►
From YouTube: IETF96-BFD-20160722-1220
Description
BFD meeting session at IETF96
2016/07/22 1220
A
C
A
E
Session,
it's.
A
D
B
D
F
D
E
C
D
D
Okay,
so
let's
go
through
our
working
group
status.
This
is
our
typical
note.
Well
at
this
point,
do
you
see
this
of
times
this
week?
I
will
also
note
that
the
working
group
chairs
are
making
a
push
to
try
to
get
more
reviewers
for
our
various
documents
get
people
involved
earlier.
So
if
you
have
some
interest
in
becoming
more
involved
in
the
active
editing
process
of
ITF,
please
let
us
know
and
we'll
help
you
do
that
so
in
terms
of
status,
we
actually
hit
a
nice
milestone.
D
D
D
We've
spoken
to
over
all
goes
our
ad
about
d
chartering
this
work
and
if
the
document
ever
becomes
interesting
again,
we
can
obviously
move
it
forward,
but
as
part
of
our
next
charter
update,
we
will
likely
end
up
removing
this
from
the
Charter.
Another
piece
of
somewhat
lingering
work
is
the
FD
multi-point.
It's
certainly
ready
for
working
group
last
call,
and
we
just
mostly
need
to
move
forward
in
the
publication
and
review
requests.
D
What
I'll
be
suggesting
at
this
point
on
the
mailing
list
is
that
we
look
for
no
one
or
two
reviewers
for
things
to
just
go
through
the
document
see
how
it
is
now
potentially
sign
them
up.
Is
the
document
Shepherd
and
actually
proceed
with
working
group
last
call
that
we
are
aware
of
at
least
one
implementation
of
the
base
BFD
multi-point
document?
The
active
tale
document
is
used
by
trill,
so
it
at
some
point
is
going
to
become
a
normative
requirement
for
them
as
well.
D
One
of
the
things
that's
sort
of
an
interesting
lingering
point
for
our
work
ethic
ation,
is
that
the
curtain
group
prior
to
their
closure
had
provided
a
serious
recommendations
for
various
protocols
to
actually
close
off
security
issues
within
those
protocols.
Their
recommendations
to
BFD
resulted
in
two
documents.
The
first
one
was
a
generic
King,
a
piece
of
infrastructure
that
mostly
provided
for
the
ability
to
do
key
agility
know
be
able
to
do
rollovers
of
keys
mid-session.
This
is
something
that
the
FD
cannot
currently
do
and
honestly,
given
how
fast
BF
he
is
done.
D
Ki
rollover
is
actually
a
very
difficult
thing
to
do.
That
said,
we
have
been
given
the
impression
from
multiple
implementers
that
VFD
security
is
not
currently
a
popular
feature.
It
is
deployed.
It
is
used
by
some
people,
but
it
is
not
heavily
used
in
the
level
of
security
is
not
considered
very
high.
At
the
moment,
the
secondary
component
that
was
recommended
from
them
was
a
profile
that
loud
for
sha
to
H
Mac
to
be
used
as
part
of
VFD.
D
Both
of
these
pieces
of
work
are
seem
to
be
well
reviewed,
pretty
solid,
but
at
the
moment,
just
sort
of
low
interest.
No
for
deployment
upon
discussion
with
evaro.
One
path
we
are
considering
to
do
take
this
stuff.
Ford
is
that
we
may
actually
want
this
stuff
at
a
later
point
in
time,
and
it
should
actually
seem
to
have
gotten
ITF
review.
So
our
inclination
is
to
look
for
document.
D
Okay,
next
life,
so
future
work.
Our
charter
for
some
strange
reason
did
actually
having
at
it
will
need
to
close
that
off
in
terms
of
clearing
stuff
out.
We
also
have
no
some
other
know
things
that
are
making
use
of
the
authentication
a
couple
of
things
we've
seen
before,
and
at
least
we'll
get
an
update
on
this
session.
We
have
a
small
amount
of
lingering
openwork
beyond
you'll,
be
items
that
are
under
the
authentication
bucket
and
past
that
point
we
don't
currently
have
anything
that
is
interesting.
This
happens
periodically
with
BF
d.
D
Once
we
finish
the
initial
work
that
resulted
in
58
8350
884,
we
actually
went
into
hibernation
for
a
while.
