ditty printer
DIGI_GETA: Invalid argument

:)

By the way, the other reason I want something to hold the printer ports
open all the time is a strange error that I sometimes get emailed to
root from lp.  Here it is:
.....................................
Subject: Problem with printer pa_laser

The printer pa_laser has stopped printing for the reason given below.
Fix the problem and bring the printer back on line.
Printing has stopped, but will be restarted in a few minutes;
issue an enable command if you want to restart sooner.
Unless someone issues a change request

        lp -i pa_laser-445646 -P ...

to change the page list to print, the current request will be reprinted
from
the beginning.

The reason(s) it stopped (multiple reasons indicate repeated attempts):
Failed to open the printer port.
(No such device or address)
......................................

It does, indeed, try again a little later and prints fine.  My theory is
that there are times when the Portserver is not ready for some reason.
With 30 printers, I get about 5 or 6 of these messages a week.  My tests
indicate that if the port is held open by something, then the problem
never occurs on that port.  Very odd.  (and yes, the ditty settings are
actually in the /usr/spool/lp/admins/lp/interfaces/*  so it should be
set right).  Looks like a driver problem to me.  So, unfortunately, my
goal is to run this "lpgetty" for EVERY printer port, not just the 3 or
4 we need to have direct printing :(

Actually, my test was with "cat < /dev/$port > /dev/null &", the lpgetty
now uses sleep, and I haven't tested that yet to see if it works to
solve the "no such device" problem... but it has two advantages- it
keeps resetting the port over time, which is good "just in case," and the
cat process starts eating up CPU time, probably because it is ACTUALLY
redirecting all the activity on the port.

Oh, by the way, have I said how much I LOVE these Digi PortServers,
otherwise????????  I DO!!!!!  HIGHLY RECOMMENDED STUFF!!
--
/--------------------------------------------------------------------\
|   Mark A. Davis,  |Lake Taylor| Voice: (757)-461-5001x431 8-4:30ET |
|    Director of    | Hospital  | mark@yy.laketaylor.org ***to reply |
|Information Systems|Norfolk, VA| from USENET remove anti-spam "yy." |
\--------------------------------------------------------------------/

 //////////////////////////////////////////////////////////////////////////////

From stevenu@poboxes.com Wed Oct 29 14:30:00 1997
Date: Fri, 24 Oct 1997 01:42:39 GMT
From: "Steven P. Underwood" <stevenu@poboxes.com>
Reply-To: sunderwood@kopin.com
Newsgroups: comp.os.vms
Subject: Re: DECserver 200 problem

I have seen this when the segment is nearing its maximum length. This
would seem to be supported by the fact that you could terminate the
segment and it works. If possible, try to shorten the entire segment
by just one length (remove the last node) and see if the problem
disappears.

Good Luck

Steve

On 23 Oct 1997 15:56:25 GMT, gartmann@immunbio.mpg.de (Christoph
Gartmann) wrote:

>Hello,
>
>we have a DECserver200 that doesn't boot anymore. Instead it tells (on a
>terminal attached to one of its ports): 
>   local 941 - transceiver loopback error
>I have replaced the transceiver, the AUI-cable, the server, still the same
>problem. Then I opened the network cable (thinwire ethernet, BNC) and
>terminated the segment immediately after the server. Now it is booting and
>behaving all right. According to the manual either the server, the transceiver
>or the cable is faulty. But this cannot be the case (see above). Thus, what
>else on the net could cause such a problem (I don't see any problem besides
>this boot-problem on the segment)?
>
>Regards,
>   Christoph Gartmann
>
>+----------------------------------------------------------------------------+
>| Max-Planck-Institut fuer      Phone   : +49-761-5108-465   Fax: -452       |
>| Immunbiologie                                                              |
>| Postfach 1169                 Internet: gartmann@immunbio.mpg.de           |
>| D-79011  Freiburg, FRG                                                     |
>+----------- Do you know MENUE, the user environment for OpenVMS? -----------+

