►
From YouTube: Contour Community Meeting - Dec 3, 2019
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
A
It's
crazy
we're
closing
in
on
2020,
so
yeah
super
super
excited
I'm
super
happy
to
have
you
all
here
and
we're
gonna
cover
a
bunch
of
stuff
today,
one
of
the
first
things
we
want
to
talk
about,
though,
is
UConn,
so
we
were
there
and
of
course
we
were.
There
were
roughly
12,000
of
our
closest
friends
over
in
San
Diego,
and
we
had
a
bunch
of
great
discussions
and
conversations
with
people
around
envoy
and
contoured,
so
I
know
Steve
and
Nick
you
we
were
at
envoy
con.
B
Yeah
yeah
so
Nick
and
I
both
that
we
both
have
speaking
slots
there,
don
khon
so
be
a
lot
of
fun.
I
know.
Next
talk
got
a
lot
of
nods
and
head
bumps
from
some
customers
there,
as
well
as
or
users
as
well
as
the
unwitting
themselves
who
I'm
you
know,
came
Brennan
branded
the
same
problems
that
that
we
had
run
into
around.
You
know
what
Nick's
talk
was
about
so
which
is
interesting,
so
I
talk.
B
A
lot
of
people
like
people,
I
had
conversations
with
first
seem
to
be
new
to
kubernetes
into
like
contour
in
the
world,
so
folks
were
interested
in
you
know
what
should
they
use?
How
do
they
get
traffic
in
other
cluster
and
that
sort
of
thing
being
that
you
know
we
write
an
ingress
controller?
It
was
a
interesting
for
them
because
they've
just
from
the
new
standpoint,
so
that
was
cool
let
ago
talks
on
the
talks
are
online,
which
is
interesting.
C
Yes,
so
for
me
it
was
interesting
to
give
a
little
talk.
I
ever
talk
about
what
we
change
or
we
can
feed
an
envoy
to
make
it
an
intra-party
and
yeah.
As
Steve
said,
oh,
but
there
was
a
lot
of
people
who
obviously
were
doing
a
similar
thing.
The
and
I
got
I'd
be
the
feedback
from
some
of
the
commentators
as
well
I'm
saying,
oh,
did
you
know
about
this
part
of
the
documentation?
C
B
It
was
bigger
than
I
expected,
I,
guess
and
crowd
wise
there's
part
of
it
with
three
for
two
people,
I
think
they're
I,
just
that
envoy
con
itself,
which
I
thought
was
great,
so
it's
yeah
yeah.
It
was
good
hearing
from
the
maintainer
'he's
about
where
things
are
going
with
the
XDS
api
that
the
three
is
coming
in
December
and
it
promised
the
back
port
fix
this
into
v2
along
that
journey.
To
help
people
migrate,
I
think
they're
gonna
leave
that
open
for
a
year,
so
that'll
be
a
direct
impact
to
contour
itself.
B
D
A
B
I
think
I
know
I
had
some
other
talks
after
my
coop
con
talk
on
Tuesday
with
some
folks
again
that
some
of
them
were
new.
Some
of
them
asked
how
they
should
basically
yeah
good
guess.
My
talkin
con
was
how
you
could
leverage
contour
to
bridge.
You
know
things
together
without
having
to
deploy
a
service
mesh
right.
So
if
you
want
to
go
down
that
road
of
deploying
a
full-fledged
service
mesh
and
all
the
side,
cars
and
all
that
overhead
in
complexity,
so
they're
interested
in
you
know,
hey
I
want
it.
I
want
to
deploy.
B
B
A
A
So
I
want
to
add
something
in
here,
because
we're
going
to
talk
about
a
new
release
coming
up
quite
shortly
here,
but
between
the
last
community
meeting
and
this
community
meeting,
we
had
a
major
milestone:
release
had
the
1.0
release
of
on
tour,
so
I
just
want
to
again
applaud
the
team
here
for,
for
making
that
happen,
really
really
awesome.