The
purpose
of
the
DFT
group
beyond
that
point
is
really
to
just
simply
provide
guidance
to
other
working
groups
that
want
to
use
VFD
technologies
and
we
keep
on
seeing
novel
things
show
up
just
this
example:
the
VX
LAN
work
that
was
being
presented
that
vo
3,
the
EVP
n
work
that
will
be
sitting
today,
we're
not
seeing
work
that
is
resulting
in
changes
to
the
base
protocol.
D
This
is
more
deployment
considerations
and
you
know
that's
having
the
FDA's
a
review
working
group
is
a
reasonable
thing
to
do
the
course.
If
we
actually
have
novel
work
for
some
strange
reason
come
up,
then
the
working
group
can
become
active
again,
don't
commit
so
in
terms
of
presentations.
Today,
we're
going
to
have
no
quick
review
of
the
VFD
yang
module,
we're
going
to
be
going
through
the
current
BF
d
authentication
module
for
being
able
to
do
optimized
authentication,
the
FT
stability
is
going
to
get
a
very
brief
update,
which
I'll
be
presenting.
D
E
B
E
The
authors
of
the
draft
are
on
that
first
page
and
I
believe
only
Greg
is
the
other
person
here,
or
maybe
zero
I'm,
not
sure
anyway,
next
slide.
So
when
we
last
met
in
Yokohama,
there
were
a
few
issues
which
were
described
about
eight
months
ago.
The
issues
which
we,
which
we
discussed
then
had
been
address,
or
mostly
addressed
in
the
latest
revision
of
the
draft.
E
E
G
E
That
kind
of
stuff-
and
you
know,
as
we
have
gift
as
we
have
the
young
model
for
more
technologies
on
which
we
have
D,
runs
then
we'd
be
having
those
parallel
models
and
the
other
thing
we
had
was
for
BFD
over
mpls
t.
You
know
you
typically
configure
BF,
gilbert,
eternal
ekiti
on
all
can
have
multiple
LSPs.
There
was
no
way
for
us
to
represent
that
purely
in
VFD,
except
by
we,
by
importing
lots
of
things
which
were
defining
the
in
the
t
model.
So
we
decided
to
augment
the
T
model.
E
The
other
change
we
did
was
regarding
the
authentication
we
had
our
own
kind
of
authentication
scheme
and
now
we're
using
the
keychain,
which
is
in
the
routing
workgroup
draft,
and
so
one
of
the
other
issues
we
had
in
terms
of
making
progress
is
some
of
the
modules
we
depend
on.
There's
been
lots
of
change.
I
mean
lots
of
churn.
If
you
know
you've
been
attending,
you
know
whether
it's
net
mod
or
rotting
workgroup,
and
all
that
you
know
they
used
to
be
routing
instance.
E
We
decided
in
dollars
to
be
rocked
to
be
grf,
centric
and
then
about
other
node
3
4
months
ago.
The
routing
instance
was
gone.
It's
been
replaced
by
network
instance,
which
I'll
be
talking
about
soon,
and
one
of
the
other
things
was
that
the
yang
doll
addition
to
landis
and
the
website
that
bueno
ogron's
used
to
be
using
using
P,
hang
on
Nia's
validation
tool.
E
Pn
is
good
for
certain
things,
but
it
doesn't
detect
certain
errors
like
xpath
errors,
and
one
of
the
things
which
was
done
in
the
hackathon
and
prior
to
that
is
karl,
has
change
his
tool
and
brought
you
to
use
county
compiler,
which
detects
xpath
errors.
Xpath
errors,
I
think,
are
the
most
common
things.
When
you
do
one
statements
or
most
statements
that
you
know
p
young
does
not
attack.
You
know
you
easily
forget
about
dot
or
have
an
extra
one.
So
we
had
couple
things
like
that
which
we
had
to
fix
next
slide,
please.
E
So
what
is
currently
not
covered,
so
BFD
/
v,
CC
v
is
not
covered.
I
think
that's
a
neck
thing
we
have
to
look
into
the
LTV
jian-yang
model
is
now
is
now
or
will
soon
be
a
war
group
document.
There
had
been
a
fair
amount
of
churn
I.
Think
the
yang
doctors
had
bunch
of
comments
so
now
that
that
kind
of
firming
up
I.
E
Look
but
having
bfg
Forbes
pseudo
wires.
The
second
thing
was
behavior
/
mpls-tp.