Steven P. Underwood,DNRC
Whitinsville,MA
stevenu@poboxes.com


 //////////////////////////////////////////////////////////////////////////////


In-Reply-To: <Pine.HPP.3.95.980318120157.20305A-100000@pollux.salleURL.edu>
Message-ID: <Pine.SOL.3.96.980318102615.29716D-100000@costello>
Date: Wed, 18 Mar 1998 10:31:11 -0500 (EST)
From: "Richard S. Shuford" 
To: Rodriguez Otero Pedro <is05424@salleURL.edu>
Subject: Re: I need help with a Digital Dec Server 100

Sir Rodriguez:

>Date: Wed, 18 Mar 1998 12:13:05 +0000
>From: Rodriguez Otero Pedro <is05424@salleURL.edu>
>Newsgroups: comp.terminals
>Subject: I need help with a Digital Dec Server 100
>Last week I bought a Digital Dec Server model 100, without any
>documentation and soft, please can anybody help me to obtain documentation
>bout this dumb terminals server.
>I visit the Digital web page, and I obtain technical information, but I
>don't obtain the soft.
>What is the red button in the case,( had a legend SW and a green led)?
>What I can do it with it(I have digital vt 320 dumb terminals)?
>And What about the DECNET protocol for Linux?
>Please email me about this questions. Thanks.
>    Pedro Rodriguez
>   La Salle Bonanova                        e-mail:   is05424@SalleURL.edu
> Universitat Ramon Llull                    Tel:      939064388


You may have a problem.  If this really is a Model 100 of the original
series of DEC terminal servers, then it speaks only the LAT protocol,
not TCP/IP-Telnet.

And you may have to have either a VMS or Ultrix computer on your local
Ethernet segment to download the software into it (via the MOP protocol).

I know of nobody using Linux with DECnet.  Anyway, LAT is not strictly
part of DECnet anyway.

 ...RSS


 //////////////////////////////////////////////////////////////////////////////

Newsgroups: comp.sys.dec
NNTP-Posting-Host: 208.31.209.254
NNTP-Posting-Date: Wed, 10 Dec 2003 19:56:36 +0000 (UTC)
Message-ID: <1ef525c1.0312101156.28445b1d@posting.google.com>
Date: 10 Dec 2003 11:56:36 -0800
From: G. W. Hairston <ghairst@mcvh-vcu.edu>
Subject: How to redirect terminals on Dec servers to a particular cluster node

How can we redirect terminals on DECservers from one cluster node to another?

Look at "preferred service".  We're not that familar with DEC service.
Any help is appreciated. See below!!!

___________________________________________________________

EX:  set host/mop MCVS65

    > show port 5

Port  5: fsutton                       Server: MCVS65
Character Size:            8           Input Speed:               9600
Flow Control:            XON           Output Speed:              9600
Parity:                 None           Signal Control:        Disabled
Stop Bits:           Dynamic

Access:                Local           Local Switch:              None
Backwards Switch:       None           Name:                    PORT_5
Break:                 Local           Session Limit:                4
Forwards Switch:        None           Type:                      Soft
Default Protocol:        LAT

Preferred Service: MCVHCLUSTER     <<<< We would like to changed this to MCVAP!

Authorized Groups:   0
(Current)  Groups:   0

Enabled Characteristics:
Autobaud,  Autoconnect,  Autoprompt,  Broadcast,  Failover,
Input Flow Control,  Lock,  Loss Notification,  Message Codes,
Output Flow Control,  Verification

 ..............................................................................

Newsgroups: comp.sys.dec
NNTP-Posting-Host: d65318a8.newsreader.visi.com
NNTP-Posting-Date: 10 Dec 2003 16:35:55 CST
References: <1ef525c1.0312101156.28445b1d@posting.google.com>
Message-ID: <3fd79fcb$0$41291$a1866201@newsreader.visi.com>
Organization: VISI.com
Date: Wed, 10 Dec 2003 16:35:55 -0600
From: Mike Dorn <mrdorn@wavefront.com>
Subject: Re: How to redirect terminals on Dec servers to a particular cluster node

[the command is]

    > change port 5 pref MCVAP

Remember, there's syntax help on the server by just typing "help"

