►
From YouTube: IETF92-PALS-20150324-0900
Description
PALS meeting session at IETF92
2015/03/24 0900
A
B
B
B
C
C
Actually
we
didn't
give
a
cop's
at
one
at
a
time
was
called
rock
martini
actually
and
and
then
we
we
created
a
working
group
called
sea
oats
and
that
then
was
renamed
to
pwe
three
and
then
I
guess
he
was
a
name
to
pals
now
and
that's
when
I
first
actually
I
met
Stewart
shortly
after
that
on
a
big
argument
about
two
bits:
the
feckin
embed
that
can
bit
exactly
actually
I
can
probably
say
what
happened.
Then.
At
this
point
it's
been
15
years.
C
There
was
a
typo
in
in
the
in
the
header,
and
so
we
basically
reverse
the
effect
on
him
back
in
bits
in
the
pw
free
header.
They
control
word,
and
that
call
you
know
it
was
no
reason
the
particular
reason,
and
so
Stewart
actually
found
that
tenancy.
He
was
the
ascii,
also
questioned.
Why
do
you?
Would
you
reverse
this
bit
that
we
couldn't
quite
explain
it?
C
That
was
the
problem
is
that
there
was
a
typo
in
the
header,
but
the
problem
is
as
soon
as
nobody
notice,
a
type
of
document
and
by
the
time
Stewart
product,
the
problem
everybody
had
implemented
it
that
way
already.
So
there
was
no
has
no
good
reason
to
change
it
anyway,
so
this
at
this
time.
So
what
I
did
in
in
in
the
past
there's
been
to
two
revisions
actually
of
this
RFC
4447
biss,
the
first
one
I
took
all
the
rat
as
we
over
the
past
15
years,
and
you
know
they
were
accepted.
C
There
was
a
lot.
You
know
many
small
technical
problems,
mostly
text,
and
you
know
clarifications
and
things
of
that
and
I
included
that
in
the
document
and
then
we
just
took
an
overpass
at
it
and
we
basically
at
this
point-
you
know
it's.
This
talking
has
been
around
for
a
long
time.
It
we
just
wanted
to
get
a
refresh
to
include
the
errata
and
a
movie
to
proposed
standards
or
whatever
the
next
step
is
we
I
believe
that
right
now,
it's
it's
ready.
C
For
you
know,
it's
last
call
we're
not
trying
to
include
one
of
the
things
that
I
wanted
to
do
in
this
document.
I
didn't
want
to
change
anything
radically
like
text
or
or
include
over
references
to
what
has
been
done
put
after
the
document
or
anything
like
that.
It's
just
a
very
simple
fix
for
for
the
errata
and-
and
you
know,
kind
of
update
to
the
current
temp
later
like
EF,
so
so
that
I
think
there's
one
thing
they
came
up
on
the
list.
C
You
know
it
doesn't
say
it
has
to
be
norther.
There
is
a
way
FTL.
These
is
opposed
because
you
should
be
able
to
send
them
in
any
order
you
want.
Unless
somebody
specifically
says
there
is
a
good
reason
to
send
in
a
specific
order,
and
so
so
I
don't
really
feel
like.
We
need
to
add
a
piece
of
text
that
says
you
know
you
can
send
in
any
order,
because
that's
implied
it
doesn't
say
you
have
to
send
it
in
order.
So
we
should
be
fine
with
that
and
in
your
comments,
yeah.
D
Clarifying
question
in
a
router
that
you
looked
at
the
fixed
in
next
release:
a
router
as
well
as
the
there
are
several
classifications
for
a
rattle
is
reject,
there's
accept
which
obviously,
you
do
have
to
do,
but
there's
also
authors
to
look
at
when
the
document
is
update.
Updated.
Did
you
pick
those
up
as
well?
Yes,.
D
C
D
C
D
C
D
D
A
E
A
D
C
C
I
mean
okay,
so
so
let
me
ask
you
a
question
here
of
them.
What
does
out
of
season
learning
about
which
number
it
is?
