►
From YouTube: IETF106-DETNET-20191121-1740
Description
DETNET meeting session at IETF106
2019/11/21 1740
https://datatracker.ietf.org/meeting/106/proceedings/
A
Since
this
is
a
independent
session,
we're
gonna
remind
folks
of
a
few
administrative
things.
First
is
the
note?
Well,
this
is
the
same
as
what
we
talked
about
the
last
session.
Please
recall
that
everything
you
say
here
is
governed
by
our
contribution
rules
and
anything
you
say
becomes
part
of
our
permanent
record.
If
you're
not
familiar
with
our
note.
Well,
it's
the
end
of
the
week
right.
You
should
know
that
by
this
point,
but
please
take
a
look
at
the
URL.
A
Sorry
I
have
to
I
have
to
remember
to
eat
the
mic.
Sorry
about
that,
so
we
are
streaming
audio
and
video
as
well
as
recording.
Please
speak
your
name
at
the
mic
when
you
have
a
comment,
also
join
in
our
joint
minute.
Taking
the
URL
is
here:
it's
also
ascent
to
the
list,
so
it
can
be
easy
to
click
on.
We
are
doing
jabber
I'm
on
jabber,
we
do
have
remote
participants
and
they
should
feel
free
to
join
the
mic,
the
mic
line.
A
If
they
have
comments-
and
we
do
have
one
remote
speaker-
the
agenda
is
slightly
modified
from
what
we
talked
about
before.
Specifically,
that
item
number
seven
we've
already
covered
in
a
previous
session
also
number
ten-
is
a
late
addition.
So
I'm
not
going
to
go
over
the
rest
of
the
typical
administrative
things.
We
did
that
in
last
session.
I
think
we're
just
going
to
jump
right
in
so
I.
Think
Andy
is
next.
C
C
So
here's
a
background
of
this
document
so
as
we
know
that
to
date,
the
working
group,
many
in
focus
on
the
data
plan
and
become
modules,
so
I
think
it's
time
to
consider
about
this
control
plane
protocol
aspect,
as
we
know
that,
as
we
know
that
the
current
architectural
document
has
already
discussed
some
control
and
controller
plan
aspects
in
the
section
and
404
font
as
according
to
the
definition
in
RC
17
for
2026,
the
controller
plan
include
actually
includes
a
control
and
the
management
plans.
C
C
So
this
statement
also
discussed
other
control
plan
issues
that
dedicated
to
dead
net,
for
example,
how
to
support
explicit
paths
and
how
to
do
the
source,
a
reservation
to
support
pre-health
functions
and
how
to
you
know
to,
and
also
talk
about
how
to
integration
with
existing
protocols.
For
example,
the
lung
I,
sir,
am
furious.
Long
is
our
IP
and
also
can
how
to
apply
this
to
the
I'm
here.
Sunday
service
six.
C
So
this
document
also
to
describe
some
aspect
about
the
management
plan
issue
that
is
eunuch
eight
unique
to
the
net
I've
example
how
to
provision
in
the
prayer
of
auction
and
how
to
do.
Then
there
are
aliens.
So
this
document
many
major
to
know
together
to
document
the
gaps.
Are
you
know
in
the
existing
control
and
solutions
of
the
document
tests
test
document
that
does
not
intend
to
you
know
to
define
some
specific
solution.
C
This?
The
solutions
will
be.
You
know,
document
in
dedicate
document,
for
example,
if
we
want
to
extension
the
RC
BT
protocol,
then
we
be
sterilized
us
in
the
case
working
group.
So
here's
a
next
step
our
about
this
work,
so
I
think
this
work
has
already
discuss
in
a
society
of
meeting
so
based
on
current
charter.
There
are
the
same
sacrament
that
this
work
is
within
the
current
charter.
C
B
Can
I
have
a
comment
on
the
previous
slide
I
understand
it's
the
long-term
plan
and
no
comments
on
that.
But
when
you
say
that
that
integration
existing
control
planes
there
is
a
sorry
six
and
we
will
hear
about
it
soon,
but
that's
a
new
data
plane.
