►
From YouTube: IETF105-CCAMP-20190725-1330
Description
CCAMP meeting session at IETF105
2019/07/25 1330
https://datatracker.ietf.org/meeting/105/proceedings/
A
B
A
A
A
Yeah
so
as
usual,
we
have
the
audio
recording.
So
please,
you
know
as
big
in
front
of
the
mic
and
state
your
name
before
speaking,
so
the
chairs
are
in
the
meeting,
echo
and
wave
your
monitor.
If
there
are
some
comments
from
there
are
remote
participants
and
for
the
minions
actually
are,
it
will
be
much
appreciated
if
anyone
could
take
some
minutes,
I
mean
key
points.
Also
they
use
a
pad
and
I,
because
you
know
our
circular
geo
sky.
It's
not
clear.
A
So
for
ip-ah
ip-ah,
you
know
I
it's
quite
very
important
and
we
have
to
follow
the
IPA
policy
on
United.
So
you
know
that
if
a
job
to
moving
to
the
next
step
in
you
know
process
we,
you
know
other
cheers.
We
are
send
out
IP,
upholding,
for
example,
before
an
individual
job
becomes
a
working
group
document
or
a
bottom
group
document
you
know
goes
to
last
call,
and
so
in
this
case
the
author's
or
the
contributors
you
know
are.
C
A
So
you
know
for
the
list.
As
always,
we
encourage
people
to
use
the
list
as
much
as
possible,
especially
we
encourage
people
on
to
discuss,
especially
the
technical
staff,
on
the
list
actively.
So,
for
example,
we
could,
you
know,
discuss
any
open
issues
on
the
list
or
even
you
know,
for
example,
I
introduced
some
new
jobs
or
you
know
some
potential
topics
on
the
list,
so
another
important
with
Munder.
It's
that
you
know
what
include
consensus.
It
is
determined
on
the
list
rather
than
you
know,
I
in
this
face-to-face
meeting.
A
E
An
update
on
the
status
of
the
working
group,
so
since
last
meeting
we
have
a
newer
of
C,
which
is
the
data
model
for
microwave
radio
links.
So
microwave
work
is
progressing
it
by
12.
We
have
now
another
C
for
the
framework
and
RFC
for
radio,
a
big
young
model
of
the
radio
link,
and
we
are
now
in
order
to
conclude
this
work.
We
are
still
missing,
adjust
the
topology
document,
which
is,
by
the
way,
progressing
quite
well.
We
also
have
two
documents
in
the
editor
tube
DLR
medulla
and
the
RSVP
T
bandwidth
availability
I.
E
E
Working
group
documents
adopted
the
since
last
meeting
up
a
lot
of
lemma,
so
we
have
been.
You
cannot
complain.
Your
chairs
are
lazy
because
we
have
six
new
working
group
document
since
last
last
meeting,
mostly
we
have
the
year
0
layer,
1
types
we
agreed
to
follow
the
way
of
working
of
Tisa.
Well,
the
traffic
engineering
types
have
been
a
decouple
that,
from
from
the
topology
and
a
dedicated
document
has
been
created
for
them.
E
These
were
these
are
the
ones
that
will
be
discussed
that
day,
for
which,
as
lot
has
been,
has
not
be
has
been
requested.
Well,
these
are
the
ones
that
will
not
be
presented
play
but
for
which
we
requested
an
update
from
from
the
others,
so
I'll
quickly
go
through
them
if
the
others
want
to
chime
in
and
provide
this
further
info.
This
is
more
than
welcome
mostly
for
these
two,
because
we
don't
have
another
date.
E
H
G
E
These
one
has
been
updated
that
reflect
the
changes
that
we
were
speaking
about
before
the
layer,
0
layer,
1
types,
layer
0
in
particular
in
this
case,
and
some
changes
regarding
the
security
section
and
updates
to
to
the
references
flexibly.
The
young
model.
Again
also,
this
one
was
updated
to
reflect
the.
E
G
E
E
I
I
E
Balaji,
can
you
send
me
a
reminder
I'll
do
that
thanks,
microwave
topology,
younger
I
can
also,
in
this
case
a
pretty
minor
a
pretty
minor
changed
changes.
Moving
the
microwave
bandwidth
from
T
link
to
T
bandwidth,
introduce
the
introduce
that
the
microwave
channel
list
I
removed
the
tunnel
concept
next
step.
Still
some
work
to
be
done,
but
I
would
say
marijuana
so
clarify
the
applicability
of
the
T
topology
on
a
micro
network,
container
definition
for
microwave
in
the
young
model
and
complete
use
case
details,
no
change
on
the
NBI
applicability
statement.
E
J
Yes,
we
are
doing
a
detailed
review
within
the
design
team.
We've
got
many
useful
comments
from
little
and
Michael
and
we
should
be
able
to
arrest
a
vacuum
update
by
September
August,
and
then
we
will
be
ready
to
provoke
a
new
Prescott.
The
only
pending
issue
is
a
normative
references
that
are
most
of
them
working
group
documents
and
one
is
individual
document
which
are
normally
work.
So
that's.
E
E
Discussion
before
this
sample,
we
needed
to
sort
out
I
mean
if
those
are
not
much
of
reference
is
there's
nothing
that
we
can
do.
The
only
thing
I
would
say
we
will
go
for
the
last
call
when,
when
the
document
is
ready,
there's
no
need
to
keep
it
part
until
all
the
references
will
become
will
be
published.
Probably
it
will
be
hold
by
the
RFC
editor
after
after
that,
but
at
least
we
will
write
everything
in
stones.
E
Basically,
the
clarification
are
related
to
the
scope
of
the
document
and
the
fact
that
it
only
covers
the
transparent
dozen
years
anti
communication.
As
usual,
pretty
pretty
intense
activity
to
see
camp
and
the
from
cecum
received
a
liaison
from
itu-t
study
group
of
15,
whose
title
is
related
topic
being
a
description
of
the
OTS
I
and
the
network
media
channel.
This
is
probably
what
we
will
see
in
our
first
presentation
from
deter
and
we
sent
a
reagent
to
any
effort
on
providing
an
update
on
our
work
on
l1
service
model.
I
I
As
you
can
see
on
this
slide,
there
are
quite
a
number
of
people
who
are
actually
contributing
to
this
effort.
So
it's
a
long
list
and
we
are
having
weekly
meetings,
open
meetings
that
have
been
announced
on
the
cecum
mailing
list,
given
the
political
circumstances
which
we
have
right
now,
so
we
have
a
call,
a
weekly
call,
every
Thursday
for
one
hour
and
as
you
can
see,
we
have
a
great
support
from
operators,
as
well
as
from
a
variety
of
different
vendors,
including
yep,
and
we
also
have
a
github.
I
The
scope
of
the
document
is
actually
to
provide
a
young
data
model
that
actually
collects
information
from
the
network
from
the
devices
and
constructs
a
topological
view
of
the
network,
based
on
the
information
that
it
receives
from
the
devices
and
provides
a
network-wide
view
of
the
network
topology
from
layer
0.
For
a
WDM
network
to
be
more
precise,
we
have-
and
this
is
the
picture-
that's
in
the
document
you
can
see
here.
Just
you
can
see.
Here's
the
devices
in
the
network,
for
example
the
optical
transponders,
and
also
the
optical
line
system
segment.