When
the
design
team
was
meeting,
we
initially
said
that
we
would
not
cover
mpls-tp
and
because
that
seemed
to
be
low
priority
on
the
t
model,
but
maybe
Greg
you
can
I
believe
you
are
working
on
a
model
for
mpls-tp.
Now.
E
F
E
E
There's
no,
our
pcs
right
now
in
the
model
and
the
question
to
the
room
is
my
take,
is
I
mean
so
the
usual
stuff
people
talk
about
is
clearing
of
counters
I,
mean
statistics
and
stuff
like
that,
but
I've
heard
others
say
that
you
know
we
should
not
be
doing
that
because
then
you
screw
up
your
controller
and
the
controller
is
doing
the
Delta
for
you.
So
that's
an
open
item.
Yeah.
E
Not
just
a
BFD
decision,
it's
across
the
routing
routing
area,
whether
what's
the
what's
the
guidelines
for
this,
the.
H
E
E
D
D
E
The
only
thing
we
were
ignoring
at
the
beginning
of
the
design
team
is
seamless
VFD,
because
we
said
you
know
we're
only
talking,
RFC's
well,
seamless,
BFD,
isn't
you
know,
there's
a
bunch
of
our
seas
of
Jeff
said
my
you
know
immediate
take
on
this
is
that
we
don't
need
any
change
in
the
model,
because
there
is
no
change
to
the
data
model
itself,
but
that's
something
else
which
we
can
take
on
the
list
and
is
could
be
for
further
study
next
slide.
Please
open
issues.
E
So
after
the
routing
instance,
you
know
compiled
failures,
I
took
to
augmenting
network
instance,
but
this
week
I've
had
some
discussions
with
the
net
mud,
guys
and
the
what
the
routing
config
model
and
I
think
what
the
net
mod
my
team
is
pushing.
Is
that
so
one
discussions
we
had
at
the
beginning
was:
you
know
some
implementations
of
some
implementations
of
VFD
rgr
fo,
where
you
know
there
are
you
know,
there's
routers
and
about
DRS
you
configure
be
obtained
brf,
others
might
be
simpler
nodes
which
are
maybe
transport
nodes
and
there's
no
vrs.
E
So
I
think
that
question
has
arise
for
other
protocols
and
what
those
Network
guys
are
pushing
for
right
now.
Is
you
define
VFD,
you
don't
talk,
mountain
your
routing
instance
or
networking
instance,
but
you
have
it
schema
mounted
based
on
your
implementation,
whether
it's
at
the
top
level
or
in
the
in
the
vrf.
E
We've
had
ongoing
discussions.
There
was
discussions
yesterday
in
the
line
work
group
as
there's
the
line
connectionist
model,
which
was
presented
again
yesterday.
I
think
there's
some
maybe
mismatch
in
the
views
as
to
how
the
OEM
protocols
use
lime.
I
think
the
line
view
is
that
protocols
like
dfv
should
be
augmenting.
The
line
model
I
think
those
OEM
protocols
might
have
the
reverse
view.
E
We
had
some
discussions
yesterday,
Greg,
okay,
Michael
I,
don't
know
if
you
guys
had
further
discussions
up,
but
Michael
is
going
to
be
present,
so
we
can
have
that
discussion
when
Michael
presents
alarm,
connectionless
model
and
I.
Think
it's
something
we'll
have
to
discuss
further.
Oh,
come
the
list.
E
D
Thank
you,
okay,
and
the
issues
that
we're
running
into
in
terms
of
know
the
standard
that
odd
models,
the
scheme
amount.
That
sort
of
thing
is
issues
that
every
single
ITF
group
is
facing
at
the
moment.
So
this
is
not
anything,
that's
blocking
our
specific
work.
This
is
us
just
dealing
with
what
the
current
state
of
the
matter
is.
I
D
Okay,
so
the
the
purpose
of
the
BFD
optimizing
authentication
feature
is
to
try
to
take
care
of
at
least
one
of
the
obstacles
for
PFD
authentication
being
deployed
and
that's,
namely,
if
you're
using
any
sort
of
meticulous
authentication,
the
amount
of
C
fuel
is
burned
by
the
authentication
operation
tends
to
be
a
little
bit
more
than
what
you
can
do
in
most
of
your
line
cards
and
since
we're
operating
a
resource-constrained
environment.
Looking
for
more
clever
ways
to
make
use
of
the
resources
without
burning
them
strictly
for
authentication
is
something
that's
under
evaluation.