Is
the
VCC
be
sorry
the
control
protocol
negotiation,
which
really
kind
of
updates
whole
session,
which
we
can
include
those
updates
and
obsolete,
that
our
see
with
the
new
one
that
will
come
out
I.
D
C
D
C
Okay,
but
here's
the
problem
is,
this
document
is,
is
well
established.
The
procedure
in
these
documents
will
establish.
This
is
ready,
for
you
know,
proposed
standard
or
standard
trac
right,
those
are
have
ceased.
They
updated
like
the
control
world
updates
and
so
forth.
That
is
not
exactly
the
same
status,
so
we.
D
Need
to
go
through
each
one
on
a
case-by-case
basis
and
see
because
that
line
will
not
be
that
line
of
updates
will
not
be
present
in
the
new
RFC
right
right
right.
Okay,
so
we
need
to
make
sure
that
the
reader
of
the
new
RFC
Rosa,
where,
by
reading
that
document
or
reading
another
document
of
the
work
that
is
presented
in
the
update
text.
C
E
C
C
C
E
I
one
comment
on
me:
Eric
gray,
at
ericsson,
I
have
one
comment
on
the
the
thing
you
were
talking
about,
with
the
order
of
the
tles
or
and-
and
it
seems
to
me
that
there's
sort
of
a
proof
of
concept
that
there's
some
confusion
over
this
and
it
you
could.
You
could
argue
that,
because
we
don't
say
it
has
to
be
in
this
order
right,
we
are
implying
that
it
can
be
in
any
order.
E
But
it's
equally
true
that
if
it's
shown
in
a
specific
order,
it's
implied
that
that's
the
older
you
might
be
consp,
ecting
it
and
so
I
think
it
might
be
useful
to
put
in
it
these.
These
may
appear
in
any
order
in
the
document.
On
the
other
point,
you're
just
talking
about
the
updates,
I
completely
agree
with
chairs
on
that
I
mean
minimally.
After
all,
things
are
considered.
You
have
to
at
least
put
a
paragraph
in
there
and
says
you
know.
Please
also
review
the
following
rfcs,
which
have
updated
this.
That's.
C
Sure,
yeah,
okay,
so
about
to
order
they
come
and
I
have.
Is
that
earlier
in
the
draft
there
is
actually
a
set
of
TL
VZ
with
the
picture
that
has
to
be
in
order,
and
it
says
this
has
to
be
in
the
order.
It
explicitly
says
that
so
the
rest
I
mean
people
yes
can
assume
the
next
one
has
to
be
in
order
or
not.
I
mean
it's
kind
of
a
clarification
point.
C
We
can,
you
know,
I
can
I,
don't
have
a
problem
to
add
some
text
I,
I
we
could
add
a
text
that
says:
did
you
know
this?
Don't
necessarily
have
to
be
in
order
or
something
like
that.
But,
as
you
know,
the
interpretation
is
that
they
should
be
in
order,
and
the
text
doesn't
say
in
that
section:
it
has
to
be
in
order.
So.
E
B
F
F
A
G
A
F
So
brief
history,
we
started
with
this
effort
was
started
in
2011
and
it
was
first
presented
in
l2vpn.
Then
it
went
little
bit
dormant.
There
wasn't
sufficient
interest
in
the
l2vpn
group.
The
service
providers
were
requesting
this
feature,
so
I
picked
it
up
and
presented
in
pw
III
in
pw3.
We
went
pretty
far
discussions
and
there
was
a
lot
of
acceptance
and
all
that
in
88,
ITF
88.
We
did
a
detail,
protocol
presentation
and
was
well
received,
but
at
the
end,
almost
towards
the
end
there
was
a
I
think.
F
The
chairs,
ND,
malice
and
other
people
suggested
that
we
should
really
take
it
to
l2vpn
working
group.
So
again
we
took
it
back
to
l2vpn
working
group
because
original
mac
withdraw
signaling
for
dynamic.
Pseudowire
were
actually
done
in
l2vpn
working
group,
so
we
took
it
to
l2vpn
working
group
since
already
in
the
last
stages
in
pw3
we
did
allah
call
and
I.
Think
then
the
group
split.
We
had
some
more
comments,
and
so
we
are.