So
for
the
rest,
actually
we
have
data
plane
already
and
we
first
should
have
maybe
data
plane
before
having
control
plane
when
it
comes
to
us
every
six,
a.
A
A
Yeah
I
think
I
think
this
person
is
focusing
on
thinking
that
controller
plane
has
to
be
signaled
and
one
of
the
things.
The
reason
we're
using
controller
plane
versus
the
older
term
of
control
plane
is,
is
that
it
may
be
fully
centralized
and
that
you
can.
You
could,
for
example,
do
resource
management
with
sr
without
having
to
signal
it.
Just.
C
A
C
B
A
So
right
now
we
do
have
an
issue.
This
was
going
to
come
up
later,
that
we
don't
have
a
definition
of
what
what
traffic
engineering
4sr
means.
This
is
an
issue
that
we've
been
talking
with,
let's
the
spring
chair
as
well
as
the
80s
and
will
continue
to
work
between
now
and
the
next
meeting,
but
I
think
that
that's
consistent
with
the
point
beyond
or
say:
let's
not,
let's
not
focus
on
SR
to
start
yeah
and
and
get
there
as
we
do
work
out
these
other
topics
yeah.
A
A
A
How
many
think
we
should
not
be
working
on
this
and
keeping
in
mind?
This
is
a
gap
analysis.
This
is
not
a
solution
document,
so
I
had
someone
who
looked
like
they're
gonna
come
to
the
like,
but
I've
and
I
made
that
caveat,
I,
think
I
said
they
sat
back
down
right,
so
they
I
think.
The
objection
there
would
be
is
is
if
we
start
talking
about
solutions.
There
are
other
working
groups
that
own
the
solution
space
and
we
shouldn't
jump
on
their
activity
and
that's
absolutely
correct.
A
So
we've
we've
said
that
there
was
a
good
number
who
thinks
we
should
be
working
on
this.
There
were
no
objections.
How
many
think
that
this
is
a
reasonable
foundation
for
our
work
in
this
area.
In
other
words,
we
should
adopt
how
many
think
we
should
adopt
this
document
as
a
starting
point.
That's
very
that's
that
that's
much
less.
Who
then
think
that
we
should
work
in
this
area?
That's
a
little
bit
of
a
surprising
result.
Maybe
I
phrased
it
in
a
way,
but
no
matter
what.
A
D
A
D
Why
we
need
an
ax
from
first
place?
There
are
some
new
applications
and
bring
some
new
requirements
for
network.
So
perhaps
we
need
some
new
technologies
rather
than
the
traditional
ones.
So
when
you
detonate
and
I
think,
people
in
the
room
may
be
very
familiar
with
these
three
features,
often
at
the
resource
allocation
to
avoid
congestion
and
the
service
protection
to
provide
a
redundant
transport
to
to
avoid
the
failure
of
the
flow
and
also
explicit
a
routine.
So
this
work
we
try
to
implement
that
netting,
yeah
yeah
SR
SR
v6
domain.
D
So
why
is
r6
first
as
a
v6
in
sauce
routine
and
as
r6
could
steer
the
packet
flows
through
the
network
according
to
an
explicit
past
with
allocated
resources
and
second
network
programming,
a
server
six
can
support
the
the
functions
that
will
be
needed
in
Internet
and
also
metadata.
A
services
could
carry
the
information
that
will
be
needed
in
that's
net,
for
example,
the
flow
identification
and
sequence
number.
D
So
why
not
here
is
the
architecture
for
a
Sophists
six
for
ten
answers
protection,
because
there
are
three
features
seen
that's
not,
and
in
this
stage
we
just
cover
the
service
protection
first
and
the
first
days
the
data
plan.
Basically,
the
packet
will
goes
through
the
ingress
node
and
in
the
ingress
node
it
will
be
inserted.
The
SR
v6
policy
that
will
steer
the
packet
from
the
ingress
to
the
relay
node
and
also
the
fluid
application.
A
sink
with
number
will
also
be
counting
in
the
SRH
and
a
general.