-- 
Mike

 //////////////////////////////////////////////////////////////////////////////


Message-ID: <71nr17$2ej$1@autumn.news.rcn.net>
References: <363DCD9D.85001A93@millerbros.co.uk>
Date: 3 Nov 1998 21:02:31 GMT
From: winter+spam@jurai.net
Newsgroups: uk.adverts.computer, comp.terminals, comp.terminals.tty5620,
    comp.terminals.bitgraph, comp.sys.hp.misc
Subject: Re: FS: Emulex 4000 Terminal Servers

In comp.terminals.tty5620 Scott Reid <scott.reid@millerbros.co.uk> wrote:
> FOR SALE
> We have many Emulex Performance 4000 32 port terminal servers for sale.
> Paid over 2000ukp for them new, many only a few months old.

> Offers around 1000ukp

Unless you're selling the software with them they're not worth that much.

For those that don't know these are pretty much clones of DEC 700 terminal
servers.  Cute little boxes.  Fans need regular attention.  I've got 6 of
them at home I use for console servers. (don't ask)  With the 16 port add-on
for 32 ports, your max speed drops to 19200, not that I'd expect anyone to
do anything outside of 9600,n,8,1 with these things.

They're fun to play with if you can get a good deal on them.

--
| Matthew N. Dodd  | 78 280Z | 75 164E | 84 245DL | FreeBSD/NetBSD/Sprite/VMS |
| winter@jurai.net |      This Space For Rent     | ix86,sparc,m68k,pmax,vax  |
| http://www.jurai.net/~winter | Are you k-rad elite enough for my webpage?   


 //////////////////////////////////////////////////////////////////////////////

Newsgroups: comp.sys.dec
Message-ID: <3695E5B1.B19BE09D@hep.phy.cam.ac.uk>
References: <772j4a$9aq@hermes.acs.unt.edu> <772t3f$ora$1@winter.news.rcn.net>
Organization: University of Cambridge
Date: Fri, 08 Jan 1999 11:02:11 +0000
From: John Hill <hill@hep.phy.cam.ac.uk>
Subject: Re: Connecting DECserver 200/MC to VS3100

Kenneth Randell wrote:

> Do you have the software necessary for this DECSERVER???
> When the DS200/MC 'boots', it sends a message out via MOP (I think),
> requesting a down-line load of the software, although frankly, I cannot
> remember
> the name of the loaded image that is stored in MOM$LOAD or MOM$SYSTEM.
> At one time, this was software you had to purchase additionally from any OS
> or other software you had; I believe the license to use it came with the DS
> 200.
> They stopped updating it some years ago; I think the last version was 3.3A,
> which gave you LAT 5.1, and some other features.
>
> Without this DS 200 software, your terminal server is essentially unusable.
>
> If you hook a dumb terminal to port 1 at 9600, n, 8, 1 (you only need to
> hook up pins 2,3, and 7, reverse 2+3), you will see the messages of what it
> is trying to do.  Modem control obviously requires more pins to be hooked up.
>
> Ken Randell
> Datametrics, Inc.
>
> David Mason Curry wrote in message <772j4a$9aq@hermes.acs.unt.edu>...
> >
> >I know next to nothing about networking, but I bought a DECserver 200/MC
> >to connect to my VAXstation 3100 so that I could have a couple of
> >terminals and a modem.  I hooked up an AUI to 10base-2 transceiver to the
> >AUI port on the DECserver and connected the 10base-2 to the VS3100 and
> >selected the 10base-2 port (at least the green light by that port came
> >on).  So I plug in the DECserver and the D1 LED comes on steady and the
> >D2 LED blinks a while, pauses, blinks a while, pauses, ad infinitum.  I
> >powered on the VAXstation and it reports that nothing is connected to
> >the ethernet port.  However, the transceiver shows that the AUI port on
> >the DECserver has at least power (The little light on the transceiver
> >comes on).  What naive thing am I doing wrong?  I would appreciate any
> >and all help as my wife has regulated my computer to the closet and I am
> >only allowed to work on it from terminals that will be in my office and
> >garage :-(  BTW, I almost forgot to mention, on the DECserver, lights
> >D3 and D4 never come on.
> >
> >Cheers,
> >David Curry


The DS200/MC boots using MOP, and expects to find  a file called
PR0801ENG.SYS in MOM$LOAD (or MOP$NAMED_LOAD if using DECnet-Plus)
on the host system.