I
It's
important
to
note
that
the
the
information
is
a
read-only
information,
because
the
information
is
just
collected
from
the
network,
we're
not
intending
to
use
this
model
to
control
the
network.
So-
and
you
can
see
here,
this-
is
this-
the
the
interface
that
were
targeting
where
the
topology
information
can
be
retrieved,
and
we
have
four
device
configuration
a
separate
interface
and
there
is
a
separate
draft
that
a
young
model
for
the
interface
configuration
that's
actually
complementing
this
topology
yang
model
for
configuring.
I
The
network
elements,
as
already
mentioned,
we
have
splits
off
or
end
off
the
type
definitions
that
we
need
for
this
later
is
your
topology
and
all
the
groupings
that
we
have
defined
that
are
used
to
describe
the
layer,
zero,
topology
and
the
optical
impairments.
So
that's
in
a
separate
di
module
and
that
layer,
zero
types
yang
file
that
should
be
imported
not
only
by
the
topology
yang
model,
but
also
by
the
interface
yang
model,
so
we're
trying
to
get
everything
pretty
much
aligned
to
make
sure
that.
I
And
this
is
the
reference
architecture,
also
an
ASCII
art
picture
from
the
document
you
can
see
here.
We
have,
for
example,
the
Rotom
device,
the
rodent
devices
can
be
decomposed
into
an
ingress
or
amplifier.
It
has
a
switching
fabric,
a
photonic
switching
fabric
to
switch
optical
channels
or
media
channels
from
one
side
to
the
other
side,
and
it
also
has
an
add/drop
lock
where
you
connect
the
optical
transponders,
where
signals
photonic
signals
can
be
terminated
or
generated
and,
of
course,
between
the
rodents.
I
We
have
the
OTS
link
and
the
OMS
link,
as
defined
by
the
itu-t
recommendations,
and
this
we
have
currently
a
very
fairly
detailed
model
of
the
OMS
link.
That
is
a
series
of
elements
like
fibers
light
in-line
amplifiers.
We
even
have
introduced
four
connectors,
an
insertion
loss
that
can
be
configured
or
that
can
be
retrieved
from
the
network.
So
we
have
a
fairly
detailed
model
already
for
the
OMS
link.
I
What
is
still
missing
and
what
we're
still
working
on
is
to
define
a
model
for
the
optical
transponders,
and
we
also
figure
out
that
we
need
to
define
a
model
for
three
our
generators.
So
that's
still
something.
That's
outstanding
and
we
have
created
issues
on
our
github
for
that
to
address
those
issues
and
refine
the
modeler
to
also
include
elements
for
three.
Our
generators,
as
well
as
optical
impairments
for
the
rodents.
I
Okay,
major
changes
since
the
previous
IETF
on
a
104
meeting
yeah.
As
you
have
seen,
there
are
a
lot
of
contributors
or
co-authors.
We
had
a
lot
of
long
list
of
co-authors
on
the
title
page
and
of
course
this
does
not
meet
the
rules
of
the
IETF,
so
we
had
to
select
a
few
co-authors
that
we
put
on
the
title
page
and
then
we
moved
all
the
other
contributors
further
down
in
the
document.
We
also
updated
the
reference
documents.
I
We
replaced
the
figures
that
we
have
used
in
our
previous
discussions,
which
were
PowerPoint,
slides
with
ASCII
art
figures
so
and
we
had
a
we
added.
Also,
then
the
definition
of
what
we
call
here,
a
media
channel
and
a
media
channel
group.
The
media
channel
is
also
defined
in
the
new
IQ
t
or
accomodation
G,
807
and
I'll.
Come
to
that
in
a
minute.
The
media
channel
group
is
something
that
we
define
in
order
to
group
multiple
media
channels
together
in
case
inverse
multiplexing
issues.
That
means
it's.
I
For
example,
if
a
400
gig
client
signal
has
to
be
carried
using
multiple
carriers
that
carries
a
fraction
of
the
client
signal
like,
for
example,
4
times
100
gig,
then
it's
of
course,
the
objective
to
put
all
the
four
carriers
into
a
single
media
channel.
If
spectrum
availability
doesn't
allow
that
end-to-end
and
of
course
we
have
to
split
the
other
signals
and
carry
them
across
multiple
media
channels
and
therefore
we
have
created
this
media
channel
group
concept.
I
We
also
had
quite
some
discussion
regarding
the
interpretation
of
the
OTS,
a
definition
that
is
defined
in
the
ITT
recommendation.
G95
9.1
and
the
OTS
I
signal
OTS
I
group,
which
is
a
grouping
of
OTS
I
signal,
that's
defined
in
the
new
recommendation,
G
807
and
of
course,
we
are
currently
reviewing.
Also
all
the
new
parameters
that
we
have
added
and
for
the
error
element
for
the
optical
transponders,
as
well
as
optical
and
parent
parameters,
on
the
line,
as
well
as
for
the
for
the
road
and
says
something
that
we
still
have
to
work
on.
I
So,
based
on
the
side
meeting
that
we
had
back
in
Prague,
we
had
a
lock
a
long
discussion
how
to
interpret
the
definitions
in
the
itu-t
recommendations
that
define
the
data
plane.
Specifically
in
itu-t
recommendation
nine
five-niner-one.
It
says
that
the
OTS
I
can
be
a
signal
which
can
be
composed
of
a
single
optical
module.
A
single
modulated
optical
carrier
can
be
a
signal
of
multiple
modulated
optical
carriers,
and
so
we
had.
I
We
were
struggling
with
that
definition
and
we
actually
put
together
a
contribution
to
the
study
group
50
meeting
that
was
held
beginning
of
July
in
Geneva,
where
we
actually
said
well,
we
are
assuming
for
our
topology
yang
model
data.
No
TSI
only
consists
of
a
single
modulated
optical
carrier,
and
we
proposed
to
actually
clarify
that
definition
in
95
9.1
to
only
cover
that
case
and
to
remove
that
that
it
may
also
contain
multiple
optical
carriers.
I
However,
we
had
also
during
the
discussion
and
in
the
meeting
that
we
had
yesterday.
We
still
figure
out
that
there
is
still
some
clarification
required.
The
G
807
is
a
new
recommendation
that's
currently
under
or
for
which
the
AAP
a
last
call
process
will
be
starts
started
soon,
and
we
agreed
that
we
will
have
to
prepare
some
lascall
comments
on
that
document
and
we
will
coronate
that
also
in
the
upcoming
meetings
of
this
or
the
group
of
people
that
are
working
together
on
this
draft.
To
submit
these
comments
and
to
study
could
15.
I
I
I
Okay,
based
on
the
feedback
that
we
got
in
that
liaison
letter
that
has
mentioned
a
few
minutes
ago,
I
think
our
model
is
still
valid
and
I
think
we
can.
Actually.
We
believe
that
we
can
move
forward
with
what
our
assumptions
are.
So
we
believe
that
the
current
yang
model
is
aligned
with
G
807
and
are
the
assumptions
that
we
have
taken
in
our
drives.
I
Is
that,
as
I
said,
that
the
no
TSI
only
has
a
single
modulated
optical
carrier-
and
we
also
allow
the
freedom
that
the
frequency
of
that
optical
carrier
may
not
or
beyond
the
ITT
grid
that
allows
us
actually
to
squeeze
multiple
carriers
together
into
a
media
channel
to
optimize
spectrum
utilization
and
a
media
channel
can
contain
multiple
OTS
I.