D
They
know
the
one
packet
of
replication
happens
and
after
the
replication,
the
payload
and
the
ipv6
header,
with
the
SRH
a
binding
to
different
as
our
basics,
policies
which
steer
the
packets
through
different
paths
and
the
in
the
elimination
note,
the
redundant
packet
will
be
eliminated
and
the
the
survival
will
transmitted
through
relay
know
to
go
to
the
egress
node
and
in
the
control
plan.
We
consider
and
we
considering
that
to
the
first.
E
D
And
after
the
after
the
controller
configure
the
action
or
the
it
should
also
configure
the
relay
node
and
the
replication
living
node
should
be
configured
to
identify
that
then
then
I
flow
spy
flow
identifications
after
application
or
elimination.
The
explicit
passed
you
next
relay
should
also
be
required
through
some
control
plan.
Extensions.
G
F
H
D
H
D
D
D
H
D
You
thank
you,
and
here
is
the:
what
do
we
need
when
we
do
a
survey
six
space,
that
than
that?
First,
we
need
a
manner
to
identify
the
10,
L
payload
type
and
also
a
suitable
explicitly
routine.
You
deliver
the
Dynaflow,
which
will
be
supported
by
the
segment
at
least
and
yes,
RH
and
also
we
should
have
a
method
of
indicating
the
packet
processing,
such
as
the
functions
required
in
the
service
protection,
for
example,
packet,
replication
uttering
and
elimination.
D
Also,
we
need
a
method
of
identifying
the
denna
flow
and
carrying
the
10.
A
sequence
number
actually
carries
queuing
in
the
forwarding.
Indication
will
also
be
considered,
but
it
is
not
in
the
current
version
of
the
draft.
Here
we
bring
some
options
to
do
this.
First,
we
can
to
carry
the
flow
identification
and
the
sequence
number.
The
first
option
is
we.
D
We
carry
the
flu
identification
and
sequence
flow
identification
sequence
number
in
the
optional
TLV
and
we
designed
how
to
how
to
link
holding
them
in
the
s
rh,
and
here,
is
that
we
design
new
functions
for
test
net,
which
is
replication
and
elimination,
use
that
that
is
basically
an
extension
of
the
current
binding
set.
It
means
that
after
replication,
it
will
binding
new
new
segment,
at
least
to
the
package,
and
also
this
one
is,
after
the
elimination
to
abiding
a
new
segment
least
to
the
packet,
and
here
is
the
example:
how
to
do
this.
D
Basically,
in
the
ipv6
header
there
is
the
source
address
and
the
destination
address
we
in
this
picture.
It
shows
as
the
end
station
1
and
station
2
and
again
the
ingress
node.
We
will
insert
as
our
age
that
currents
that
the
east
source
address
is
ingress
node
and
the
next
the
destination
address
is
transit,
1
and
in
the
SRH.
The
seminar
list
shows
the
relay
node
1
and
the
transient,
not
one
and
the
in
transit,
not
1.
D
D
D
I
Data
plane
where
we
would
also
need,
on
a
per
hop
basis,
the
appropriate
shaper
for
the
latency
control
that
we
needed
at
net
and
then
compare
that
natively
six
data
plane
with
the
SR
v6
solution,
and
then
you
know
explain
what
the
benefits
would
be
right,
because
I
think
that
that
is
hard
to
vet
just
from
looking
at
this
solution.
Obviously
the
interesting
part
is
the
transit,
not
the
stuff,
on
the
edge
with
the
duplication,
duplication,
I
think
that
part
would
be
certainly
the
same.
J
D
J
D
Yeah,
actually
in
the
in
a
draft
I
described
this
part
because
in
the
previous
design
you
know
I'm
PRS
data
plan
solution.
There
is
we
can
find
the
service
sub
layer
and
forwarding
sub
layer.
Obviously,
because
it's
label
stack
right,
but
in
sr6
it
is
cannot
be
separate,
very
obviously,
which
part
is
forwarding
sub
layer
and
which
party
is
service,
a
bigger
sub
layer,
because
the
the
metadata
that
is
needed
by
the
server
sub
layer
is
carrying
inside
the
SRH.