D
The
dockyard
right
now
really
has
had
no
substantive
changes,
since
the
last
time
was
presented.
The
security
implications
about
leaving
it
in
this
form,
where
we're
not
always
working
through
assigned
packets,
was
the
main
consideration
we
had
to
work
through.
The
request
from
the
security
directors
was
that
we
should
actually
go
talk
to
the
sa
AG
people
to
see
if
they
had.
Any
considerations
looks
like
so
we
had
breakfast
with
some
of
them
on
Monday
morning
and
the
commentary
that
we
had
received.
You
know
from
the
two
that
showed
up
was
no
sorry
prospective.
D
If
the
proposed
solution,
you
know
again,
we
authenticate
on
state
changes
only
you
know
we're
up.
You
know
once
every
second
tool
send
a
dedicated
packet.
That
will
not
necessarily
be
a
state
change
and
it's
mostly
to
just
make
sure
that
we
don't
start
subject
to
man
in
the
middle
types
of
tags
next
slide
and
the
breakfast
that
we
had
had
Alan
decock
and
Randy
Bush
there
and
for
the
most
part
they
seem
to
be
generally
okay
with
the
idea,
especially
for
the
problem
space
that
BFD
his
tendon
is
attempting
to
cause
a
handle.
H
Actually,
the
question
I
send
the
commission
to
the
list
before
the
meeting,
so
if
we
are
sending
single
authenticated
packet
are
among,
their
stream
are
not
authenticated.
Packets
I
can
imagine,
for
example,
are
in
hardware-assisted
implementation.
I
authenticated
packet
will
be
shipped
for
check
to
their
control
plane.
So
in
the
meantime,
if
we
have
very
short
interval
are
non
authenticated
packets
who
get
in
and
they
will
be
resetting
the
dead
timer.
D
Possibly
so
and
I
think
the
piece
of
the
procedure
that
does
need
to
be
spelled
out
is
that
in
the
window,
where
you're
expecting
an
FN
decade,
attacker,
they
show
up,
there
needs
to
be
potentially
an
extended
window
longer
than
the
detection
multiplier
to
say
the
sessions
actually
supposed
to
be
up
or
down.
Yes,.
H
D
And
as
we
see
here,
they
were
generally
okay.
With
this
idea
of
leaving
things
know
partially
unauthenticated
Ellen
did
add
an
additional
suggestion
that
we
use
something
other
than
monotonically
increasing
sequence
numbers.
This
allows
us
to
avoid
some
of
the
predictive
attacks
that
we
have
so
based
on.
Having
some
sort
of
key
algorithm
would
be
used,
or
the
sequence
number
is
predictable
by
each
side.
That
has
the
key
without
allowing
the
item
in
the
middle
just
simply
spoof
it
out.
So
in
cases
where
a
signing
operation
is
that
not
actually
being
done?
D
So,
in
terms
of
the
next
steps,
the
draft
has
been
stable
for
a
while.
Now
this
is
a
useful
feature.
This
does
help
us
with
our
scenarios
where
we
went
to
play
stronger,
no
cryptographic
mechanisms,
one
way
to
actually
approach
this
is
adopted
then,
along
with
the
other
authentication
work
tried
to
do
it
as
a
batch.
D
No,
this
is
again
from
the
working
group
status,
not
something
that
is
traditionally
gotten
a
lot
of
interest
in
the
work
working
group
and
interest
among
operators
is,
you
know,
modestly
low,
but
the
problem
of
security
problem
security
stuff
is
the
interest
for
it
always
happens
long
after
the
fact,
if
you
don't
have
the
feature
and
a
security
issue
becomes
known,
having
a
solution
in
RFC
somewhere
is
actually
a
useful
thing.
So
the
authors
are
looking
to
have
the
working
group
adopt
the
document
and
potentially
last
call
it
relatively
soon.
Thereafter,
okay,.
D
So
any
further
comment
on
this
proposal.
We
will
be
taking
adoption
to
the
email
list,
just
as
a
attempt
to
get
daggett
know
some
generally
working
group
and
just
show
of
hands.
No
don't
need
a
home
for
this.
You
know
who
here
thinks
this
is
actually
work,
that
the
working
group
should
take
up
yeah.
D
F
D
So
again,
I'm
playing
virtue
mahesh,
no
just
as
a
brief
update
on
the
DFT
stability
show
of
the
authors.