The software is still on the layered products CD distribution for both
VAX and ALPHA.  The license was sold bundled with the hardware. However.
even though PAKs were issued, the software doesn't use LMF, so there is
no need to register them. The meaning of the LEDs are:

D1 - power ON/OFF
D2 - diagnostic (ON=self-test passed;OFF=fatal error or testing;blinking
          =non-fatal error)
D3 - software download status (ON=software loaded, OFF=load in progress;
          BLINKING=load failed)
D4 - network activity (%age of time on indicates traffic levels).

From your description, you are getting self-test errors which are considered
non-fatal..

One thought - does the AUI side of the transceiver have heartbeat (==SQE test)
enabled?  Some of the older Ethernet devices require heartbeat (and I certainly
have it enabled on my transceivers attached to the two DS200/MC that we still
run).

Regards,
John Hill

--
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
+  Dr. John C. Hill                  University of Cambridge,       +
+  E-mail: hill@hep.phy.cam.ac.uk    Cavendish Laboratory,          +
+  Phone:  +44-1223-337243           Madingley Road,                +
+  Fax:    +44-1223-353920           Cambridge, CB3 0HE, U.K.       +
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

 //////////////////////////////////////////////////////////////////////////////

Newsgroups: comp.sys.dec
Date: 8 Jan 1999 15:22:48 GMT
Subject: Re: Connecting DECserver 200/MC to VS3100
From: David Mason Curry <dcurry@silo.csci.unt.edu>

I want to thank everybody for the help!  I now understand what I need to
do to get a DS200 working... but mine never will since I didn't get the
software with it.  I say never, because DEC/CPQ/some_spode said that, yes
they would be more than happy to sell the software for personal home use
and that they would give it to me for the cheap personal price of $1,300.
What planet are they from????  They can bite me :-)  Oh, well... I will
just have to wait until someone sells a DS200/MC w/software on eBay.  At
least I only wasted $8.50.  I have never transacted with DEC before, but
now I understand why everyone bitches so much.  If the rest of the world
worked like DEC, we would be expected to pay $1,500 for the software to
run a 10 year old 1200 baud modem on a Trash-80.

Cheers,
David Curry


 //////////////////////////////////////////////////////////////////////////////


Date: Fri, 08 Jan 1999 11:02:58 -0500
From: Jeff Morgan <Jeff.Morgan@coats.com>
Newsgroups: comp.sys.dec
Subject: Re: Connecting DECserver 200/MC to VS3100

David Mason Curry wrote:
>
> I want to thank everybody for the help!  I now understand what I need to
> do to get a DS200 working... but mine never will since I didn't get the
> software with it.

David:

  Consider locating a DECserver 90L+ terminal server. The software is
built-in to the ROM so it is live on power-on. Uses MMJ connectors with
8 serial ports. They should be available in the $100 range by now what
with VT style terminals dying off for pc's, nc's, wbt's and
javastations...


  You'll have to use the serial port on the VAXstation for the modem.
The 90L+ doesn't support modem control.

  Also, since you have the server and supposedly a license to run the
software, you might make friends with your local DEC Field Service
office and see if they will "loan" you the kit from CONDIST.

                                *JM*

 //////////////////////////////////////////////////////////////////////////////


Date: Fri, 08 Jan 1999 10:37:12 -0600
From: Chris Scheers <asi@airmail.net>
To: David Mason Curry <dcurry@silo.csci.unt.edu>
Newsgroups: comp.sys.dec
Subject: Re: Connecting DECserver 200/MC to VS3100

In your original message you said that when you powered up the VS3100,
it reported that nothing was on the Ethernet.  Do you mean that you got
an error from the console?  If so, did you terminate the network?

