Re: ANSI T1.231 ?

"C. M. Heard/VVNET, Inc." <heard@vvnet.com> Wed, 10 June 1998 21:41 UTC

Delivery-Date: Wed, 10 Jun 1998 17:41:35 -0400
Return-Path: heard@vvnet.com
Received: from cnri.reston.va.us (ns.cnri.reston.va.us [132.151.1.1]) by ietf.org (8.8.5/8.8.7a) with ESMTP id RAA27011 for <ietf-archive@ietf.org>; Wed, 10 Jun 1998 17:41:34 -0400 (EDT)
Received: from kickme.cisco.com (kickme.cisco.com [198.92.30.42]) by cnri.reston.va.us (8.8.5/8.8.7a) with ESMTP id RAA15367 for <ietf-archive@cnri.reston.va.us>; Wed, 10 Jun 1998 17:43:57 -0400 (EDT)
Received: from proxy2.cisco.com (proxy2.cisco.com [192.31.7.89]) by kickme.cisco.com (8.8.5-Cisco.2-SunOS.5.5.1.sun4/CISCO.GATE.1.1) with ESMTP id OAA09560 for <trunk-mib@external.cisco.com>; Wed, 10 Jun 1998 14:36:03 -0700 (PDT)
Received: (from smap@localhost) by proxy2.cisco.com (8.8.7/8.8.5) id OAA06010 for <trunk-mib@external.cisco.com>; Wed, 10 Jun 1998 14:36:01 -0700 (PDT)
Received: from shell16.ba.best.com(206.184.139.148) by proxy2.cisco.com via smap (V2.0) id xma005996; Wed, 10 Jun 98 21:35:57 GMT
X-SMAP-Received-From: outside
Received: from localhost (heard@localhost) by shell16.ba.best.com (8.8.8/8.8.BEST) with SMTP id OAA11405; Wed, 10 Jun 1998 14:32:02 -0700 (PDT)
X-Authentication-Warning: shell16.ba.best.com: heard owned process doing -bs
Date: Wed, 10 Jun 1998 14:32:01 -0700
From: "C. M. Heard/VVNET, Inc." <heard@vvnet.com>
X-Sender: heard@shell16.ba.best.com
To: atommib@bellcore.com, trunk-mib@external.cisco.com
Subject: Re: ANSI T1.231 ?
In-Reply-To: <3.0.32.19980610081901.009de9d0@shultz.argon.com>
Message-ID: <Pine.BSF.3.96.980610125046.22224A-100000@shell16.ba.best.com>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset="US-ASCII"

On Wed, 10 Jun 1998, Ken Chapman wrote:
[ ... ]
>>> What is meant by "digital signal level"?  DS1, DS3 and SONET?
>>> Or is SONET broken down by OC-1, OC-3, ..., etc.?
> 
>>It means anything that for which you have one or more
>>PerfCurrentCount objects -- DS1, DS3, Sonet STS-N
>>Section/Line/Path signals, various kinds of Sonet
>>VT Path signals (to the extent that they aren't already
>>covered), and their far-end counterparts, where applicable.
>>
>>That's a lot of thresholds!
>
> I'm afraid I didn't make myself clear.  I KNOW there are a lot of 
> threshold registers.  I'm trying to figure out how many DEFAULT
> threshold registers there needs to be.  My question was in the
> context to the particular sentance I quoted.

Ken,

I'm sorry, I'll try to make this reply a bit more helpful, if I can.

> So, let me ask it another way.  If I have a box with a DS1 port, a DS2
> port, an OC3 port and two (2) OC12 ports, I'll have five (5) separate 
> threshold registers for the current SES errors (one for each port).
> How many default threshold registers do I need for SES 15 minute interval
> errors to satisfy ANS1 T1.231-1997 claues 9.1.5.2, paragragh 2?
>
> a) two (one for DS1/DS2 and one of SONET)
> b) four (one for DS1, one for DS2, one for OC3 and one for OC12)
> c) some other set.

If I understand what clause 9.1.5.2, paragragh 2 says, then

c) some other set.  Exactly what this set is depends on how your
box can subdivide SONET payloads.

For the DS1 port, we have the following SES-related PerfCurrentCount
objects:

   - dsx1CurrentSESs
   - dsx1FarEndCurrentSESs

The thresholding requirements for DS1 are specified in clause 9.2.1.3
and Table 19 of T1.231-1997. Thresholding is always required for the
DS1 path near-end 15-minute SES counter;  and in applications
where far-end parameters are required, thresholding is required for
the far-end 15-minute SES counter also.  So the answer here is that
up to two default threshold objects are needed for the DS1 stuff.

T1.231 does not cover DS2, so the next part is a somewhat speculative
extrapolation.  I invite the trunk-mib folks to jump in and correct
this if they feel that this is wrong.  From a quick reading of
draft-ietf-trunkmib-ds1-mib-08.txt it appears to me that the same
two SES objects would apply.  However, a _different_ pair of defaults
would be needed for DS2, since it is a _different_ digital signal
level than DS1.  You would not, hoever, need a separate set of
defaults to cover any DS1 subchannels, since you already have that
set above.

The thresholding requirements for SONET specified in clause 9.2.3.3
of T1.231-1997 are that thresholding is required for all required
parameters and for any application-specific and optional parameters
which are implemented (in other words, for all PerfCurrentCount
objects which exist).  The relevant set of SES objects therefore is

   - sonetSectionCurrentSESs

   - sonetLineCurrentSESs
   - sonetFarEndLineCurrentSESs

   - sonetPathCurrentSESs
   - sonetFarEndPathCurrentSESs

   - sonetVTCurrentSESs
   - sonetFarEndVTCurrentSESs

Life starts to get quite complicated here, so let me suppose that
the SONET ports on your box are an OC3c and two OC12c's.  Then there
are no VT objects and only one set of path objects per port.  So you
need 5 default thresholds for the OC3 stuff and another 5 for the
OC12 stuff (one set of OC12c _defaults_ will suffice for both ports),
assuming that the far-end parameters are implemented.  Matters get
more complex if the SONET payload could be split up; you would need
a set of defaults for an STS-1 path, and for any VTs to the extent
that they were distinct from stuff that was already there (such as
the DS1's).

Summary:  if you support far end parameters, you will need to have
_at least_ 14 default thresholds for SES objects (2 for DS1, 2 for DS2,
5 for OC3, and 5 for OC12).  If you support only near-end parameters
then you need _at least_ 8 default thresholds (1 for DS1, 1 for DS2,
3 for OC3, and 5 for OC12).  More will be needed if your SONET ports
support non-concatenated payloads and/or VTs.

I hope this helps.

> 	Ken

Mike
--
C. M. Heard/VVNET, Inc.
heard@vvnet.com