D
So
although
there
is
no
sub
layer
seams
in
the
data
data
encapsulation,
but
inside
the
process,
the
device
process,
it
will
be
serviceable
there
and
forwarding
sub
layer
and
each
sub
layer
will
do
their
own
job.
But
they
will
find
the
corresponding
information
inside
the
SRH
from
different
part.
But
but
it
is
not
seems,
like
the
I'm
sure
as
data
plan,
but.
J
I
think
this
is
what
is
my
major
problem
here
that
if
you
are
looking
here
to
the
transport,
node
and
I,
think
this
is
but
previous
commander
or
circum
entities
that
in
the
transport,
nodes
or
transit
notes,
you
need
that
net
forwarding
sub-layer
how
to
recognize
the
flows
if
it
is
within
your
s,
r,
v6
header,
which,
by
definition
should
not
be
touched
by
the
transit
nodes.
Actually.
D
If
more
functions
will
be
needed
for
their
transit,
node
I
think
the
transient
node
should
has
a
seat
to
indicate
that,
because
there
it
needs
some
special
processing
or
functions.
That
is
reasonable,
but
in
this
case,
I
have
already
said
that
in
this
case
the
only
service
protection
is
included
because
replication
and
the
elimination
is
the
main
focus
in
the
current
draft
and
the
perhaps
after
this
part
of
work
is
accepted
or
people.
Think
of
this.
That
is
reasonable
and
we
can
go
further
to
the
other
functions
about.
Maybe
the
transcend
out.
D
J
D
I
will
in
this
picture
it
shows
that
a
Service
protection,
Needham
Ella
metadata
of
flow
identification,
a
six
number
right,
so
we
put
in
the
option
one.
We
put
these
two
information
into
the
optional
TLV
SRH
and
in
service
protection.
We
also
provide
some
special
functions
to
do
this,
including
replication,
elimination
and
ordering.
So
we
will
extend
the
current
functions
of
sr6
programming.
D
We
need
new
functions
for
that.
We
have
already
discussed
this
before
and
also
in
the
left.
We
see
that
we
also
need
some
resource
allocation
in
this
resource
allocation.
Maybe
we
need
new
functions
for
scheduling
or
resource
reservation.
If
so,
we
will
also
do
the
extensions
to
the
functions
and
if
there
is
new
metadata
needed-
and
maybe
we
will
end
more
data
into
the
seed
as
arguments
or
inside
function,
that
will
be
a
lot
of
implementations.
J
D
D
J
D
A
D
The
relay
node
and
the
elimination
node
in
the
replication
on
an
elimination,
node
and
in
this
picture
that
is
the
relay,
know
the
one
r1
and
really
know
the
two
are
to
the
arch.
R1
will
do
the
replication,
and
that
is
a
new
function.
That
is
that
function
is
an
extension
of
the
current
abiding
function
of
the
sr6
network.
Programming
and
also
in
r2
elimination
will
be
done.
So
there
is
a
new
function
for
elimination
and
after
the
elimination
they
were,
there
will
a
new
as
sorry
as
are
H
abiding
into
the
packet.
Okay.
H
H
D
D
H
D
Actually,
here
there
are
some
description
in
the
function
design
we
we
use,
we
use
some
funk.
We
use
some
information
bringing
from
the
previous
as
Irish
header,
which
in
this
picture
in
this
slide,
we
can
show
that
we
bring
the
sequence
number
and
flow
and
education
in
the
optional
TLV
in
the
first
SRH
right
and
hinder
after
the
replication.
D
H
A
We've
done
a
good
job
of
filling
up
time,
so
we're
why
we
were
ahead.
We're
now
right
on
right
on
time.
So
I'd
like
to
ask
the
folks
in
line
to
try
to
keep
it
brief
and
I'll,
also
give
a
preview
of
what
what
the
chairs
are
gonna
say
on
this
is
any
new
functionality.
New
sr
v6
function
like
this
version.
We
have
to
take
to
that.
The
working
group.
D
I
Total
Secord
again
I
mean
my
understanding
is
that
there
is
an
SR
architecture
right,
and
would
you
really
be
good
to
say
that
what
how
are
we
doing?