So
again,
this
was
a
attempt
to
make
use
of
VFD
to
measure
the
stability
of
a
link.
This
is
not
being
used
to
actually
take
a
session
down.
This
is
simply
where
you
actually
want
to
use
an
existing
stream
session.
That's
going
across
the
link
to
try
to
detect
periodic
disruptions,
and
you
know
the
main
purpose
is
to
see
or
what,
if
there's,
some
form
of
loss.
D
You
know
this
was
again
take
advantage
of
the
fact
that
we
have
sequence
numbers
going
across
the
authentication
and
watch
for
drops
in
the
sequence
number
packets
just
to
see
if
things
are
there
and
that
way,
if
you
happen
to
see
a
loss
of
a
certain
number
of
packets.
No,
this
gives
you
some
level
0
statistics.
The
general
idea
is
that
we
want
to
use
sequence
numbers.
D
We
could
use
cryptographic
mechanisms,
but,
honestly,
you
know
that's
just
going
to
slow
things
down,
so
part
of
the
proposal
does
include
it
in
the
bottom
point
here:
a
no
auth
tlv.
This
is
simply
there
to
make
use
of
the
existing
cryptographic.
General
header
looks
like
so
they've,
mostly
just
done
some
text
publishing.
They
have
done
some
cleanup
in
terms
of
reusing
the
no
off
and
try
ensured
among
the
optimizing
authentication
field
and
the
packet
delays.
Nope
then
put
as
a
implementation
details
specifically
how
it's
actually
done
beyond
that
is
out.
D
The
scope
of
the
document
draft
is
stable
and
they're.
Looking
for
working
group,
adoption
now
give
this
well.
This
actually
fits
in
the
authentication
bucket
of
work,
since
it
actually
touches
that
feature.
This
really
is
a
standalone
featured
for
you
know
trying
to
actually
detect
pack
of
loss
across
the
FD
session.
D
I
D
E
What
are
you
presenting
first
I'm
Selig.
H
H
Non-Standard,
so
with
interaction
well
with
cooperation
from
I
Tripoli
ITF
managed
to
resolve
that
how
to
run
on
the
sub
second
intervals
on
the
individual
concision
thinks
so
effectively
creating
a
set
of
identical
sessions
with
different
discriminators,
and
usually
you
run
it
with
the
same
time
interval.
But
we
realized
that
the
solution
proposed
in
72
month
30
does
not
address.
H
H
Multicast
address
or
use
a
local
broadcast
our
next.
What
okay
and
if
we
are
in
IP
MPLS
Network,
then
encapsulation
of
VFD
control
packet
could
be
either
with
ap
header
or
with
an
or
ACH
encapsulation.
So
if
its
IP
header
then
associate
the
channel
will
be
used
to
indicate
whether
ipv4,
ipv6
and
IP
address
will
be
destination
so
to
prevent
the
routing
out.
H
H
E
E
H
H
H
I
H
K
The
link
aggregation
standard
has
been
updated
recently.
If
you
incorporate
these
this
multi
chassis,
while
it's
not
called
multi
shot,
it's
called
dual
redundant.
It's
called
distributed
resilient
network
interface,
not
your
redundant
anything
and,
and
this
effectively
standardizes
a
the
dual
home
case,
and
it
also
standardizes
two
nodes
on
each
side
anymore.
H
K
So
I
would
suggest
that,
because
of
the
work
having
been
done
in
eight,
oh
two
dot
one
to
handle
this
multihomed
case,
those
considerations
will
have
been
analyzed
and
may
have
been
changed
fundamentally
in
the
link,
aggregation
and
so
I
think
it
would
be
necessary
to
coordinate
this
with
eight
oh
two
dot
one
and
not
to
work
independently
here.
So
I
would
request
that
you
and
send
an
email
to
the
I
Triple
E
ITF
coordination
group
and
to
bring
their
attention
to
this
draft
and
request
that
the
data
to
one
takes
a
look
at
it.
K
H
K
D
Where
eyeballs
are
good?
Okay?
So,
since
we're
looking
for
a
potential
adoption,
this
working
group
again
we'll
take
it
to
the
mailing
list.
But
could
we
have
a
show
of
hands
of
people
who
think
this
is
work
that
is
worth
adopting.
F
L
D
F
F
D
H
H
H
H
Architectural
elements
construct,
so
you
need
to
monitor
a
multi-point
to
pour
into
point-to-multipoint
and
it
probably
even
multi
point
to
multi-point
LSPs.