This
is
an
update
on
with
the
comments
next
slide.
F
Please
so
the
as
we
as
you
guys
know,
Mac
withdraw
signaling
is
actually
it's
an
optimization
to
mitigate
the
black
holes
during
pseudo
I
switch
over.
It's
already
done
for
dynamic
slower,
as
I
said
this
is
this
is
to
do
it
over
the
static
pseudo
wires.
So
this
this
mechanism
actually
is
clone
from
the
dynamic
match,
because
signaling
for
which
is
already
defined
in
the
4762
and
pseudo
as
status.
Signaling
was
actually
taken
from
the
pseudo
I
status.
F
So
yeah
here
is
just
a
brief
graphical
display
on
where
it
is
used
a
little
bit
of
animation.
Here
you
want
to
click
yeah.
So
if
you
have
a
fault,
if
it's
empty,
you
is
dual
home
to
p1
and
p2,
and
if
the
primary
pseudowire
is
on
going
to
the
p1
and
you,
if
you
had
a
fault
you
the
mech,
you
need
to
the
empty
use
which
is
over
to
the
standby,
pseudowire
and
more
click.
And
since
this
is
a
static
sort
of
wire,
you
need
to
do
a
signaling
of
mac.
F
We
draw
one
more
click,
and
so
this
is
how
it
progresses.
Mtu
sends
the
mac
withdraw
signal
over
the
static
pseudowire
has
described
in
this
draft,
which
then
gets
propagated
to
all
of
the
remote
pease.
The
whole
idea
there
is
to
so
that
yeah
the
whole
idea
there
is
that
all
those
remote
p
is
p
6,
5
4,
3
2,
who
were
sending
the
traffic
for
stations
behind
the
MTU.
They
were
they.
Those
remote
piece
did
not
know
anything
about
the
pseudowire
going
down
on
P
between
mt,
u
and
P
1.
So
they
were.
F
F
That's
the
utility
of
this
feature,
okay,
so
that
the
Wi-Fi
on
the
static
pseudowire,
so
the
this
mpls-tp
you
know
as
is
taking
hold,
especially
in
the
access
networks
and
thats,
a
static,
provisioning,
LSPs
and
that's
doubt,
tails
with
the
static
provisioning
of
the
pseudo
wires.
So
this
nessa
see
it's
this
om
signaling
for
the
statics
of
the
wire
next,
so
the
coming
back
to
the
central
topic
here,
the
response
to
the
comments,
the
you
know.
F
These
comments
were
quite
comprehensive
and
Stuart
provided
those
who
Thank
You
Stuart
so
that
I
have
not
addressed
all
the
comments,
but
I
had
taken
the
most
of
the
the
general
bigger
comments.
The
first
concern
was
that
magnet
rho
signaling
on
ldp
was
actually
you
know.
Ldp
uses
the
tcp,
so
there's
a
guaranteed
delivery,
reliable
acknowledged
in
order
and
this
one.
Since
this
is
not
using
tcp,
it
says
it's
over
the
static
pseudowire.
So
how
do
you
guarantee
the
delivery?
And
so
the
the
comet
is
correct?
F
D
F
D
A
F
I
actually
added
a
paragraph,
because
you
made
that
comment.
I
said
yes,
I
will
and
I
added
yeah.
I
made
it
very,
very
clear
that
this
is
a
best-effort
scheme.
It's
an
optimization
scheme
and
there
are
others
other
mechanisms
by
which
the
node
will
relearn.
So
if
you
have
a
bidirectional
unicast
frame
exchange
going
on,
then
you
will
learn
and
the
second
is
the
you
know:
the
aging
timer
sorry.
F
The
IRA
7361
is
an
optimization
the
opted
for
the
in
the
73
61
that
the
scheme
there
is
that,
if
you,
if
we
go
back
to
the
previous
badger
that
graph
here,
the
in
73
61
says
that
the
p1
should
send
the
Mac.
We
draw
oh
yeah.
Instead
of
having
the
P
to
send
Omega
door
to
every
the
p1
sends
the
Mac
we
draw
with
a
specific
list
of
mac
addresses
so
that
you
don't
have
to
flush
all
the
mac
addresses
from
the
other.