Debt
net
NSR
with
the
ability
support
both
SRS
MPLS
and
SR
v6?
The
gaps
for
both
of
them
may
be
different
in
terms
of
forwarding
plane
features,
but
the
architecture
should
and
could
be
the
same
right
so
I
kind
of
hate.
The
idea
of
saying,
oh
today,
we
think
SR
v6
is
hip
and
leaf.
I
D
You
Tony's
for
a
comment.
Brief
respond.
I
I
haven't
considered
this
problem
from
the
view
of
the
sales.
Actually,
it
is
said
in
the
previous
data
plan
solution,
strapped
that
the
if
we
use
as
our
MPs
it
is
supported
naturally,
by
the
current
MPs
data
plan
solutions.
So
there's
there
is
no
more
job
as
our
MPs
I'm.
F
F
Don't
you
don't
necessarily
/
the
architecture
allows
arbitrary
splits
and
arbitrary
merges
and
I
was
just
trying
to
understand,
making
sure
that
you
weren't
making
an
assumption
of
one
single
replication
going
on
on
a
on
a
path.
So
if
you
go
back
to
your
bigger,
yes,
that's
that
one
there
so
first
off
our
one
in
the
general
architecture
is
allowed
to
be
replicated,
have
multiple
replications
there
and
then
t3
could
be
our
3.
D
Actually,
I
think
it
is
a
very
valuable
comment
and
we
consider
this
problem
from
beginning.
We
design
it
this
way,
because
if
we
replicate
it
again,
yeah
replicate
pass,
it
will
makes
the
things
more
complex.
You
know,
because
we
allows
this
case.
This
is
just
the
simplest
one.
If
we
do
the
replication
again
in
one
of
the
paths
and
we
merge
in
that
path,
it
will
makes
the
there
will
be
header
and
header
and
header,
so
that
will
makes
the
case
really
complex.
D
F
Well,
what
I
think
we
need
to
write
what
we
need
to
do
in
the
next
version?
What
you
need
to
know,
I
think,
is
to
take
some
of
the
arbitrary
replication
P
or
function
the
designs
that
we
have
in
the
framework
document
and
demonstrate
how
you
would
solve
those,
because
this
is
easy
to
solve,
and
we
just
need
to
make
sure
we
can
solve
the
general
case
or
put
the
limitations
in.
J
Elijah,
a
variation
in
the
document
in
the
prayer
of
section
in
section
for
the
tree,
you
are
speaking
about
point-to-multipoint
two
point
parts
and
just
thinking
about
that.
We
intend
to
to
work
also
on
some
some
controller
paint
stuff.
So
I
don't
understand
that
term.
So
please
clarify
what
does
it
mean
because
I
think
as
Stewart
also
referenced,
that
that
you
can
do
any
number
of
replication
elimination
during
the
path?
So,
but
does
it
mean
that
point
move
the
point
to
point.
D
Maybe
this
word
is
a
little
confusing.
I
can
explain
it
or
remove
it,
but
you
might
in
my
understanding,
it
just
means
that
it's
just
it
is
not
the
P
2
and
P
it
after
the
replication,
the
elimination
exists.
That
is
what
I
mean
from
the
first
place,
but
it
makes
you
a
little
confusing
I.
Can
we
can
talk,
talk
about
more
how
to
describe
this
case.
I
think.
D
A
K
Little
to
the
look
of
the
structure
of
the
peers
and
over
detonate
and
either
similar
to
that,
he
has
an
over
as
our
MPs
and
the
left
part
and
the
right
part
is
autumn
TSN,
TSN
network
and
mid
parties.
There
is
our
v6
to
me
and
that
c1
and
it
will
sense
the
critical
patriotism
to
p1
and
p1
and
the
AG
ed
notable
as
I
said
as
our
v6
and
we
are
performed,
the
factions
of
mapping
and
the
publication
and
the
elimination
and
send
it
to
the
P
to
another.