Job
super
super
fun
tier
to
be
a
part
of
this
journey
with
y'all.
A
D
So
I
think
the
the
evolution
of
ingress
route
to
HTTP
proxy
has
caused
some
community
members,
some
anxiety,
especially
those
with
large
deployments,
so
I,
don't
wanna,
so
I
think
it
would
be
great
to
have
some
tooling
to
help
those
folks
and
offer
some
assistance.
I
know
we've
already
talked
about
it,
so
it's
not
exactly
foreshadowing
but
I
I
know
because
of
that
change.
Some
people,
you
know
some.
Some
community
members
plans
had
to
change
with
an
existing
deployment
plan
or
something
like
that.
B
Yeah
I
had
similar
similar
questions.
Maybe
the
same.
The
same
person
I'm,
not
sure,
but
you
know
they
were
came
with
it
was
more
of
you
know:
do
we
have
plans
to
get
there
I
think
we
do
in
the
upcoming
release
of
1.1
to
help
folks,
you
know
move
from
from
ingress
route
to
HTTP
proxy,
so
it's
coming
it's
just
more
of
like
they
wanted
to
kind
of
see
that
and
how
that
work.
C
Awesome
yeah
I
actually
does
those
thanks
for
reminding
him
and
they.
The
other
thing
that
was
important
was
that
the
USB
to
work
in
scene
networking
is
actually
sort
of
spinning
up
and
it
looks
likely
that
there
will
be
a
working
group
created
in
the
near
future.
I
attended
a
meeting
of
what
probably
will
and
not
becoming
a
working
group,
and
it
was
pretty
positive.
C
The
English
the
English
way
to
expect
takes
a
couple
of
cues
from
some
of
the
stuff
we've
done
in
contour
with
delegation
and
I
walk
them
through
how
we
changed
from
delegation
to
inclusion,
and
why
and
a
few
other
problems
that
I
saw
with
the
with
the
Victor
spec.
Everyone
was
happy
to
hear
what
I
had
to
say
and
was
going
to
hear
what
everyone
else
had
to
say.
C
We
had
a
meeting
with,
so
it
was
maintain
us
from
contour
ie,
the
engine,
X,
engrossed,
controller,
Kong,
East,
EO
and
then
a
button,
then
those
sort
of
Google
people,
the
seen
networking
people
all
right
now,
I'm
so
pretty
useful
meeting.
Although.
E
C
F
They
create
as
interest
groups
around
different
areas,
for
example
the
runtime,
the
security,
the
storage
and
networking
also
got
its
own
sig
as
well.
I
make
I'm
wondering
if
it's
worthwhile
for
some
of
us
to
be
involved
in
signal,
work
and
I'll
paste
the
link
here
to
the
Chad.
This
is
a
Oh
Jonas
posted
it.
So
there
you
go
Jonas,
so
I'm
wondering
if
it's
a
worthwhile
for
one
of
us
to
start
participating
and
assuming
Dave,
you
raise
your
hand,
there's
a
Google,
so
yeah
I'm.
G
F
G
F
F
F
F
F
Wondering
if
some
of
the
you
know
as
part
of
that
working
group,
that's
working
on
kubernetes
v2,
ingress
that
the
kind
of
participating,
assuming
that
some
of
those
things
will
eventually
bubble
up
to
to
seek
network
as
well
but
yeah.
If
nothing
is
happening,
then
I
guess.
The
good
thing
is
that
the
good
folks
from
envoi
are
driving
that
sick,
which
is
I,
guess
a
good
thing
for
us.
F
Help
they
just
like
nothing's
happened
yet.
Do
you
mind
if
you
pink
Mutt
client
directly
and
ask
him
hey
what's
going
on
when
we're
gonna
jumpstart
this
music?
It
might
be
that
he's
overloaded
that
he
doesn't
have
the
time
of
the
resources
to
do
this,
and
we
could
potentially
even
pick
up
a
more
of
a
leadership
position
here.