So
for
the
purpose
of
optimizing,
this
spectrum
in
case
inverse
multiplexing
is
used,
and
all
these
arm,
ltsi
that
are
belonging
to
a
single
out
TSI
group
have
to
be
corroded.
I
So,
therefore,
the
requirement
from
our
side
is
that
network
has
to
provide
a
pipe
and
to
end
from
A
to
Z
crossing
multiple
rodents
over
the
media.
Channel
of
the
same
sizes
is
switched
in
the
switching
fabric
or
in
the
WSS
switching
matrix
and
such
that
we
get
an
end-to-end
pipe
for
that
contains
all
the
OTS
is
belonging
to
the
same
virtuous
I
group.
So
that's
a
type
of
the
same
size
from
A
to
Z.
That
carries
all
these
signals.
I
Okay,
as
I
said,
we
still
have
to
work
together
with
the
itu-t
study
group
15,
to
clarify
these
issues
and,
as
I
said,
we
will.
We
are
going
to
coordinate
and
send
some
last
aap
last
call
comments
on
8
to
7,
and
this
is
the
yang
at
the
preview
of
our
apology
yang
a
model
we
have
currently
on
our
github.
I
It
augments
the
tea
topology
yang
model.
So
it's
augmenting,
for
example,
the
te
link
attributes
here
are
a
number
of
attributes
that
are
listed
so
for
the
OMS
line.
For
example,
we
defined
generalized
way
there
are
some
power
parameters,
then
the
the
media
channel
group
and
the
media
channels
inside
the
media
channel
group.
I
I
Yeah
we
are,
we
have
it
github
we're
using
github
to
actually
develop
our
yang
model,
our
game
file,
and
there
are
a
number
of
issues
there
and
everybody
has
access
to
this
github
and
can
actually
create
issues
or
submit
pull
requests
in
case
he's
proposing
some
changes
which
we
are
then
reviewing
together
and
then
we're.
If
the
changes
are
okay,
we're
accepting
that
them
and
then
move
on.
Please
have
a
look.
This
is
the
link
to
the
github,
it's
also
I,
think
available
on
data
tracker
or
our
layer
0
at
apology,
young
model.
I
I
The
most
important
issues
that
we're
currently
addressing
is
we
need
to
define
what
the
OT
SI
capabilities
or
the
transponder
camp
capabilities
are
so
all
the
operational
modes
and
the
configuration
that
possibilities,
the
optical
transport,
provides,
have
to
be
defined
and
have
to
be
retrieved
in
order
to
support
path
computation
in
a
on
the
topology,
because
when
you
create
a
new
service
from
A
to
Z,
of
course,
the
Tewa
transformers.
At
the
end,
the
list
of
the
photonic
connection
have
to
share,
have
some
capabilities
in
common.
Otherwise
the
traffic
or
the
connection
doesn't
work.
I
So
in
order
to
make
sure
that
we
have
to
list
all
the
capabilities
that
the
two
end
transformers
have,
and
also
the
optical
impairment
parameters
determine,
to
some
extent,
also
the
the
length
of
the
connection
or
how
many
ops
you
can
actually
cross,
and
those
also
depend
on
the
operation
mode.
So
that's
important
information
that
has
to
be
provided
before
the
service
is
actually
established
and
then,
of
course,
the
optical
transponder
has
one
configuration
based
on
its
capabilities.
I
We
still
have
to
define
what
optical
impairments
should
be
provided
in
the
model
for
rodents.
So
when
an
optical
signal
traverses
a
we
have
to
define
what
optical
impairments
in
terms
of
crosstalk,
for
example,
or
other
impairments,
so
those
are
still
missing,
and
the
last
or
one
of
the
other
important
open
issues
is
the
definition
of
three
origin
or
ages.
We
had
some
discussion
yesterday.
Already
there
are
different
types
of
three
original
raters,
uni-directional,
regenerators
or
b
direction.
I
Are
we
generators,
so
these
kind
of
devices
that
are
used
in
today's
network
have
to
be
supported
by
the
model
next
steps?
Of
course,
we
will
continue
our
discussion
or
interactions
with
the
iqt,
as
I
said,
we
will
have
to
keep
the
alignment
of
the
topology
model
with
the
interface
model.
That's
going
to
be
presented
in
a
few
minutes
by
a
Gabriela
and,
of
course,
we'll
have
to
add,
then
also
the
model
for
three
original
raters
and
manage
the
open
issues
on
our
github.
So
that's
pretty
much
it.
Regarding
this
draft
questions.
G
Actually,
Gabriela
speaking
sure
is
not
a
question
but
is
foundation
as
when
we
introduce,
for
example,
to
our
generator,
also
the
layer,
layer
layer,
zero
type
draft
should
be
updated
accordingly
and
then
okay,
we
are
working
with
your
importer
of
the
drafter
in
order
to
make
a
or
keep
this
that
draft
you
see
with
our
okay.
I
I
think
that's
that's
our
general
strategy
that
we
have
one
young
file
that
contains
the
type
definitions
for
Holly
zero
work,
including
the
topology
model,
as
well
as
the
interface
model.
Yes,
Tara,
a
single
source.
That's
import,
then,
into
the
yang
models
for
for
the
interface
as
well
for
topology,
and
maybe
there
are
others
just.
A
I
A
So
on
here
regarding
the
assumptions,
I
think
actually
are
dis.
Reham
assumptions
are
correct,
but
you
know
are
based
on
the
itu-t
recommendations.
It
says
that
OTS
I
may
consist
of
one
thing:
go
our
carrier
or
a
group
of
carriers
or
sub
carriers.
So
for
the
first
item
is
cracked,
but
I'm
not
sure
on.
My
question
is
that
if
your
model
can
support
OTS
I
consists
of
you
know
a
group
of
carriers
or
sub
carriers.
I
Yeah,
we
deliberately
did
not
want
to
have
that,
because
we
believe
that
it's
sufficient
to
have
an
OTA
Seiichi
that
can
contain
multiple
OTS
eyes
so
and
to
have
a
second
level
of
inverse
multiplexing,
and
that
was
the
message
of
the
the
proposal
or
the
justification
in
the
itu-t
kharak
contribution.
We
do
believe
that
this
over
complicates
the
management
of
the
WDM
network.
A
Yeah
I
know
you,
you
know,
purpose
of
you
and
you
know
a
contribution
to
ITT,
but
you
know
this
contribution
of
lopata
was
protected.
You
know,
I
hate
UT
I
mean
so
because
we
have
actually
we
don't
define
data
slam
technologists,
usually
from
the
conscious
land
perspective.
We
should
have
followed
the
ITT.
You
know
recommendations
so
on.
I
Optical
transponders
that
are
available
today,
I
think
all
the
vendors
are
actually
using
a
single
optelec
optical
carrier.
That's
modulated
to
carry
digital
information,
so
the
in
the
liaison
statement,
and
also
in
the
information
that
I
received
from
the
ITT
meeting,
they
said
in
the
future.
There
might
be
other
modulation
techniques
that
may
use
subcarriers,
but
today,
I
think
such
modulation
schemes
do
not
exist
in
in
in
real
networks.
So
therefore,
I
think
we're
okay
for
now
and
in
the
future,
if
required,
we
can,
of
course,
six
and
remodel.