The correct way to connect a DS200 to a VS3100 is to put the transceiver
on the DS200, put a T connector on both the transceiver and VS3100,
connect a cable between the Ts and put a 50ohm terminator on the unused
leg of each T.  Without the terminators, the self test will probably
fail (and the network won't work).

As has been mentioned, you need the download file.  The license to use
it comes with the DS200 hardware.

Once you have the network connection and download file, you need
something to download the software from the VS3100 to the DS200.  For
that you need either DECnet or LANACP (depending on your version of
VMS).


It sounds complicated, but its not so bad.  I'm not too far from you,
give me a call if you have any questions.

Good luck!


-----------------------------------------------------------------------
Chris Scheers, Applied Synergy, Inc.

817-237-3360 (Voice)    817-237-3074 (Fax)    Internet: asi@airmail.net




Date: 8 Jan 1999 16:15:09 GMT
From: ByerRA <byerra@aol.com>
Newsgroups: comp.sys.dec
Subject: Re: Connecting DECserver 200/MC to VS3100


You have to have software in order to run the
DECServer 200/MC.

If you need it, I can e-mail you a ZIP file with the install software set so
you can install it on your machine.



 /////////////////////////////////////////////////////////////////////////////


Message-ID: <WLwp3.707$dP4.15535@dfw-read.news.verio.net>
Date: Tue, 03 Aug 1999 07:21:26 GMT
Organization: None of Your Business
Newsgroups: comp.sys.sun.hardware
From: Travis Dixon <travisd@shell.clark.net>
Subject: Re: Terminal Servers

John-Paul Pagano <fdmitri@concentric.net> wrote:
>
> Can anyone recommend a good terminal server solution for a Sun shop?

I'm partial to the Bay/Xylogics/Nortel/whatever "Annex" series -
important thing here is that they don't send a break to the serial
lines upon a reset. The RA4000 will support up to 72 lines which
appear on RJ21x telco-style connectors - so you can make them appear
however you want.


 /////////////////////////////////////////////////////////////////////////////


Newsgroups: comp.terminals
Organization: General Dynamics Communication Systems
Message-ID: <3974D033.9D9A8659@GD-CS.Com>
References: <GtJc5.327973$MB.5423623@news6.giganews.com>
X-Mailer: Mozilla 4.73 [en] (WinNT; U)
X-NNTP-Posting-Host: n206241180182.gdgsc.com
Date: Tue, 18 Jul 2000 21:46:27 GMT
To: Thomas <thollay@emcosales.com>
From: "Scott G. Hall" <Scott.Hall@GD-CS.Com>
Subject: Re: IOLAN TS-Server to a Windows box????


Thomas wrote:
>
>         I need to connect a IOLAN Terminal Sever to a Windows 98 box  and
> print port configurations. What kind of cable do I need for connecting COM1
> to an > AUI (15 pin female) port


An AUI port is an Ethernet 4-pair non-transceiver port.  The COM1 port on
your PC is a asynch-serial port.  The two don't go together.

However, you can get an external transceiver adapter for the AUI port on
the terminal server (either 10-Base-T or 10-Base-2 (thinnet)).  Then
connect it to the ethernet network going to your PC.  Then with the right
Win-98 drivers, the terminal server will look like additional serial ports
to Win-98.

-- 
Scott G. Hall			General Dynamics Communication Systems
ph: 919-549-1189		North Carolina Systems Center
email: Scott.Hall@GD-CS.Com	Research Triangle Park, NC   USA


 //////////////////////////////////////////////////////////////////////////////

References: <3A59A119.E6CEAD46@nospam.kpnqwest.com>
Date: Mon, 08 Jan 2001 14:29:55 GMT
Organization: telocity
Newsgroups: comp.sys.sun.hardware
Message-ID: <D9k66.15660$Iu.2948071@newsrump.sjc.telocity.net>
From: Raf LaPietra <raf@u5.telocity.com>
Subject: Re: annex terminal server