K
P2,
then
at
the
traffic
of
Theoden,
will
be
recovered
to
send
it
to
see
that
was
the
pedicle
structure
of
Tiran
over
to
the
net.
Now
that
we
take
a
look
at
mr.
lewin
of
taking
cover
seriously
relation
structure
of
tears
over
as
our
v6,
we
have
occurred
several
options
for
encapsulation
of
tears
and
over
as
our
v6,
the
further
wines
just
repackaging
packet
and
there's.
Second,
wine
is
inserted.
K
A
trading
to
tear
at
our
v6
packet
will,
together
with
some
control,
innovations
and
the
third
one,
is
just
this
speaker
because
we
think
about
them,
as
are
a
cheater
overhead
of
authorities,
elite
hole,
a
turn,
so
I
think
we
here
in
order
to
save
their
overhead,
and
we
should
a
salute.
The
payload
of
the
TSN
and
the
insult
of
the
payload
Bay
loaded
into
them
as
ihe.
K
Just
follow
the
as
IH
and
two
days
are
56,
where
I
did
note
martyr
support
the
following:
kids
in
Cumberland,
the
fourth
one
is
Oh
Becca
nice
gentle
flows,
including
the
stream
identifications
and
a
second
for
fr
yard
user
inside
the
Teton
domain,
and
the
third
one
is
forever
our
yard.
When
the
note
is
she
hasn't
replication
or
elimination
parade,
and
it
is
the
capitulation
structure
for
anxiety
for
TSN,
we
have
a
just
unique
Lydia
minor
change
for
the
si
D
and
with
the
bank
world.
Combat
mobility
with
the
authority
definely
Intel
in
another.
K
Entrust
next
page
will
give
a
teacher
Union
detail.
Interaction
photo
anxiety.
We
have
a
defenders.
Si
te
didn't
area
is
ideal
similar
like
this.
That's
a
curvy
and
with
the
64
bits,
long,
employee
identifications
and
there
are
secret
secret
number
and
the
other
piece
are
reserved
for
the
future
functions.
K
An
ability
family
of
new
to
new
and
as
a
v6
programming
functions
that
the
entropy
prevarication,
Terran,
lsit
and
an
appeal
elimination
force
for
the
pantry
elimination
function,
and
this
is
just
till
corruption
stopped
and
in
the
future.
Maybe
we
need
to
discussing
and
then
elaborated
to
functions
and
the
moment
I
would
have.
I
would
like
to
take
a
example
over
the
NEP
replication
for
the
Terran
anxiety
and
the
talented
node.
We
will
present
a
original
packet
and
a
generator
to
duplicated
packet
patron
Terra.
I
ib
v6
@
IH.
K
K
Now,
on
the
decision
to
give
a
summary
for
the
data
plan
considerations,
then
the
two
aspects,
the
third
one-
is
written
in
intaglio,
replication
and
elimination
order
functions
and
the
second
is
the
aging
order
processing
and
to
the
management
and
control
information
summary.
We
included
the
first
vegas
for
exit
of
elimination.
K
Here,
that's
at
the
third,
why
the
TSN
screen
identification,
that
here's,
an
art
rank
information
and
the
second
is
IP
v
6
for
the
trait
and
third,
an
attorney
is
the
idea
is
to
the
revi
6
destination
address
and
therefore
it's
the
ipv6
traffic
or
class
and
and
to
the
next
date
and
further,
although
we
will
solicit
it,
a
comment
for
this
is
dropped
and,
secondly,
we
want
to
elaborate
the
first
and
the
second.
Certainly
we
want
to
other
brands
and
they
improve
their
implementation.
For
this
suggestions.
That,
though,
about
the
truck.
D
That
draft
is
published
about
one
year
ago,
and
the
contents
of
the
draft
is
very,
very
similar
of
this
one.
If
you
haven't
read
it
I
recommend
you
read
it
this
first
and
because
the
it
is
too
similar
with
each
other,
I
I
have
to
say
I,
prefer
you
put
this
part
of
work
as
and
contribution
or
comment
on
the
mailing
list,
because
I
cannot
see
a
lot
of
more
work
in
this
draft
compared
to
our
previous
one.