I
M
E
E
E
E
Probably
be
careful
when
writing
in
the
draft,
something
like
don't
write,
we
assume
that
only
a
single
carrier
can
be
part
of
any
TSI
but
say
something
like
not.
Yes,
I
can
be
composed
by
single
of
multiple
carriers.
Within
the
scope
of
this
document
we
are
assuming
we
are
considering
just
the
scenario
of
single
carrier.
G
Okay,
so
this
draft
is
actually
complimenting
the
previous
draft
described
by
editor
and
is
working
more
at
the
lower
level,
so
represent
the
optical
impairment
at
the
interface
level.
So,
while
the
previous
draft
is
augmenting,
the
the
topology
disrupt
or
the
models
described
in
this
draft,
augmented
the
interface,
the
idea
of
interface
models,
so
that
the
scope
is
to
give
the
possibility
or
expose
all
the
optical
impairments
motors
and
the
the
way
that
in
interface,
DWM
interface
works
to
the
to
the
controller.
G
Document
history
is
just
quite
a
new
working
document.
He
has
been
promoted
to
two
zero
zero
ration
in
March,
so
we
didn't
have
any
presentation
in
Dragon
AIT
f-104
and
has
been
updated
in
July.
What
has
been
changed
is
the
regime,
some
parameters
that
are
reflecting
better,
the
the
meaning
of
the
signal
they
represent.
We
added
actually
two
new
parameters.
One
is
the
BOD
rate
and
the
second
one
is
the
total
input
power
received
at
transceiver.
Then
we
fix
some
typos,
and
so
so.
The
next
step
is
ok.
G
Realign
the
document
based
on
the
latest
itu-t
definition
and
models
work
very
closed
to
keep
the
draft
in
sync
with
the
optical
impairment.
Apology,
young,
modest,
either
in
terms
of
terminology
and
in
terms
of
contents,
so
I
remind
that
these
two
documents
are
complementing
each
other,
so
it's
important
that
they
have
the
same
terminology
and
they
don't
present
any
parameter
duplication,
because
this
would
be
misleading
and
so
on.
G
L
Let's
get
them
all.
I
was
a
holder
of
this
job
that
has
to
clarify
so
in
this
interface
model,
there
are
also
possibility
to
define
different
carriers,
which
is
to
address
your
your
concern.
We
still
need
to
work
out
how,
let's
it
bridge
those
two
drafts
together,
but
at
least
from
that
perspective,
we're
okay,
Thanks.
Thank.
G
I
C
I
N
I
F
Okay,
this
is
a
hobby,
come
here
from
Harley
and
welcome
to
lair
what
the
first
draft
is
going
to
be
present.
Today,
it's
layer,
one
types
which
is
a
new
draft,
but
this
is
a
grading,
the
track
meeting
to
have
a
common
session
between
the
among
a
few
draft,
including
the
odeon,
topology
or
Tintin.
Oh,
and
also
they
layer,
one
service
model.
So
we
generate
common
types
and
groupings
and
put
a
model
in
this
draft.
F
So,
firstly,
let's
look
at
what
is
proposed.
The
model
relationship
and
the
motivation
is
to
okay
use
the
same
philosophy
with
what
has
been
done:
40
generic
model
and
extract
the
technology
specific
model,
the
types
with
the
common
types
between
multiple
models.
So
here
we
just
focus
on
layer,
1
and
more
explicitly
odeon
common
types,
so
this
model
would
be
imported
by
three
different
modules,
including
the
Audion
topology,
OTN
danau
and
the
layer.
1
active
the
service
model
and
this
module
itself
does
not
require
I
had
any
importing
from
other
models.
F
So
this
can
be
used
as
a
fundamental
one,
and
in
this
table
we
just
enumerate
what
has
been
included
in
the
current
layer.
One
types,
so
there
are
two
groups
of
basic
type
of
the
first
way
is
OD,
you
type
and
the
second.
Why
is
client
signal?
So?
Basically,
audio
type
is
used
as
a
switching
type
II
and
we
are
using.
F
We
have
specified
all
Audio
types
list
in
RFC,
scented
ones,
39,
which
is
person
I
to
teach
it
awesome
all
died
and
also
their
RFC
79
63,
which
is
from
the
treatise
at
sacrament
43,
and
for
the
client
signal.
We
have
different
granularities
for
Ethernet,
STM,
n,
OC,
fibre,
channel
and
cycle,
so
we
have
reference
to
the
cross
bonding
data
playing
standard
as
well,
and
besides,
the
types
were
also
having
similar
groupings,
which
would
be
imported
by
multiple
models,
including
basically,
besides
audio
type
and
the
client
signal.
F
So
there
is
consecutive
effort
to
harmonize
also
there
one
models,
especially
for
the
OTO
one,
and
make
it
a
consistent
between
OTN
topology
tango
and
there
one
connectivity
service
model,
and
there
is
also
the
open
issue
to
check
with
our
FCAT
407,
which
is
a
guideline
of
the
young
model,
to
avoid
annoying
mr.
patch
and
after
we
do
that
we
will
repress
the
further
working
class
go
and
the
end
after
review.
E
Just
comment
from
a
problem,
let
me
say
process
point
of
view.
I
would
say
that
that
now
that
the
flow
to
be
followed
in
progress
in
the
work
should
be
types
topology
and
like
like
the
divinities.
So
yes,
this
will
be
the
next
billion
zero
types
and
the
layer.
One
types
will
be
the
next
in
the
Python
to
be
moved
forward
right.
F
Yep
just
so,
the
slice
is
also
following
this
order.
I
believe
is
attract.
So
after
the
types
we
are
going
to
present
to
the
audience
Apollo
young,
which
has
been
silent
for
a
while
and
last
time,
is
mantra.
Okay,
him.
So
so,
let's
reveal
the
summary
of
the
changes.
This
is
now
it's
portion
number
seven,
but
what
we
compared
a
is
Washington
bursaries,
which
is
our
last
presentation
in
Montreal
as
well.
So
for
this
one
we
have
chaff
the
compliance.
F
Firstly,
we
added
the
supported
Aquinas
signals
to
indicate
the
capabilities
of
the
OT
node,
and
we
also
harmonized
with
the
latest.
He
generic
model,
which
is
now
in
the
RFC,
editor
queue,
and
basically
there
are
two
basic
classes.
The
first
one
is
to
add
the
labels,
tab
information
and
for
out
here
in
so
since
it's
it's
digital
once
and
it's
just
always
set
one,
and
we
also
change
the
past
correspondent
pass.
F
Specific
characteristics
so,
basically
for
the
attributes,
we
extend
the
link
and
did
be
a
little
bit
and
the
first
band
ways
we
use
audio
type
and
the
numbers
to
indicate
the
the
each
half
see
of
you
in
as
a
list
and
a
few
parameters,
including
TPN
tributary,
slot
tubular
slot
granularity
and
list
added
into
the
teal
able
to
indicate
the
label
start
and
the
heart
and
step
restrictions.
Then
so
this
worker
is
taking
this
table
and
we
also
request
for
the
last
coin,
the
end
after
review
after
the
previous
one.
F
E
Left
one
to
the
young
daughters
are
pretty
much
overloaded
with
reviews
etc.
Isn't
it
the
tea
types,
something
quite
simple
that
we
can
skip
at
a
young
daughter
view?
Do
you
feel
comfortable
that
would
progress
said
that
at
the
drafter,
without
a
young
method
of
you
for
such
a
simple
such
a
simple
draft.
