RE: Consensus Check - SDLC Objects for RS-232 MIB

["Alan K. Bartky, Manager of Technology" <[e-mail]>]

I believe that you will find that in actuallity only the NRZ/NRZI parameter
and the port idle patter pertains directly to SDLC.

The other parameters are used by either
X.25/HDLC and other synchronous protocols.  Certainly most of the
parmeters are simple to implement as many implementations (HDLC ones)
can just default the values.  Here are some of the reasons we at Sync
Research think the parameters are important.


(These comments are based on Wayne's email message of December 22, 1993)

(1) Sync Port Role.

All of Sync Research's latest products have the ability (regardless
of running Sync or Async) to be software configured as DCE or DTE.
This field is of importance to us.  It is also certainly usefull
for even fixed devices for information purposes.  Devices that 
can't be reconfigured should certainly know and could report their
role as DCE or DTE.

(2) Encoding

Only SDLC (to my knowledge uses NRZI encoding), but this is an 
easy parameter for HDLC (only allow/report NRZ) and is required for
SDLC.

(3)RTS control and (4)delay.

In out products we support RTS switched control for both bisync, univac
and SDLC protocols.  Proper configuration of these parameters is 
critical to Synchronous port operation. (Yes we do support SNMP
for Bisync ports!)

(5) Mode of Operation

There are other protocols besides SDLC that are HDX protocols (again
we support Bisync with SNMP support).  Setting of this parameter
is crucial for SDLC.  HDLC, Frame Relay, etc. can set/report FDX.

(6) Port Idle pattern.

This is usefull and required for some SDLC devices.  It seems that
flags as an option was missed (i.e. options should be mark, space or
flags).  Again HDLC/Frame relay ports can easily default this parm
(to flags) and not make it setable.

(7) Minimum number of flags

This is usefull for determining problems.
Some devices do not support single flag separation (especially at high speeds).
If an attached device requires more
flags than 1, setting this object will allow it to operate (We have
a lot of field experience on Frame Relay, X.25/HDLC and SDLC that
have required us to insert more than one flag in order to get
an interface to work properly).  The description might need to be
rewritten to beter define it's need.  Devices that don't support
adding or controlling the number of flags between frames could hard
code it.

(8) Runt and (9) Giant Frames

Frames that are too small or large is a usefull concept for all
synchronous protocols!

Regards,
Alan K. Bartky
Manager of Technology
Sync Research Inc.
7 Studebaker
Irvine, CA. 92718-2013
[phone]
FAX: [phone]
email: [e-mail]








--- Begin Included Message (edited) ---

From POP2-Server@synclib Fri Mar  4 16:31:56 1994
    

Date: Thu, 3 Mar 94 10:25:51 -0500
Message-Id: <[e-mail]>
From: Bob Stewart <xap.xyplex.com![e-mail]>
To: xyplex.com!pa.dec.com![e-mail], xyplex.com!cisco.com![e-mail]
Subject: Consensus Check - SDLC Objects for RS-232 MIB
Content-Length:        503
Status:   

A few people spoke in favor of adding some SDLC-related objects to the RS-232
MIB.  They were reposted just recently by Wayne Clark, SNA SDLC MIB editor.
They could be added as a separate compliance group for SDLC, but I didn't do
so due to the general apathy.  The following test of consensus is intended to
insite response.  Depending on the response, I may try the other way around.

******** Consensus Test *********

Are there strong objections to leaving out the proposed SDLC objects?

	Bob
    
--- End Included Message ---