But
here.
F
F
F
A
H
F
H
F
F
F
Make
that
clear
do
well
there.
Well,
you
know
it
says
the
static
over
static
pseudowire.
There
is
no
ldp
in
bed,
yes,
yeah,
okay,
the
couple
of
like.
Oh
yes,
so
you
know
the
general
scheme
is
that
you
will
relearn
with
the
you
know,
bi-directional
uniform
unicast
for
him
exchange
and
obviously
you
have
the
mac
address
aging.
Usually
the
default
time
is
300
seconds.
So,
if
you
know
in
the
verse
trick
worst
case
scenario,
you
would
have
a
black
hole
in
45
minutes.
F
Another
point
is
that
you
don't
want
to
actually
keep
retrying
forever,
because
the
utility
are
the
benefits
of
mac,
withdraw
signaling
to
flush
mac
addresses
the
case.
You
know
as
the
time
goes
by,
so
you
know
it's
it's
a
very
small
window.
You
want
to
do
the
flush
and
you
know,
carry
on
the
next
slide,
please.
F
So
the
other
comment
was
you
know
what,
if
you
are
using
the
big
mac-
and
you
know
if
that
list
is
too
big.
There
is
no
fragmentation
and
you
know
how
do
you
handle
all
that?
So
you
know
the
response
is.
Actually,
there
are
a
couple
of
ways
you
can
handle.
That
one
is
that
it
is
a
sequence
number
based,
so
you
eat,
you
can
treat
a
concise.
F
A
F
F
H
F
F
B
F
F
Yeah
so
you
handle
the
you
know
the
MTU
exit
by
the
empty
list
or
break
break
down
the
list
into
multiple
videos.
Okay,
the
second,
the
next
slide.
Please
yeah.
Sorry
there
was
a
yeah.
The
concern
is,
how
do
you
handle
the
out
of
order
again?
This
is
not
a
reliable
delivery,
but
we
do
have
a
sequence
numbers.
F
So
what
we
have
stated
in
the
draft
is
that
if
you
have
a
out
of
order,
so
if
you
have
a
lower
sequence,
number
arrive
later
than
the
higher
sequence
number
you
you
actually
ignore
that,
and
but
you
do
a
kit,
so
the
highest
sequence
number
takes
the
presidency.
So
now,
when
that
happens,
if
you
ignore
that
list,
you
will
you
will
not
flush
the
subset
of
the
MAC
addresses,
which
again
is
fine,
because
then
you
know,
if
you
did
the
means
those
mac
addresses,
they
would
be
relearned
anyway
by
other
mechanism.
F
So
so
it's
not
a
calamity
that
you
lost
a
subset
of
mac
addresses
next
time
the
fixed
three
yeah.
We
do,
the
three
retries
actually
I.
Can
you
go
back?
I
did
not
talk
about
the
sequence
number
rap.
There
is
an
RFC
4385,
we
do
do
sequence
number
in
the
control
word
and
this
RFC
4385
actually
describes.
How
do
you
handle
that?
How
do
you
detect
a
wrap
and-
and
how
do
you
handle
it-
the
overflow
so
I'll
go
to
the
next
slide.
I,
ok,.
F
So,
yes,
we
have
33
retries
comment:
oh
the
fixed
interval,
so
the
yeah,
the
original
draft
said
that
you
know
if
you,
if
you
don't
receive
that,
if
the
sender
doesn't
receive
the
ACK
yeah,
you
say
you
know
you
resend
after
1,
second
interval
3
fixed
intervals,
and
the
comment
was
that
you
know
in
case
of
scaled
configuration
you
there
may
be
many
pseudo
was
doing
this
and
you
cannot
guarantee
that
everything
so
I
think
the
better
thing
and
I
have
changed.
The
text
is
that
you
can
do
the
exponential
back
up.
F
So
yeah
I
mean
we
had
the
that
was
very
good
comments,
very
comprehensive.
There
was
a
lot
of
other
comments.