K
We
have
a
32-foot
their
website
for
the
similar
department.
We
have
got
five
similar
part
related
to
this
one,
but
ice
from
my
site.
Those
five
was
thought.
You've
read
from
this
one
and
your
table
is
forgotten
to
turn
it
over
as
our
v6
right,
and
it
did
why
you
focus
on
here
then
override.
Obviously.
D
B
B
D
B
K
I
thought
it
may
be,
you
me
and
my
colleague
our
will
change
their
cut
in
the
website.
If
they
had
this
identical,
we
will
always
do
it.
Even
if
they
are
event
I,
don't
know
the
continue
are,
did
it
our
can
research
it
independently,
and
then
my
suggestion
and
and
from
my
side
then
you,
your
father,
is
ethically.
Is
it
even
front
with
what.
K
A
L
Grandma
I'm
changing
from
titi.
My
presentation
today
is
the
detonator
ques
cure
s.
The
drafter
has
been
presented
in
ITF
103
and
we
got
little
feedback.
So
this
meeting
we
hope
to
get
more
feedback
out,
after
mainly
applying
existing
ques
work.
Under
that,
let
less
quick.
Let's
do
a
quick
recap
of
the
ITF
cues
ITF
cues
model
has
been
proposed
and
adopted
by
the
RTE
gwg
working
group.
It
mainly
includes
policy,
a
classifier
action
and
the
target,
and
it
many
user.
L
If
they
wish
deficiency
80,
the
surface
is
diffserv
for
short
under
the
tip
short
tips.
A
piece
of
architecture
is
defined
in
our
press,
824
and
75
and
and
real
MP
else.
Ifs
of
is
also
defined
and
in
there
ITI
RFC
third
3270.
So
there
are,
there
are
many.
There
is
marshal.
Work
has
been
done
and
stable
for
ITF
cues.
So
we
summarize
the
delayed
kills
requirements
in
RFC
8655,
so
they
are.
The
ester
follow
shown
a
telnet
and
long
dead.
L
Flood
flows
could
coexist
with
existing
class
of
service
and
the
death
net
flows
must
be
differentiated
from
law.
That
traffic
not
attract
traffic,
cannot
destroy
destructor.
There
Detlef
flows
and
the
net
flows
can
be
shaped
and
schedule.
Their
aggregation
of
the
led
flows
must
be
supported
and
must
others
more
others
can
be
provided
so
their
death
netquest
draft.
L
So
this
is
a
data
narratives
of
model.
We,
the
tips
of
model,
includes
their
classifiers.
Yes,
if
he
traffic
conditioners,
PHP
and
queen.
So
we
extended
the
tips
of
model
for
that
land.
We
define
that
latter
class
virus.
We
may
use
their
behavior
aggregate
the
be
a
classifier
order.
Marty
field,
EMF
classifier,
which
is
defined
before
we
we
are
used.
We
may
be
used
to
be
a
and
MF
classifiers
in
that
net.
We
also
add
aloo
that
latter
and
classifier.
It
means
it
named.
That's
not
flow,
it
is
DF.
L
For
short,
it
may
be
used
as
label
in
death.
Lat
MPs
encapsulation
and
maybe
user
flow
and
edification
in,
as
always
six
encapsulation
may
be
in
later,
and
so
we
are
used
with
field
to
class,
find
that
net
and
the
low
dense
net
flow.
Second,
we
defined
that
last
dscp
includes
the
lat
IP
DSP
and
that
amperes
DSP
the
sector
value
we
haven't
defined.
We
will
want
to
discuss
with
others
in
the
mailing
list
and
their
define
a
value.
L
Second,
third:
we
define
death
net
traffic
conditioners,
including
meter
maid
marker
shaper
dropper.
In
the
way,
at
a
low
action,
their
scheduler.
We
will
define
or
definite
profile
to
to
do
the
meter
of
action
and
to
see
why,
if
they're
death
knell
flow,
is
meat,
meat
eater.
Mr.
kettle
and
fire
a
provirus
requirement
so
men
we
that
define
our
death
natural
/
hot
behavior
that
PHP
we
define
a
veterinary
field
should
be
and
defined,
er,
T
and
service
class.