O
E
E
J
P
N
For
future
way,
so
we
went
through
this
process
with
a
louder.
Sorry,
we
went
through
this
process
many
times
already
and
I
think
I
recommend,
if
it's
possible
and
then
to
do
it
them
to
do
it
as
early
as
you
can.
Okay,
there
is
a
lot
of
good
feedback
comes
from
the
doctors,
okay
with
respect
to
readability,
and
so
with
respect
to
the
gravity
and
correctness
of
the
definitions.
N
F
O
F
So,
okay,
so
in
the
next
one
is
next
draft.
It's
a
young
model
for
ot
intono,
which
is
also
related
to
with
the
previous
two
and
the
mainly
impulse
is
a
layer
one
types,
so
we
have
also
did
some
changer
to
the
text
and
the
model,
so
taxi
is
basically
form
a
checking
and
make
every
section
complete
and
the
compliance
with
RFC
eighty
407
and
the
young
model
Chandra.
We
also
harmonized
with
the
T
generic
model.
The
same
is
too
similar
as
Audient
apology.
F
Similarly,
we
augment
ins
a
tea
tunnel
model
in
a
few
different
segments.
The
first
plan
is
attributes.
We
augment
itano's
the
tea
bandwidth.
We
also
indicate
audio
types
for
for
teen
to
represent
to
the
bandwidth
information
and
again
TPN
tstst
t
s
last
range
types
as
that
kind
of
seeing
Santa
representing
the
Audion
labels.
So
basically,
this
is
a
consistent
ways
what
we
have
been
doing
in
the
Audion
topology,
so
the
next
step.
This
work
is
also
technically
technically
stable.
E
J
So
what
is
the
problem
statement?
What
is
the
problem
we
are
trying
to
address
with
this?
Rafter
is
tooth
problems.
The
first
problem
is
to
describe
by
the
ethernet
assess
link.
So
what
do
we
have
defined
up
to
now?
Is
we
have
a
key
topology
and
t
tunnel
models
which
are
low
to
set
up
a
tunnel
in
the
network
for
any
type
of
server
T
top
network
like
o
TN
and
after
we
create
this
o
TN
tunnel?
For
example,
we
can
map
thanks
to
the
Ethernet
line
senior.
J
We
can
describe
how
the
ethernet
traffic
from
the
access
link
can
be
mapped
and
transported
over
the
T
tunnel.
What
is
missing
is
what
addresses
links.
How
to
represent
is
assess
leaks
in
terms
of
apology,
how
to
describe
the
interlayer
relationship
between
the
businesses
link
and
a
seven-layer
TTP,
because
if
I
want
to
create
a
tunnel
to
carry
traffic
from
Atlantis
from
a
given
access
link,
there
should
be
a
back
another
tissue
cavity
between
this
SS
link
and
the
TTP.
J
Otherwise,
e
the
traffic
from
the
client
cannot
assess
the
tunnel
and
what
are
the
capabilities
from
the
access
link
in
terms
of
villain
classification
and
operations
which
have
to
be
configured
during
the
in
decline,
signal
model
and
one
requirement?
We
have,
for
example,
in
the
F
which
people
wants
to
configure
per
you
and
I
Bennett
profiles,
so
how
you
can
configure
and
a
vendor
profile
on
an
ass's
link
and
the
solution
is
to
have
a
client
apology
for
Ethernet
the
where
we
report
the
access
links
and
the
inter
layer
relationship.
J
Is
yet
another
technology
specific
commentation
of
the
generic
T
topology
model
when
the
switching
technology
is
a
layer
2
as
switching
capability,
and
this
type
of
switching
technology
is
a
generic
T
can
be
also
used
to
support
the
ethernet
client
seniors
as
before,
and
that's
the
eighth
another
technology
that
we
can
use.
We
can
represent
the
same
topology,
both
the
server
T
and
a
client
topology.
Next,
please.
J
That
we
did
is
basically
to
create
this
new
one
topology,
a
model
which
provides
both
T
and
client
descriptions
we
as
an
augmentation
of
the
generic
T
topology
next
and
that's
okay,
we
can
go
script
quickly
here,
it's
the
generate
the
augmentation.
So
what
is
the
next
steps?
Okay,
this
approach
is
quite
consistent
with
the
existing
at
EF
modus
80
is
just
an
augmentation
of
the
topology,
where
the
client
is
basically
providing
the
capabilities
that
are
used
by
the
client
signal,
which
has
been
recently
adopted
as
a
working
document.
J
We
are
relying
for
problem
statement
under
1
on
this
document,
also
in
the
transport
MPI
design
team
to
represent
the
SS
links
to
carry
services
over
at
the
end.
So
we
think
we
can
start.
We
can
move
this
worker
to
the
working
group
and
asked
for
a
pole
and
his
work
is
also
available
and
we
use
the
same.
J
A
young
model
as
before,
with
Toro
Tien
and
other
seniors,
and
one
that
that
we
got
recently
is
whether
we
need
packet,
t
types
and
packet
e
topology,
as
an
impact
on
the
concept
on
the
structure
of
the
document.
I
think
this
can
be
addressed
by
the
working
group
and
we
need
to
unseat
discussing
with
the
authors
of
those
drafts,
because
not
very
clever
that
they
are
generic
or
at
least
T
specific,
these
two
documents.
E
J
The
reason
for
TS
that
there
are
some
attributes
in
the
T
topology,
which
can
be
useful
for
us,
for
example
the
interlayer
lock,
which
allows
you
to
link
the
client
apology
with
a
server
topology.
You
have
also
information
about
the
client
ID
server
ID,
which
is
a
spoke
whose
exposing
this
topology,
and
also
you
have
the
Plaga
D
that
you
can
use.
If
you
want
to
stitch
at
us
as
links
topology
between
two
different
domains,
these
are
I,
think
them
three
major
reasons
and
make
sense.
Thank
you.
J
Q
Q
Q
J
Q
F
Q
J
Q
Q
J
F
J
Q
J
Is
not
so
it's
this
ever
you
have
the
tunnel
model
the
teeth
on
the
model.
We
can
be
OTN.
What
is
by
the
Italian
motor
stops
the
tunnel
from
this
point
to
this
point
now,
what
you
have
to
say
is:
you
have
to
say
which
traffic
goes
here
and
that's
the
climb
senior
the
current
senior
say:
buy
the
currency,
a
reference
to
a
nod
to
a
link.
What.
Q
J
N
Qualification,
so
basically
talking
about
where
to
source
per
se
were
talking
about
the
investment
of
their
to
service
right.
So,
for
example,
if
you
have
there
are
two
VPN
vision
which
has
some
kind
of
T
constraints
or
optimization.
Then
you
will
do
that
yeah,
okay,
so
it's
not
about
rich
purity
per
se,
its
it's
about
traffic
engineering
aspect,
especially
in
this
verse.
In
the
second
thing
you
can
think
about.
This
is
basically
a
one
hop
infinitely
which
is
basically
supported
by
and
to
end
the
tunnel
in
the
solar,
well
yeah
right.
N
J
Oh,
no,
no!
Okay!
There
are
two
cases,
that's
why
we
have
the
second
problem
statement.
If
you
would,
if
you
cannot
do
an
end-to-end,
OTN
tunnel,
then
you
ever
know,
then
you
have
a
basically
in
principle
to
type
of
options.