Also,
I
want
to
address
them
all
here,
but
it
made
the
truck
better
and
I
think
we
feel
that
we
have
it's
now
in
a
much
better
shape.
So
next
slide,
please,
yes,
so
I
think
you
know
we
have
asked
for
a
new
channel
type
x
28
and
we
feel
that
we
should
give
go
for.
The
last
call
again
comments.
H
A
B
If
you
feel
strongly
enough
about
it,
I
suggest
that
you
put
your
comment
on
the
list
in
there
and
see
if
you
get
any
positive
reaction.
Ok,.
D
F
F
F
Will
have
to
show
where
to
move
to
and
all
that
stuff
and
and
the
sender
doesn't
know
where
it
should
move
to
like,
for
instance,
in
this
case,
then
p
2
would
have
to
know
which
sort
of
our
phase
and
what
that
should
have
our
coincides
on
the
other,
remote
Pease
right
here.
This
is
just
saying:
hey,
plus
this
make
it
unknown,
it
will
get
broadcasted
and
it
will
relearn
the
one.
F
B
F
B
D
B
That's
marked
on
the
carpet
that
guarantees
that
you're
in
the
range
of
the
camera
so
that
anyone
who's
watching
out
in
the
web
or
watching
the
recording
at
some
point
in
the
future
can
actually
see
who's
speaking.
So
that's
the
reason
why
there's
a
pink
box
on
the
carpet?
That's
so
now
you
all
know
well.
D
D
D
Okay,
so
what
not?
D
What
this
is
about
is
about
defining
the
behavior
of
spfd
on
the
VC
cv
channel
for
a
signaled
and
prostatic
pseudo
wires
and
uncle
slightly
unconventionally.
It
covers
both
are
the
l2tp
case,
as
well
as
the
over
and
mpls
PSN
case.
So
55
58
5
covers
this,
but
we
don't
very
often
talk
about
it
in
this
in
this
working
group,
so
we're
interested
in
l2tp
signal
pseudo
wires,
static,
pseudo
wires
and
a
future
revision
of
the
draft.
D
E
D
So
the
the
next
thing
that
we
need
that
we
need
to
design
as
part
of
this
is
the
the
the
mechanism
for
getting
the
spfd
discriminators
will
either
use
provisioned,
discriminators
or
use
the
the
alert
mechanism
that's
currently
being
proposed
in
the
seamless,
PSD
alert
draft
and
l2tp.
When
that's
used
for
advertisements.
There
is
a
draft
that
describes
how
we
add
that
to
lttp
encapsulation,
where
the
is
your
a
call,
we
can
either
send
BFD
natively
or
we
can
send
it
in
UDP
IP,
where
it's
going
in
UDP
ap.
D
It
will
go
so
as
it
does
today,
except
that
the
sbf
d
UDP
destination
port
will
be
used.
So
that's
actually
really
just
cookie
cutter
stuff,
where
l,
two
SS
for
l2tp
or
the
pseudowire
ACH
encapsulation
mechanism
is
used.
We
will
have
to
request
new
iono
types
to
identify
that
it's
spfd
being
carriage
next
slide.
Please
the
initiator
operations
are.
We
need
to
new
cv
types
for
the
operation
for
the
operation
option.
D
D
Bfd
seamless
specification,
the
capability
selection,
and
this
is
something
that
I
think
we
need
to
discuss
in
this
room
for
a
little
bit.
There's
a
question
of
what
is
the
order
of
preference
between
s
BF
d
and
b
FD,
and
I
understand
that
you
can
have
both
of
them
can
currently
working
as
well.
So
the
the
question
that
the
authors
posed
in
their
original
slides
was:
can
they
be
independent,
I
order
of
preference?
D
Someone
who
worked
in
this
work
quite
a
long
time.
I
would
be
worried
about
that,
because
I
would
like
to
make
sure
that
every
time
two
systems
come
up
every
time,
I
provision
some
systems,
the
behavior
is
exactly
as
I
expect
so
I'm
guessing.
We
may
end
up
with
all
three
cases,
one
the
other
or
both
Greg.
D
B
And
email
sighs
have
a
comment
as
a
participant
rather
than
as
a
chair.