Then
there
are
death,
not
Queen.
We
as
we
one
known
tur
this
so
far.
L
L
A
Resources
and
it's
a
little
hard
to
understand
how
a
what
a
debt
net
PHP
is
relative
to
per
flow
resources.
Could
you
talk
about
how
those
two
things
relate?
This
was
the
same
question
from
103
and
I.
Don't
think
it
was
been
addressed
from
the
last
time
I
was
presented.
I
can't
remember
what
meeting
it
was
that
it
was
presented,
but
yeah.
L
A
L
L
A
Alright,
is
anyone
interested
in
hearing
more
about
this
topic?
That's
that's
not
good!
There's,
no
one
who
seems
interested
in
this
work.
I
think
you
have
to
do
a
much
better
job
of
seeing
of
communicating
how
it
relates
to
the
rest
of
the
working
group
activity.
Try
doing
that
on
the
list,
but
yeah
maybe
take
a
step
back
and
come
out
at
a
different
way,
because
people
aren't
understand,
I,
think
the
way
I
read
it
is
people
aren't
understanding
how
this
relates
to
the
working
group.
Okay,.
B
So
why
Louis?
The
idea
was
that
we
would
have
time
to
talk
about
it
and
it
was
a
year
ago
when
we
had
the
workshop
and
we
are
proposing
another
one.
So
just
a
quick
update,
I'm
speaking
as
an
individual
as
usual-
and
this
is
this-
is
grouping
of
TSN
standards
and
projects
in
in
three
categories:
based
technology
configuration
solutions
and
profiles,
a
timeline
is
given
compared
to
the
last
ITF
in
Bangkok,
where
we
discussed
list
of
I'm
just
I
like
that.
B
If
you
are
puzzled
on
the
TSN
technology
that
we
have
so
many
standards
and
projects,
you
are
not
the
only
one.
That's
why
we
have
profiles
for
dedicated
for
particular
verticals
to
select
tools
out
of
the
toolkit
and
specify
their
use.
The
color
code
here
is
that
the
darker
shade
is
more
progress
to
work.
The
lighter
is
a
less
progressed
in
the
profiles.
The
ongoing
ones
are
the
IEC
I
Triple
E,
six
way
to
to
join
project
foreign
in
the
industrial
automation
vertical.
B
B
Bangkok
presentation
I
would
like
to
highlight
two
items
here.
The
really
good
news
out
of
last
week
is
that
the
times
in
our
time,
synchronization
standard,
the
802
da20
as
general
as
precision
time
protocol
profile
of
the
1588
PTP
has
been
approved
by
the
AIT
Plato
to
itself
and
sent
to
I
typically
standard
Association
revision
committee
approval,
hopefully
early
next
year.
So
we
are
practically
speaking
done
with
the
long-awaited
new
time.
Sync
revision,
and
there
are
some
new
work
items
here
in
the
resource
configuration
part
of
the
TSM
pillars.
B
I
would
like
to
highlight
here
that
work
is
ongoing
to
specify
a
new
reservation
protocols,
the
brace
protocol
being
a
so-called
link
local
registration
protocol
on
its
own.
It's
not
really
useful,
but
we
need
an
application
on
top
and
the
ongoing
work
is
the
resource
allocation
protocol
for
that.
This
is
a
new
project
compared
to
one
year
ago,
and
the
two
together
will
provide
the
new
resource
allocation
and
this
work
has
been
started
to
me.
B
The
industry
automation
needs
based
on
all
the
discussions
with
the
key
contributors
they
consider
it
applicable
for
death
as
well
and
I've
been
encouraging
them
to
contribute
to
that.
Not
so
a
heads
up
these
contributions
may
show
up
here.
Another
new
project
is
the
allowed
IP
version
2,
which
was
largely
triggered
by
the
lsv.
Our
work
going
on
in
IETF
and
summarize
article
802,
is
on
individual
contributions.
Individual
participation,
just
like
ITF,
you
are
welcome
to
contribute,
participate.