So
thus
there
is
a
combination
of
BFD
and
the
protocols
to
be
used,
for
example,
in
the
multiple
into
point.
It's
a
point
to
point
B
of
d,
&
n
point-to-multipoint,
it
obviously
the
BFD
for
a
multi-point
network.
H
But
in
addition
to
that,
the
question
is
so
how
you
identify
the
scope
of
your
BF
d
session,
and
originally
the
proposal
was
that
to
use
the
same
encapsulation,
of
course,
with
a
gal
label
as
being
used
by
the
corresponding
traffic
type.
That
creates
a
lot
of
complexity
and
probably
unnecessary
arm
overhead.
So
thus
we
propose
to
use
another
alternative.
It
can
go
to
the
next
slide
place
so
which
uses
ACH
encapsulation
and
provides
the
context
in
a
POV
when
she
follows
the
view
control
packet
and
that
can
be
done
either
for.
H
D
Okay,
so
we
do
have
jabber
comments
from
episod
when
it
first
ask
for
some
commentary
on
both
alternatives,
which
you
proceed
to
go
into
a
little
bit.
His
second
bit
of
commentary
was
that
he
personally
thinks
that
they're
nerds
for
the
first
form
of
encapsulation,
I
used
the
l
to
header
inside
the
gal,
gash
payload
anything
for
the
percent.
D
H
J
E
So
I
have
some
have
some
questions
can
come
and
some
of
them
I
put
on
the
list
already,
and
we
spoke
about
in
the
best
k
meeting.
So
I
think
the
draft
doesn't
explain
very
well
in
the
alternative
proposal
that
there
is
no
IP
header
or
maybe
it
does,
but
not
in
every
place.
That's
one
thing
I
think,
also
to
clarify
that
the
LS
beeping
is
needed
only
when
in
the
second
proposal
and
not
looking
the
first
right.
Yes,
thank
you
for
the.
H
Yes,
the
proposal
currently
assumes
that
we're
using
ACH
encapsulation
so
the
thus
there
is
no
IP
source
address
in
VFD
and
part
of
the
bootstrapping,
whether
it's
with
ls15
or
some
other
means
because,
for
example,
was
interesting
discussion
at
the
best.
They
have
a
proposal
to
use
vgp
for
the
bootstrapping
to
distribute
discriminator
through
vgp
yeah.
E
H
D
Be
where
is
the
conclusion
that
I'd
come
to
for
my
reading?
Is
that
the
encapsulation
issues
no
push
it
a
little
bit
more
for
best
in
terms
of
how
this
is
meant
to
be
utilized,
whereas
the
has
a
review
requirement
from
mpls
there's
not
much
in
the
way
of
actual
changes
to
BFD.
It's
enough,
that's
correct,
so
this
may
be.
One
of
those
cases
were
best
may
be
the
primary
first
home
for
the
document
making
it
review
for
the
other
two
working
groups
now.
D
This
sounds
like
a
conversation
should
get
kick
off
with
the
chairs
of
all
three
groups:
okay,
okay,
thank
you.
Click
show
who
has
the
blue
sheets
and
have
you
not
signed
the
blue
sheets?
We
have
a
couple
people
in
the
back
corner,
so
whoever
has
the
sheets
if
you
could
forward
them
back.
D
G
Okay,
how
do
I
write?
My
name
is
Mike
home
from
Holly
today,
I
want
to
introduce
this
I'm
connect,
Solis
OEM,
llamo
geo,
brave
late,
since
we
have
in
you.
Can
you
come
a
meeting,
be
talking
with
land
group
and
receive
a
lot
of
comments
as
we
splinted
his
model
into
space
is
a
basic
model
into
to
model,
14
connector
and
connect
or
ended,
and
that
another
a
glaze
and
we
have
do
a
lot
of
work
to
build.
This
is
a
connected,
som,
Yamato
and
present
ate
a
presentation.
G
G
G
G
Just
you
can
look
at
this
motor
design
in
the
back
up
a
size,
exactly
okay,
we
have
introduced
this
so
connect
kinetically,
so
mi
amor,
the
update
in
la
working
group
and
all
some
online
and
offline
discussion.
So
in
this
meeting
we
cling
ups,
a
passive
passive
performance
measurement
attribute
eames
from
the
connectors
over
them.
Your
motive
and
no
way
closes
passively.