My
own
personal
druthers
is
that
we
as
a
working
group
just
make
a
choice
for
one
being
the
default
one
of
the
three
cases:
I,
don't
key
I,
don't
have
any
particular
opinion
as
to
which
one
should
be
the
default,
but
I
would
like
to
see
a
default
so
that
it's
so
if,
if
people
just
naively
bring
up
systems
without
provisioning,
this
it'll
work.
G
G
I
have
a
question
to
the
previous
slide,
where
there's
a
reference
to
a
CH
encapsulation
at
previous
slide,
so
in
case
of
a
CH
encapsulation,
there
is
no
IP
address
to
return
to
so
how
reflector
will
identify
the
source
of
s
BFD
echo,
request
and
return,
because
the
idea
of
SB
FD
is
that
reflector
doesn't
create
any
state
associated
with
a
session.
There.
D
F
G
D
D
D
Okay,
next
slide
next
slide,
so
I
think
it
would
be
really
useful
if
people
can
put
some
of
the
comments
that
they've
made
onto
the
list,
we
could,
if
we
can
have
a
list
discussion
on
this.
The
authors
are
keen
to
close
off
the
open
items
and
to
submit
a
new
revision
and
they're
going
to
ask
for
this
to
be
adopted
by
the
working
group.
Given
the
existence
of
spfd
I
think
the
working
group
does
need
an
opinion
on
it
and
don't
need
to
specify
a
solution
for
using
that
type
of
OEM
technology.
I
I
D
D
D
D
If
we're
negotiating
this,
because
if
you're
running
ldp,
you
will
have
to
have
the
default
negotiation
mechanism
working,
so
I
think
it's
just
that
they
didn't
put
the
text
in
rather
than
there's
no
intention
to
do
it
and
I
think
we
have
no
no
choice
but
to
solve
its
Institute
to
document
the
design
for
use
in
an
ODP,
signaled,
environment,
I'm.
Sorry,
I,
misheard
that
you
said
I
know
any
other
questions.
H
D
Don't
think
so
I
mean
BF
BFD
exists
anyway,
and
I
can't
see
as
ever,
deprecating
it
in
any
immediate
term
done
time
frame.
So
this
is
to
enhance
BFD,
but
I
am
presumably
when
you
want
to
introduce
some
measure
of
scaling
that
you
might
not
easily
achieve
with
BF.
Diva
BFD
has
got
some
useful
properties
anyway,
yeah.
E
J
G
E
D
So
the
answer
is
I,
don't
know
the
answer
that
question,
but
I
asked
the
author's
where
shouldn't
you
have
a
or
b,
and
they
said
no,
I
might
want
to
run
a
and
B
so
and
I
didn't
pursue
it,
but,
like
they
tell
me
they
need
to
run
a
and
B.
We
need
to
figure
out
how
to
have
get
both
ends
running
in
the
same
mode,
a
B
or
am
to
be
okay,
feel.
B
D
B
A
I
Hello,
yo
B
from
javi
this
topic
that
used
to
discuss
the
design
for
our
two
week
in
a
young
model.
Okay
I
see
a
nice
life.
Okay,
they
see
you.
The
history
revealed
in
the
Honolulu
IETF
491,
so
we
present
the
are
tweaking
a
young
configuration
a
model
in
the
past
working
group.
Then
after
the
meeting,
we
received
a
positive
response
from
the
several
experts
to
cooperation
to
define
the
young
for
review
so
because
of
the
working
group
I
used
to
change
it.
I
So
this
draft
is
also
moved
to
the
past
working
group
together
the
appropriate
working
group
to
go
on
to
discussion
okay.
So
we
have
a
meeting
to
discuss
the
design
for
the
earth
who
we
being
young
so
according
to
the
operation,
so
that
we
said
have
a
design
team
for
arrow
to
Whitney
Young.
So
this
is
from
a
sciphone
santoshi
show
and
a
commensal
so
and
also
that
we
go
on
to
welcome
more
members
and
a
cooperation
from
the
industry.
I
We
hope
to
add
the
beginning
to
negotiate
the
common
models
for
our
to
Virginia,
so
that
we
can
finish
the
young
walk.