One
option
is
that
the
end-to-end
tunnel
is
an
eternity
tunnel
and
then
you
have
then
that's
why
you
have
a
second
issue
and
there
you
have
to
manage
an
eternity
topology
and
an
eternity
LSP.
So
what
you
do
you
you
created
a
to
attend
attack
or
detergent.
J
Analyst
becomes
three
links
in
the
Eternity
topology
and
then
you
set
up
an
iterative
tunnel,
but
you
still
have
to
steer
the
traffic
from
the
access
link
into
this
eternity
tangle.
So
so
that
and
that's
I
said
this
document
addresses
both
you.
You
can
expose
an
eternity
topology
which
can
be
based
on
top
of
a
new
TN
topology,
as
well
as
a
multi-layer
topology
or
as
you
can
deploy
it
as
adjust
Ethernet
switches
which
support
eternity
LSP.
J
J
N
You
are
saying
I
understand
is
that
you,
in
most
of
the
cases
you
have
a
single
say
and
wind
tunnel,
which
basically
supports
one
leave,
a
one-foot
over
which
you
will
deliver
service
in
more
general
case,
it
doesn't
have
to
be
a
single
tunnel,
but
it
could
be
say
two
or
three
with
nodes
in
between
which
is
basically
can
terminate.
They
say:
Donald
regroup
the
service
and
put
it
back.
So
in
this
respect,
they
are
not
so
much
different
from
Ethernet
switches,
except
they
have
like
that
rigid.
N
R
S
Okay,
okay,
hello,
everyone:
this
is
Chile
I'm,
going
to
provide
into
this
practices
about
how
to
apply
the
CMP
s
to
the
Beyonder
Oaxaca
coaching
next,
since
last
since
they're
our
top
team
of
this
job,
that
we
make
some
updates
as
soon
as
there's
updates.
That's
a
based
on
the
comments
just
before
the
opposition.
First
up,
this
is
some
discussion
about
how
to
report
into
the
OT.
S
You
see
a
link
with
healing,
where
other
one
fingers
will
try
to
discover
this
and
a
second
one
is
some
it's
based
on
the
comments
that
from
as
in
maybe
blossom
Bangkok
meeting.
Where
are
some
tags
about
how
to
know?
Should
you
tell
the
odious
any
link
with
awesome
weather?
This
is
obviously
falling
how
to
use
it
food
into
it,
societies
and
also
just
the
office
after
way.
Some
minutes
later,
this
comes
this
graph
that
we
also
received
some
other
comments.
S
The
first
common
testing
has
just
about
the
case
without
a
3i
that
is
things
with
surah,
at
least
to
describe
a
trip
or
anything
about
audio
singing
with
healing
also
ways.
They
save
some
comments
about
the
attacks
referring
written,
try
to
align
with
some
other
document,
not
to
overlap
with
the
currently
like
OD.
Okay,
arrived
in
document
analysis,
kind
of
comments;
okay,
okay,
yeah,
it's
proteins
and
I
think
it's
pretty
simple
and
comments.
E
O
O
O
O
O
This
is
a
picture
that
is
and
has
ASCII
art
in
the
document,
and
this
is
the
I
think
the
weak
point
of
the
document
I
tried
to
come
up
with
a
reference
model.
I
need
help
to
go
with
this
model
and
see
if
it's
actually
correct,
I
think
I
had
so
I
have
had
cross
comments
on
it,
so
people
saying
this
is
correct:
it
should
do
it
like
that.
No
no!
This
is
strong.
You
should
do
it
like
this
and
then
so.
O
O
So
we
put
the
terminology
of
using
him
we
actually
given
back
to
our
CV
4201,
where
link
bundles
are
actually
defined
as
logical
interfaces
on
or
out
there.
We
actually
do
the
same
saying
where
the
Flexi
group
reflects
see
links
on
these
boxes
here
it
in
in
and
out
I
think
this
is
the
second
time
I
put
them
back
in.
O
We
can
advertise
those
interfaces
into
the
routing
system
and
we
can
actually
set
up
an
LSP
on
top
of
those
that
the
deflection
network
and
if
we
go
to
real
life
deployment,
they
can
use
all
of
this
part
of
this
any
subset
and
we
can
actually
control
the
network
not
only
through
a
control
plane.
We
can
also
use
an
NMS
or
a
centralized
controller
or
it
can
be
modeled,
even
if
you
want
to
anything,
is
possible
and
actually
I
think
any
combination
of
them,
I
sort
of
think.
It's
also
possible
that.
O
That
line
break
was
not
in
my
slide.
Stone
has
been
introduced,
so
when
we
do
flexi
configuration,
we
can
do.
We
have
a
lot
of
alternatives
are
certainly
previous
slides.
We
can
use
an
animus
only
set
up
everything
down
to
the
last
bit
on
the
MPLS.
Lsp
became
the
combination
or
the
Animus
and
conflict
gm+
control,
plane
or
any
type
of
the
method
we
discussed
before.
O
O
Okay,
so
let's
start
with
an
example
where
we
have
an
MS
or
a
controller,
so
the
controller
talked
to
the
shin
on
Bo's
and
flexy
group
is
established
across
and
then
we
leave
the
Flexi
client
out
just
for
in
this
example.
But
this
information
could
be
exported
to
the
control
plane
and
then
to
any
routing
system
that
we
have
and
then
yeah.
So
then
we
have
a
network
where
we
have
a
mix
of.
O
It's
kind
of
playing
a
trick
on
me,
so
you
want
to
set
up
a
connectivity
from
here
to
there
and
the
straight
link
you
would
where
you
could
go.
If
you
only
use
term
cost
metrics
would
be
going
across
from
here.
So
now
we
introduce
the
condition
that
we
actually
want
to
use.
Flexi
cable
links
all
the
way
from
A
to
B,
and
then
we
actually
get
something
that
traverses
the
network
that
way
I,
don't
know
it
it's
the
best.
You
can
find
a
better
way,
but
I,
don't
think
so.
E
O
O
Yes,
between
two
boxes,
it
termed
originates
and
one
box
terminus
on
the
other,
but
in
the
control
claim
you
can
see
which
boxes
that
has
flexi
links
attached
and
to
which
spoke
in
the
box.
It
go.
So
you
can
pick
interfaces.
Someone
called
a
daisy,
chaining
I
think
I
was
deeper,
but
yeah.
That's
what
we
do,
but
that's
what
you
do
always
on
these
nuts.
You
you
probably
do
you
do
exactly
the
same
thing.
O
O
So
next
steps
well,
since
we
didn't
present
at
last
time,
we
kind
of
did
go
in
hibernate
for
three
months
and
now
we're
back
here,
I'm
prepared
to
work
further
than
this.
But
yes,
now
it's
kind
of
hard,
because
I
need
the
input
from
very
working
room
and
the
only
time
only
way
to
get
that
is
actually
making
a
working
group
document,
and
we
have
other
reflection
document
that
actually
depends
on
this.
So
we
should
really
make
it
a
working
group
document
quite
soon,
I.
E
O
N
S
S
S
O
O
S
O
C
M
To
set
up
my
recipes
to
hoagies,
where
I
care
is
the
bandwidth
of
the
link,
whether
there
are
flexi
capable
or
not,
I
need
to
her
lips
over
a
blue
or
yellow
things.