F
A
Can
do
that
for
looks
like
there's
a
PR
or
the
Signet
work,
Charter
proposal
and
it's
linked
from
the
meeting
notes
and
that's
the
last
thing
that
can
gone
November
7th,
but
there's
it's
still
open
late.
It's
nothing
there.
It's
still
forming
well
yeah
Dave,
you're,
absolutely
right
and
that
nothing
else.
F
A
H
We
probably
he
did
decide
soon,
whether
word
the
ordering
in
which
we
release
contact,
1.1,
missus,
contour
1.0
one
because
on
boy
one
point
twelve
point:
two
is
not
going
to
drop
until
next
week.
The
other
main
things
in
contour
one
point:
one:
the
prefix
rewrite
support,
which
is
last
major
missing
feature
that
we
brought
over
from
ingress
route
and
the
yeah,
and
it's
a
bunch
of
test
cases
and
kind
of
corner
cases
that
Dave
is
fixed
around
TCP
proxy.
G
H
G
Do
we
do
it?
Do
we
do
another
release
on
the
100
series
just
for
that
on
release
just
just
to
step
back
a
little
bit?
I
don't
want
to
combine
but
keep
putting
the
past
in
the
past,
going
for
I,
don't
want
to
combine
security
secur
later
at
races
and
about
the
feature
releases.
The
reason
for
that
is,
you
should
us
uses
a
contour
should
never
feel
you're
in
a
position
of
like
oh
I
want
to
apply
that
update.
G
You
know
it's
important
security
update
but
yeah
like
there's
some
functionality
changes
and
my
users
are
complaining
like
we
should
never
put
you
in
that
position,
so
taking
a
leaf
out
the
way
that
the
go
project,
which
is
they
will
always
do
a
separate
release
for
security
issues,
as
opposed
to
bug
fixes
our
improvement
so
I'm
trying
to
figure
out
how
how
our
timings
gonna
gonna
shake
out.
It
could
be
that
we
do
I'm
a
point
release
on
the
100
series.
G
It
could
be
that
we
do
one-to-one
release
and
then
soon
after
a
1.1.1
to
pick
up
the
Envoy
change,
perhaps
the
less
preferable
instant
just
do
it
is
to
roll
everything
into
1.1,
because,
as
I
said,
we
don't
ever
want
to
put
anyone
in
a
position
of
like
hey,
you
shoot.
You
know
the
supported
policy.
Is
you
use
the
latest
release
which
will
be
1.1
and
like
well?
There's
some
functionality,
changes
or
a
thing
I
didn't
understand,
especially
perhaps
around
there
we
found
a
lot
of
educators
anticipate
practicing
we're
just
things
were
inconsistent.
G
So
in
cleaning
up
hosing
consistency,
we
don't
put
you
in
a
position,
you
be
like
well,
one
of
my
users
complaining
it's
broken,
so
we
had
to
roll
back,
and
that
means
I'm
using
the
vinyl
version
of
Bomboy.
So
there's
one
of
those
some
of
the
things
we
have
to
consider.
This
is
a
slight
phase
ordering
issue
that
the
when
we
get
together
internally
to
chat
about
this
is
actually
after
this
meeting.
G
So,
ideally,
it
should
be
in
the
other
way
around
that
we
could
have
had
a
better
answer
going
forward,
but
just
to
just
to
be
aware
that
we
were
further
on
top
of
the
on
risk
situation.
We've
made
made
all
the
changes
we
need
to
in
the
product
to
accommodate
envoy
1:12
and
as
soon
as
those
bits
are
ready,
we're
going
to
be
ready
to
roll
out
of
the
room
of
these.
G
You
go
it's
it's
it's
it's
a
fairly
good
whoo.
It's
always
it
was
my
try
saying
the
goal
of
wonder.
One
is
not
to
said
not
to
set
the
world
on
fire,
but
make
the
main
goal
is
to
pick
up
some
of
the
things
which
we
didn't
get
done
for
1.0
and
also
to
put
us
in
a
place.