Annotated
consists,
concern
concerns
as
a
beginning:
okay,
next,
okay,
so
the
design
team.
We
discussed
the
scope
of
the
arrow
to
VPN
because
in
the
past
ten
years
are
to
review
the
very
popular
and
the
all
kinds
of
our
VPN
solution
have
been
proposed
here.
So
this
goal
we
try
to
define
that
what
to
design
by
this
team
and
the
water
not
will
be
designed
by
this
team.
I
I
This
is
the
basic
components
of
the
air
to
whip
you
so
that
the
common
configuration
and
operation
will
be
defined
under
the
the
young
on
the
different
types
of
the
eric
who
live
here,
because
the
different
era
to
live
here
different
types
of
the
air
to
VPN
there
may
be
different
configuration
and
is
the
operation
Peter.
So
this
included
the
RDP
based
repeatable
us
and
VP
RS,
and
the
PGP
based
the
bleep
ers
and
the
PDP
auto-discovery
based
we
be
ours,
and
also
that
is
a
state
hehe,
p
SS
and
the
static
of
a
PRS.
I
So
this
is
the
the
haps
of
the
art
will
appear
will
be
taking
into
account.
So
now
the
one
remaining
usual
use
the
CCC.
That
means
the
security
cross
connect,
because
this
is
a
user
to
some
extent
use
the
different
from
the
common
reference
model
of
air
to
VPN.
But
in
fact
there
are
several
implementation,
so
we
will
discuss
later
this
you
use
as
a
vendor
specific
young
model
or
try
to
defend
in
the
working
group.
Okay,
yes,
young
definition.
I
We
will
take
into
account
the
configuration
data
and
the
operation
data
and
in
the
RPC
and
the
notification
so
that
a
user
completed
the
young
definition
for
the
models,
not
a
universe,
cope
I,
think
that
the
art
will
appear
because
the
number
of
the
types
of
the
art
will
appear
so
I
think
this.
We
can
be
a
future
draft
so
and
also
that
uses
some
work
related
with
the
arrow
to
VPN
or
the
silverware
is
real.
I
That
means
that
the
OEM
they
have
the
VCC
VPN
and
the
trees,
and
also
that
I'm
pr's
tpo
em
for
civil
aware,
especially
regarding
a
related
with
the
performance
monitoring,
so
this
Walker
will
not
be
taking
into
account
by
this
team.
So
this
work
we
will
leave
to
the
OEM
experts
to
do
the
young
model,
work,
yeah,
okay,
next,
okay,
and
also
when
we
prayed
until
the
worker
in
the
past
working
group.
But
in
fact
the
art
will
appear,
has
much
relations
with
other
work.
I
So
we
will
do
some
the
during
that
the
cause
of
the
desire.
We
will
do
the
liaison
work.
This
included
the
lay
down
with
the
best
working
group,
because
the
PDP
based
vp
RS
and
the
PCPD
auto-discovery
based
the
Riviera's,
will
be
taking
into
account.
So
the
liaison
work
will
be
done
between
the
past
working
group
under
the
best
working
group
and
also
that
using
the
last
meeting
stillwater
reminders
that
they
are
to
repeal
the
existing
are
to
VP
amoeba.
I
Ok-
and
this
is
the
design
work
plan,
so
the
first
we
will
discuss
the
architecture
design
because
we,
firstly,
we
try
to
defend
the
architecture
of
the
air
tubes
in
young
models.
So
we
try
to
finish
this
work
in
the
middle
of
the
May
this
year,
so
this
will
define
this
is
using
the
work.
I
include
sakova
definition
and
also
the
component
division,
because
this
is
because
we
try
to
do
you
truly
divided
the
art
we
be
a
young
model
design
walk
into
the
different
apart,
so
there
is
a
they
say
user
into
account.
I
This
included
the
common
component
and
also
the
different
there
may
be
different
aircraft
for
the
different
art
Olivia
young
model.
So
this
is
the
tool
hire
two
walks
we'll
be
taking
into
account
into
this
face
after
the
face.