I,
don't
care
if
they're
achieve
200
with
flexi
or
not
it's
matter
of
setting
up
to
the
capita
capacity
between
the
neighbors
using
plexi
or
not.
But
I
don't
see
the
use
case
of
setting
up
our.
O
O
Weight
you
kind
of
interrupted
me
so
say
if
the
way
way,
bigger
bandwidth
and
it's
possible
to
split
them
up
in
granularity.
So
you
can
actually
allocate
bandwidth
on
a
flexi
client
for
a
specific
purpose,
and
you
can't
do
that
on
Ethernet
on
a
regular
Ethernet.
So
in
regularly
that
you
mix
packets
over
the
same
same
link
in
Flex
and
in
flexi,
you
can
send
traffic
on
a
dedicated
bandwidth
that
it's
just
for
you.
O
C
M
T
T
H
T
T
They
uses
a
flexi
on
later,
no
flexi
client
flexi
is
configured
underneath
even
and
it's
done
on
a
very
coarse
granular
basis
by
configuration.
So
the
idea
that
it's
dynamic
and
we
can
switch
the
bandwidth.
You
know
with
a
control
protocol
to
Tilly's
point,
which
I
think
I
understood,
there's
not
a
handle
for
the
control
protocol
to
get
ahold
of
unless
you're,
exposing
it
algorithmic
ste
for
66
beat
layer,
in
which
case
it's
not
Alya
flexi
and
it
probably
violates
8o
2.3.
T
N
To
help
you
know,
I
think
you're
talking
with
guys
from
different
levels.
Okay,
so
it
is
true
that
on
the
on
the
flying
player,
it
just
Ethernet
right
but
but,
for
example,
how
to
actually
implement
this
Ethernet,
it's
a
different
story
and
require
different
orchestration
or
control
mechanism.
It
just
like,
for
example,
when
you
are
castrates
a
layer,
3
VPN
service
from
the
client
point
of
view,
just
leave
MVP
enough
from
the
orchestrator
point
of
view.
It's
many
many
different
other
things
he
put
in
setting
up
tunnels,
steering
services
and
so
forth
right.
N
N
Then
it's
basically
it
answers
both
children
and
do
it
because
they're
thinking
about
like
from
the
client
point
of
view,
is
just
purely
son.
Yet
ok,
just
not
from
the
client,
is
a
pure
way,
our
three
BPM
but
actual
implementation.
That
makes
very
big
difference
whether,
for
example,
the
f3
BPM
is
T
rather
I
can
say
it
or
not
and
other
things
right.
So
it
requires
different
models
and
different
control
mechanism.
E
B
Okay,
so
I
think
people
are
just
concerned
that
those
intermediate
nodes
you're,
showing
switching
of
the
flexi
on
and
I,
think
you're
doing
that,
but
I
guess
what
it
would
come
down
to.
It's
how
you
advertise
the
bandwidth.
The
flexi
is
capable
of
optimizing
better
for
different
bandwidths.
Instead
of
nailing
up
a
whole
save
hundred
gig
on
a
wavelength,
it's
able
to
to
do
more
granularity.
C
B
I
guess
how
I
guess,
Julianne
area
Mary
said
that
we
would
not
look
at
necessarily
having
to
prefer
a
flexi
route
in
that
it's
flexi.
The
preference
would
be
more
that
you
would
have
better
bandwidth,
optimization
for
those
links.
So
when
it's
advertised,
you
would
be
able
to
see
that
you
could
support
this
intermediate
rate,
and
so
it
would
be
not
so
much
that
it's
advertised
as
flexi.
B
It
would
be
advertised
more
that
it
can
do
this
granularity
and
they
only
so
I
think
you
know
it's
just
as
when
we
had
SDH
signal
or
TN
signal
you
advertise
different
granularities
that
you
could
support
and
that's
what
people
are
looking
for
in
their
Browning
tables.
So
here
that's
wrong
thing.
You
miss
it's
not
so
much
that
it's
flex
EU
when
I
will
text
you
want
to
advertise
the
types
of
bandwidth
that
can
be
supported
on
that
way
and
that's
what
you
would
go
hunt
for
then.
O
T
So
the
way
I
understand
this
is
you
wouldn't
advertise
it
as
a
flexi
climbing?
What
advertise
it
is.
An
Ethernet
link
with
very
good
QoS
characteristics,
things
that
could
hit
a
very
fine
level
of
jitter
of
things
that
you
link.
They
could
hit
a
very
fine
level
of
latency
things
that
could
handle
a
very
strict
key
or
less
policy,
but
it's
Ethernet
and
it
gets
advertised
and
propagated
through
the
routing
and
the
control
protocols
as
Ethernet,
which
we
can
already
do
today.
A.
T
O
T
O
T
Nothing
to
grab
onto
the
Flexi
group
there
is
nothing
exposed
to
grab
on
to
and
advertise
I
can't
send
it
in
routing
I
can't
send
it
in
a
signaling
protocol
to
Igor's
point
yeah.
The
orchestration
would
be
different.
Why?
Because
it's
in
the
interface
configuration
that
I
configure
the
Flexi
groups
and
the
channels
and
associate
them
together,
not
in
the
control
protocol
and
not
on
the
priority.
So
it's
very
static.
It
can
be
changed
and
then
that
can
be
reflected
into
the
routing
into
the
control
protocol.
E
Yeah,
this
is
the
most
complex
a
session
since
M
KSTP.
E
L
T
The
significant
Cove
Ericsson
again
it
the
risk
of
confusing
the
situation
even
further.
There
is
work
on
a
flexi
switch
technology.
Flexi
switched
I
mean
for
tying
sister
right.
You
have
flex
T
switch
sub-layer.
That
would
be
something
that
we
possibly,
that
we
would
possibly
look
at
C
camp
to
create
a
control
protocol
before
right,
but
there's
a
whole
lot
of
extra
work,
there's
a
whole
lot
of
work.
That
needs
to
be
done
instead
of
the
15
first,
oh.
U
E
P
P
P
Okay,
here
is
our
design
of
the
erotic
of
interfacing
Flossie.
We
different
Madhu,
sir
blocks
is
Flossie
group
and
flask
hunt
and
a
measurement
as
an
interface,
and
it
opposed
argument
for
ITF
interface.
Emotive
grant
is
actually
contending
for
secret
and
they
can
be
convicted
based
our
results
and
some
fallacies
out.
P
This
is
our
young
model,
trees
and
okay.
This
is
what
we
have
to
after
previous
last
status
meeting.
We
because
plastic
Ron
is
decoupled
from
philosophy
guru,
so
it
is
a
motive
as
a
new
type
of
interface,
so
a
new
France
account
can
be
more
appropriate
motive
and
easy
to
be
recommend
to
support
more
some
new
Mac
layers
features
such
as
a
flow
can
show
and
the
way
you
want
to
keep
this
module
simple.
P
P
Do
is
a
flat
c
interface
module.
So
here
is
a
separation.
Where
is
this
module
and
the
way
that
to
coordinate
ways?
Another
document
ulcers-
and
we
know
no
sizes
difference
way
present
here
and
the
wither
back
to
post
exists
difference
on
the
mailing
list,
and
maybe
it's
a
working
group
and
another
folks
here
past
to
present,
as
here
'dear
under
give
us
some
comments.
R
P
Is
less
right?
So?