We
continue
to
deprecating
aggressor
out
and
get
it
out.
That's
always
that's
always
been
the
plan,
because
there
is
enough
confusion
with
the
number
of
different
ways
you
can
talk
about
ingressing.
There
are
two
fulness
there.
G
There
are
two
from
the
sea
and
see
if
they'll
be
three
by
next
year
and
potentially
four
we
have
two
of
our
own,
so
keeping,
even
though
we
want
to
will
always
support
all
the
things
that
upstream
do
what's
the
cognizant
that
sixty-four
ways
to
talk
about
the
same
thing
isn't
very
beneficial
for
anybody.
So
four
things
which
are
deprecated,
which
you're
not
going
to
get
new
development
work
in
them,
want
to
make
sure
that
we
have
a
path
to
get
you
off
then
onto
HTTP
proxy.
G
That's
part
of
that's
part
of
what
makes
been
working
on
you
you're
a
little
bit
bit
circumspect
in
how
how
the
translation
tool
is
working
so
maybe
make
you
could
talk
a
little
bit
about
how
you
have
message.
People
using
this
tool
like
again
it's
not
it's,
not
gonna
set
the
world
on
fire
is
literally
a
tool
that
translates
during
aggressive
objects
to
proxy
objects,
so
that
you
can
do
to
the
to
the
degree
that
mechanical
translation
is
possible.
You
can.
We
can
remove
that
as
a
roadblock
for
upgrading.
C
C
There
are
most
things
like,
as
I've
said,
once
we've
finished,
making
sure
that
the
that
HTV
proxy
is
featured
complete
with
respect
to
ingress
route,
then
we
should
be
able
to
support
all
the
features.
There
are
some
sort
of
edge
cases
where
it
doesn't
match
well.
So
in
that
case,
the
tool
just
warn
you
it'll
most
of
the
time
it
will
still
give
you
HTTP
proxy,
but
it'll.
Warn
you
and
say:
I
couldn't
do
this.
You
know
the
way
that
you've
done
this
in
English.
Now
it
doesn't
work
in
HTTP
proxy.
C
Ideally,
the
idea
is
that
eventually
it
should
point
to
some
documentation
is
worth
right
now.
Obviously,
just
love
the
husband,
but
so
the
intent
is
that
for
a
corpus
of
English
routes,
you'll
be
able
to
run.
You
run
a
set
of
run
this
command
across
each
of
them,
and
it
will
give
you
a
HTTP
proxy
ready
to
be
spat
out
somewhere
and
then
reviewed
by
a
person
and
then
applied
to
the
cluster.
As
a
corollary
to
that.
C
The
other
thing
that
I
will
be
working
on
for
this
week
is
to
try
having
a
look
at
trying
to
change
the
semantics
of.
If
you
have
an
overlapping
English
throughout
an
HTTP
proxy
I'm,
going
to
change
it
so
that
the
HTTP
proxy
will
overwrite
the
English
well
currently,
I
think
it
just
makes
them
both
invalid,
and
so
the
idea
being
that
you
can
apply
the
hey.
C
Should
the
new
hash
to
be
a
proxy
check
that
everything
works
and,
if
not
roll
it
back
by
deleting
the
object
and
if
it
does
then
roll
forward
by
deleting
the
English
route,
Thank
You
migration
path
to
get
to
from
English
to
H
difficulty.
So
that's
sort
of
the
plan,
yet
the
hoping
to
be
able
to
get
the
get
that
change
in
for
one
point.
One
part
epoxy
is
a
separate
tool
in
a
separate
repo
with
its
own
release
guidance.
C
A
G
B
G
Unprepared
all
the
prepared
notes
we
have
so
maybe
if
there
are
any
committee
members
on
the
call
we
can
throw
it
open,
because
this
is
really
your
opportunity
to
ask
those
questions
may
put
us
on
the
spot,
goes
to
explain
how
things
work
or
why
they
well,
they
don't.