I'm
a
young
opening
issue.
So
the
next
step
we
will
cover
adoptions.
It's
our
connector
ram,
your
module
and
working
group.
Okay
stacy
is
a
basic
snod,
connectors
OEM,
your
modules
update.
D
G
In
this
line,
basic
mall
do
this
in
this
really
use
a
trust
case.
You
trust
his
notes
to
dig
two
graves
different
type.
Our
difference
test
ponds,
location
list
type.
So
it's
a
tweak
design.
This
model
we
both
consider
realtor
and
equipment,
said
Bailey
taste,
so
this
trance
case
note
as
real
extensible.
G
So
technology
by
CFA
program
can
choose
this
relative
blog
extending
all
just
can
aid
his
rental
case
in
Tuesday's
test
pointer,
since
we
have
already
up
with
since
the
land
connection
module
is
extending
by
the
ICF
polity:
module
okay,
so
it
build
is
a
relationship
with
different
the
techni
text,
s
upon
the
location,
so
technologically
Miyamoto
can
aid
I'll
insert
as
relative
case
in
Tuesday's
test
pond
twice.
So
this
hope
it
was
so
there's
a
whole
test.
Pondle
locate
locations
laser.
G
E
It's
a
something
which
we
discussed
yesterday
and
I
think
which
greg
alluded
to
also
in
the
line
meeting
yesterday.
So
now,
there's
certain
constructs
like
tea
tunnels,
where
you
know
the
model
is
fairly
complex
and
on
the
config
whys
and
the
operational
wise.
We
really
do
not
want
to
duplicate
that
tree
now.
Maybe
we've
scheme
a
lot.
We
can
do
that,
come
not
sure.
But
how
do
you
can
you
say?
Do
you
I
mean
I
think
we
spoke
about
leaf
ref
or
something.
E
G
Think,
okay
should
be
done.
Okay,
if,
because
there's
a
tag,
tea,
tea,
tea
party
network,
perhaps
the
sea
in
the
basic
multi,
we
don't
have
a
design.
A
relative
test
point
pattern
for
tea:
you
can
either
relativity's
location
and
the
user
some
leaf
reference
reference
to
that
eased
party
is
a
llamar
do
so
it
it
kinda
Hera
takes
a
team,
adios
attribute
and
unions,
difficult
difficult
attribute
and
also
can
provide
this
out.
You
know
the
lambda
relationship,
okay,
so.
G
G
You
may
wish
to
us:
I
mean
you
may
define
disease.
Rpc
blogs
can
provide
on-demand
on
the
money
method,
but
also
we
use
this
configuration
blocks,
but
you
can
just
a
twosome.
You
know
practice
attribute
into
the
relative
on
your
point
and
so
it
can
use
or
both
a'practice
upon
the
undermanned
master.
Okay,.
D
So
this
is
just
that
at
some
point
you
may
be
making
recommendations
to
the
VFD
yang
design
team
for
some
our
pcs
to
be
at.
Is
that
correct,
yep,
okay-
and
my
last
comment-
is
that
for
the
seamless
BFD
case,
discovery
of
the
discriminators,
the
main
thing
were
you
thinking
that
this
should
be
discovered
through
the
igp
or
since
your
main,
in
direction.
Right
now
is
through
the
topology
model.
Would
you
be
expecting
the
discriminator
to
be
in
the
technology
man,
okay,
sit
sit.
Why.
G
We
alcaman
is
the
topology
model
since
days
and
a
wake
and
sees
a
lamb.
Llamada
is
a
kind
of
OEM
topology,
and
the
polish
model
Kim
shows
a
different
test
point
a
relationship,
and
we
also
ate
some
relative
test
Ponte
in
this
module
to
complete
this
relationship.
Another
thing
is
just
a
way
it
is
just
well
I'm
going
to
do
is
to
you
provide.
G
This
is
test
point
allocations,
relationship
and
if
it
won't
you,
you
know,
if
want
to
use
a
UPS
in
some
technology
networking
instance,
perhaps
some
job
in
Dubai
small
do
design
team
can
use
peer
among
our
NASA
message
to
mounted
lime
module
into
his
network
instance
module.
Okay,
taxes,
okay,
so
it
now
too
I
mean
is
naturally
effects.
Networking
instance
it.
Now
we
have
a
lot
of
master
to
use
this
to
you,
use
one
module
into
another
blocks
unit.
Okay
is.