Okay,
yes,
I
see
I'd
like
to
coordinate
Raza
officers,
and
but
since
there
are
quite
different
opinions,
so
maybe
we
internally,
we
will
decide
to
update
the
documenting
our
way
and
there's
a
working
group
who
may
help
us
to
make
a
decision
how
to
his
works,
are
shall
be
processed
and
the
would
like
to
como
people
to
help
us
to
reveal
this
document.
S
P
P
A
P
Okay,
so
the
question
the
first
day
is
because,
if
you
module
supported
to
mdhil
and
the
architecture,
the
young
models
and
the
architecture
also
can
provides
ability
to
require
to
help
you
to
do
the
configuration
transaction
it
can
help
you
intended
Edisto
the
intended.
It's
though
they
can
help
you
to
verify
whether
this
mode,
this
data
can
be
configured
whether
asa
can
be
whether
it's
in
print
and
in
the
operational
data
store.
A
P
R
R
Another
point
is,
if
you
automatically
add
a
new
client,
usually
it's
done
in
sequence.
You
cannot
add
to
it
her
on
some
day
on
the
first
day,
if
you
do
it
her
one
by
one
for
example,
today
and
this
week,
then
you
can
send
the
to
working
relations
in
sequence
and
so
that
the
plane
can
still
regards
the
first,
the
previous
condition
as
current
a
and
it's
a
new
combination
as
a
kind
of
B.
So
it
doesn't
matter.
We
provided
her
in
the
communication
module
as
a
hunter,
a
II
or
Canada
B.
R
U
H
S
Okay,
thank
you
little
chilly
again
for
this
protein,
since
it's
at
some
it
model
in
a
way
for
powerful,
flexi
next.
Okay,
this
is
a
content
of
this,
so
this
is
an
updated
of
the
draft
of
progressing
or
at
UTS
a
top-15
and
conversing,
with
a
draft,
and
also
for
questioning
the
minister
next
step
and
for
update.
So
it
just
as
us,
we
made
a
point.
The
force
of
III
doesn't
expire,
the
value.
S
I
think
this
can
be
usually
the
case
of
a
mismatch
checking
and
a
second
one.
We
added
to
configuration
more
the
for
Policy
Canada.
It
could
be
master/slave
mode,
it
could
be.
We've
got
this
requirement
from
Cisco
from
some
vendors,
and
we
also
found
some
related
information
related
at
escorting
you
for
a
CI
and
the
soda
where
I
sorry.
We
either
flesh
you
kinda
to
sorta
location
information.
S
We
give
someone
s
as
a
why,
when
you
know
why
we
want
to
do
this,
we
also
had
a
new
answer
from
Cisco
and
the
photo
fluorescein
necessarily
sort
of
updated
accordingly
and
next
time.
This
is
a
clinical
location.
Sort
of
a
locus
information
we
didn't
worry
about
is
a
lot
because
was
the
first
time
we
don't
want
to
put
this
information
in
2000
because
we
are
thinking
about.
It
was
it's
differently.
Oh,
it's
similar,
okay,
after
some
discussing
with
the
guys
from
actually
some
all
also
recalled.
S
Some
information
from
flexi
I
did
with
continuous
in
the
same
at
the
first
group
of
the
fleshy
planet
has
just
some
ethics
over
yeast
notified.
So
was
single?
Okay,
that's
the
phone
network.
A
point
of
a
opposed
for
a
second
and
in
a
cortical,
the
existing.
You
know
sim
that
water
layer
it
can
be
aware
of
the
existence
of
a
saucer.
That's
why
I
would
do
this
in
the
modeling
and
next
we
add
some
explain.
Okay,
the
model
value
number
five,
my
calendar
configuration
we
just
use
also
to
do
a
mismatch.
S
Mapping
from
different
level
and
for
the
right
side
of
this
is
an
idea
that
we
add
the
true
configuration
mode
of
a
flex
eternity.
It's
amazing
that
you
for
accelerating
from
some
description
idea
but
in
the
country
is
not
indeed
over.
1823
I
got
some
feedback
from
question
14.
We
should
talk
about
this
over
the
question
learning
to
sing
together
you
from
then.
Why
took
all
the
information
over
Sierra
I'm
standing
currently
either.
S
Two
moles
of
the
first
static
mode
with
us
very
uses,
a
country
to
to
configure
both
the
transceiver
side
and
the
receiver
side
and
and
a
father
ii
know
that
we
just
say
it's
masculine.
Even
more.
We
just
used
to
configure
the
soft
side.
Another
computer
information
can
be
conveying
that
were
destinations
through
overhead
and
the
current.
This
is
an
update
on
the
photos.
We
seen
the
progress
in
your
life
and
activities
that
was
15
I,
think
that
is
no
okay.
S
Public
progress
in
your
country
that
working
young,
a
2.1,
was
about
another
publicity
years
and
for
715
I
think
they
concentrated
as
8.8
23
I
mean
didn't,
have
one
to
income
that
the
Frazee
I.
True,
wasn't
true
that
hey
the
Pharisee
instance
and
that
way
also
submit
a
pricing
model,
comes
video
sent
to
the
management
question
question.
S
Question
what
do
we
define
another
kilobytes,
also
some
local
scope.
Okay,
next
step
whistling,
who
currently
the
requirements
is
required,
a
courier.
We
have
a
lot
of
disgust
and
violation.
We
incur
the
ways
for
a
CIA
date
upon
an
editor
with
also
some
guys
from
question.
14
question
11
our
single
I
said
so
there
was
that
yeah
from
Anacin.
Thank.
T
You
thanks
things
in
the
group
now
speaking
from
Erickson.
This
is
this
model
seems
more
consistent
with
my
understanding
of
flexi,
and
my
question
is:
have
you
confirmed
this
with
OIF?
Have
you
run
this
bio
I
up?
It
says?
Surely
he
went
by
your
slide,
one
for
Title
II
to
move
forward,
but
I
was
reading.
That
so
is
this
confirmed
to
be
like
a
model
that
would
reflect
the
old
away,
a
flexi,
2.0
interface,
I
guess.
T
T
Missing
in
the
work
that
I'm
familiar
with
is
a
model
to
be
able
to
configure
by
a
yang
a
flexi,
2.0
interface
yeah,
that's
what's
missing,
so
I
was
a
little
distraught
at
the
previous
hunt.
The
previous
draft
that
said
that
they
were
defining
flexi
is
a
new
type
of
interface,
because
really
it's
an
Ethernet
interface,
like
you
presented
here
right.
G
F
How
many
hallelujah
this
comment
applies
to
both
this
presentation
in
the
previous
one
I,
don't
know,
we
understand
that
there
are
related
to
war,
people's
OIF
and
itu-t,
but
the
more
importance
in
Syria
is.
We
also
need
to
design
a
young
model.
That's
fits
into
the
Hawaii
of
young
system
and,
as
we
need
to
look
into
other
young
model
specifies
here.
So
as
far
as
I
know,
I
didn't
see
any
merit
written
directly
augmentations
to
other
models.
T
P
Email
from
my
colleague
and
Sims,
there
are
quite
different.
There
are
quite
a
lot
of
issues,
so
maybe
we
are
for
the
mother
Jean.
Maybe
it's
a
lot
of
issues.
We
should
further
discussions,
so
I
think
I
think
that
we
need
flux.
The
young
lady
mother,
insist,
I
have
but
I
don't
think
we
can
do
this
walk
face
of
this
document.