HP3000-L Archives

January 1995, Week 3

HP3000-L@RAVEN.UTC.EDU

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Eero Laurila <[log in to unmask]>
Reply To:
Eero Laurila <[log in to unmask]>
Date:
Thu, 19 Jan 1995 23:07:38 GMT
Content-Type:
text/plain
Parts/Attachments:
text/plain (43 lines)
Ross Scroggs ([log in to unmask]) wrote:
: Ok, we've all suggested the things to test, which don't seem to have worked.
: SOOO!! Break out the data scope/protocol analyzer, e.g., HP 4952. Plug it in
: at the DTC end of the cable and start the data flow with the original setup,
: not the new COBOL program. Two questions will be answered immediately.
: 1) Did the Intermec printer send a suitable XOFF?
: 2) Did the HP-3000 stop transmitting data until receiving an XON?
: If (1) is false, fix the printer! If (2) is false, call the HP RC!
: What makes question (1) true?
:   The Intermec must send an XOFF with hex value 13, not 93(=80+13, i.e.,
:   parity bit = 1).
: What makes question (2) true?
:   The HP must stop transmitting within one to three characters of receiving
:   the XOFF. The Intermec must allow these few characters of run-on from the
:   HP-3000. It's not viable to send an XOFF when your buffer is 100% full,
:   you send one when your buffer is x% full where x < 100. Verify with
:   Intermec what their XOFF strategy is.
 
This is exactly what I would expect an HP CE/RCE to do - hook up a protocol
analyzer and make sure the points 1) and 2) above;- was a good Xoff character
sent at appropriate time and did the DTC port stop from sending.
 
The way the DTC data flow works is that the host 3000 sends data in big lan
frames - say 1500 bytes each (if there's that much to send).  The DTC and
the host control the data flow through AFCP (Avesta Flow Control Protocol),
which typically allows for two(2) AFCP packets window.  Now, please don't
think it will take the time for the host to see the Xoff --.  No.
The DTC port will detect the Xoff and stop sending data;- the host may
still be sending to the DTC but that's where it stops.  This will cause
the DTC to hold on to the data and further on stop sending acknowledges/
opening up the AFCP packet window any further.  By the time the host has
sent it's current AFCP windows worth of data, it'll simply stop waiting
for more send window from the DTC.
 
As Ross recommended:
: If (1) is false, fix the printer! If (2) is false, call the HP RC!
 
Also, as someone recommended, you can use 2392A or some other terminal
in place of the printer to see that the DTC really stops sending on Xoff
<cntl-S> and resumes on Xon <cntl-q>.
 
Eero - HP CSY Networking lab.

ATOM RSS1 RSS2