In article <3A59A119.E6CEAD46@nospam.kpnqwest.com>,
Tom Lanigan  <Tom.Lanigan@kpnqwest.com> wrote:
>
>Can anyone help, how the hell do you set the default gateway address on
>one of these devices. Its driving me nuts as the CLI is not very helpful
>These are supplied by Sun a a remote management  solution for headless
>servers in a sun Rack. Cunningly we can no longer find the documentation
>for these so any pointers gratefully received.
>
>Thanx,
>Tom Lanigan

Try docs.sun.com :

http://docs.sun.com:80/ab2/coll.572.7/CLUSTNETHW/@Ab2PageView/920?DwebQuery=annex&oqt=annex&Ab2Lang=C&Ab2Enc=iso-8859-1

 //////////////////////////////////////////////////////////////////////////////


Newsgroups: comp.unix.solaris
Message-ID: <aJse7.4711$35.448710@iguano.antw.online.be>
Date: Wed, 15 Aug 2001 11:09:14 -0000
From: Karel Adams <k_adams@glo.be>
Subject: Re: HELP---BlackBox (iolan+) Terminal Servers connected to SUN


"Jim" <dogphace@alberta.com> schreef in bericht
news:gxme7.82993$sM.22063296@news2.rdc1.ab.home.com...
>
> Just wondering what anyones experience has been with BlackBox term servers
> and hooking them up to SUN's.  These term servers are a re-badge of the
> iolan+ term servers that Chase Research sold.  I guess they have been bought
> by Perle a while ago.
> Does anyone have a configuration that has been reasonably reliable for them?

(snip)

Chase have indeed been taken over by Perle maybe two or three years ago or so.

I have used iolan and iolan+ on other unixen and always found them very
reliable although support was never very impressive, Chase had trouble
to find and keep good people.

Can't say what the Perle support is worth. Can not help with an iolan
configuration as I have only used iolan to attach lots of serial
terminals and printers to a single server.

OTOH I have recently seen a setup where many Solaris machines had their
console ports attached to a Cisco router, I think it was a 2500. It
certainly is available with 8 or 16 serial ports and works very well.

Karel Adams.

 //////////////////////////////////////////////////////////////////////////////

Newsgroups: comp.terminals, comp.sys.dec
References: <b78217fa.0108301500.42d21dbf@posting.google.com>
Message-ID: <4KAj7.1034$bB1.45698@news.cpqcorp.net>
Organization: Compaq Computer Corporation
Date: Fri, 31 Aug 2001 00:19:12 GMT
From: Hoff Hoffman <hoffman@xdelta.zko.dec.nospam>
Subject: Re: terminal servers and serial devices

  Follow-ups have been set to comp.sys.dec.

In article <b78217fa.0108301500.42d21dbf@posting.google.com>,
    ant_selleck@hotmail.com (ant) writes:
:
:There are a number of products on the market which allow access to
:serial devices over a network. Essentially they seem to operate by
:making the devices "appear" to be local to a host, whilst in reality
:they are actually mapped to the physical ports of a remote terminal
:server.
..
:I understand that some early DEC terminal servers, operating under the
:LAT protocol had this type of functionality. Does anyone know if this
:was the case?

  Most (all?) DECserver terminal servers have these general abilities, with 
  the Local Area Transport (LAT) protocol being common to most of (all of?)
  these widgets and both LAT and IP terminal protocols common on the more 
  recent DECserver terminal server models.

  Remote displays via X Windows are also available -- the display is remote, 
  but appears to be a terminal that is locally connected to the OpenVMS host.

  For product information on current DECserver models, you may want to visit 
  the URL:

     http://www.dnpg.com/products/

  The earliest models of the DECserver series, LAT-11, and particularly the 
  earliest (the PDP-11-based Ethernet Terminal Server products) are at least 
  as old as 1985.
  

 ---------------------------- #include <rtfaq.h> -----------------------------
      For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    
 --------------------------- pure personal opinion ---------------------------
   Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com


 //////////////////////////////////////////////////////////////////////////////

Use Linux and TermPkg to make a PC into a serial terminal server:

    http://www.linuxlots.com/~termpkg/

 //////////////////////////////////////////////////////////////////////////////
 \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
 //////////////////////////////////////////////////////////////////////////////