So
we
will
prepare
the
plan
for
the
concrete
data
model
design.
So
the
concrete
data
model
design
included
a
common
component
under
me
under
the
super
Civic
pass
of
the
different
types
of
the
air
to
Vivian's.
I
So
I
think
this
is
we
try
to
pro
whole
the
initial
draft
before
the
next
idea
of
a
meeting,
but
according
to
the
pasta
experiments.
So
the
work
is
a
very
challenging
so
that
we
will
propel
the
work
plan
after
the
first
phase
architecture
design,
so
that
a
disgusted
in
the
working
group
to
see
if
it
can
comply
with
the
charter
of
the
air
to
pass
a
working
group.
Ok,
next
one!
Ok!
I
A
I
A
B
E
Wikis
that
we
made
reference
to
at
the
beginning
of
the
meeting
I
know
that
is
there's
quite
a
few
drafts
one
drafts
on
there
and
I'm
not
sure
which
ones
overlap.
It
looks
like
there's
some
redundancy,
because
some
of
them
are
listed
is
l2vpn
and
some
were
listed
as
pals,
but
if
you
could
reconcile
those
as
a
result
of,
though
the
work
and
the
working
methods
I
appreciate
it.
Okay.
I
Yeah,
okay,
definite,
okay,
I
think
of
data.
We
will
that
in
the
beginning,
/
that
a
user
some
days,
the
young
model
coordination.
What
a
message
that
has
been
proposed
that
you
know
I
ATF:
we
will
comply
with
this
method.
Okay
to
earn
our
were
walking;
okay,
that's
one!
Okay,
so
I
saying
this!
So
you
finished
so
I
hope
that
we
get
more
comments
to
help
her
to
help
us
to
defend
your
models
effectively.
F
A
E
I
I
missed
this
part,
in
fact,
the
univer
last
the
meeting
the
day
proper,
the
similar
opinion,
so
that
I
used
because
there
for
the
VP
ours
for
the
some
the
pws.
There
are
some
special
contribution
regarding
the
AC
attacher
security,
so
that
especially
for
repairs,
there's
a
lot
of
the
villa
operations.
So
this
is
the
this
Walker
will
be
defined.
This
unit
is
your
model
yeah.
E
Yeah
Nick
del
regno
verizon
a
couple
things
one
be
happy
to
help
out
with
the
definition
source
engine
email
join
the
list.
The
second
is
don't
think
we
should
include
vinterberg
vendor
implementations,
vendor
specific
implementations
like
ccc.
In
our
definitions,
we
should
be
in
the
business
of
writing
Hank
models
for
the
RFC's
that
we
produce.
I
G
B
B
D
J
Episode,
my
name
is
mommy
hookah,
death
and
and
presenting
this
draft
on
behalf
of
all
theaters,
so
the
precise
will
be
sure
because
you,
this
working
group
and
the
pharma
group
that
did
already
done
the
work
to
identify
the
problem
space
and
the
gap
analysis
that
we
are
addressing
in
this
draft.
So
in
the
next
life
pieces,
as
you
see
in
there
are
RFC
74
39,
it
won't
fight.
J
Why
we
have
it
right
now,
ipv4
specific,
so
this
draft
is
trying
to
fix
this
void.
I
did
pride
so
exactly
two
simple
updates,
as
you
can
see
in
the
next
slide,
please
so
the
first
one
is
just
an
update
to
the
BGP
another
I
attribute.
That
is,
don't
define
it
in
that
section.
So,
in
the
current
version
of
the
RFC,
the
PIAA,
this
is
all
the
uncoated
for
octet
and
the
new.
J
J
So
the
drug
is
suggesting
to
define
a
new
type,
which
would
be
an
able
to
include
128
bits
rather
than
the
one
that
is
today
only
capable
to
go
to
d
2
230
it's
so
there
will
be
a
the
river
section
that
we
need
to
be
updated
in
the
in
the
RC,
60s
74.
Just
to
say
that's
so,
as
you
can
see
them
in
the
next
slide,
please
it
this
drop.
Is
it
just
feeling
something
with
that
has
been
unified
in
general,
this
RFC?