So.
If
this
anyone
out
there
in
radioland,
we
hit
answer
your
questions,
heads
right.
E
Here
just
want
to
touch
on
that
point,
they've
made
about
the
security
updates
and
how
that
would
work
with
the
supporting
the
current
latest.
Are
you
looking
to
you
know
support?
Maybe
the
N
minus
2
just
for
security
updates.
So
then
they
don't
have
to
choose
between
feature
and
security
you're.
Looking
to
do
that
sort
of
thing
or
you
know
any
ideas
on
where
it
might
go.
G
We
should
definitely
it's
the
the
current
the
current
like.
We
only
support
the
latest.
The
latest
release
policy
is
not
not
particularly
flexible
there.
We
should
definitely
talk
about
if
there's
an
opportunity
to
change
that,
obviously
that
that
has
a
cost
like
in
in
time
and
maintenance
overhead,
but
which
I'm
not
qualified
to
speak
about.
D
G
It's
probably
a
good
point,
good
time
to
point
out
on
the
contour
website.
We
have
a
compatibility
table
between
contra
lists
and
envoy
releases.
This
table
is
very
brief
at
the
moment,
but
it's
gonna
grow
over
time.
One
of
the
restrictions
in
back
wanting
security
fixes
is
usually
the
fixes
aren't
to
our
product
they're
to
envoy
and
almost
on.
Would
you
not
have
a
backporting
policy?
G
They
only
they
bared
their
policies
effectively
the
same
as
what
we
model
arise
on
they
hate
they
don't
back
off,
fixes
toll
diversions,
which
puts
in
somewhat
difficult
position
that
we
have
to
make
an
older
version.
Contour
work
with
a
newer
version,
one
boy
which
is
a
risky
proposition,
Evangeline
see
we.
We
are
kind
of
between
a
rock
and
a
hard
place
because
of
there,
because
if
there-
because
they
don't
back
port
fix
it,
so
we
yeah
that
doesn't
matter
our
options,
a
lot
and
kind
of
drives
us
in
this.
G
This
motion
of
we
followed
them
very
very
closely.
Wherever
they
move,
we
move
as
quick
as
we
can,
because
I've
seen
what's
happened
to
other
projects
like
you
steer
that
are
stuck
on
their
own
formation
of
envoy
there
and
they're
in
really
really
unfortunate
situation.
I
don't
want
to
put
our
users
in
that
cool.
Thank
you.
H
I
I'm
happy
to
do
it
I
think
that
what
Nick
is
working
on
will
be
like
a
good
segue
into
into
providing
an
upgrade
path
for
that
easily
for
people.
I
just
wanted
to
do
it,
because
I
think
that
a
lot
of
folks
getting
started
out
with
communities
and
ecosystem
kind
of
look
to
home
charts
to
be
like
what's
available,
that
I
can
use
so
I
wanted
to
get.
I
F
Yeah,
so
that
attack
actually
and
another
thing
there
that's
happening
in
that
community
is,
you
know,
there's
this
open
container
initiative
around
buildings
and
OCA
registry
that
will
be
able
to
host
artifacts,
like
continual
images,
but
also
home
tours
at
the
same
time,
in
the
same
repo.
So
the
need
for
charts,
museum
and
other
repositories
is
gonna
wind
down,
eventually.
H
I
H
Be
awesome,
Peter,
I
think
the
other
thing
that
I'm
really
interested
in
with
their
helmet
art
is
having
their
two
or
three
different
deployment
options
which
which
people
need
which
currently
you
end
up
tweaking
the
deployment
yeah
well
a
little
bit
for
so
I.
Think
there's
just
one
or
two
different
things.
You
would
do
slightly
differently
if
you're
deploying
to
like
a
local
kind
cluster
or
if
you
were
deploying
to
AWS
or.
A
A
I
added
a
added,
a
link
in
the
the
channel
added
to
the
notes
as
well,
so
we
we
recently
did
this
for
harbor,
and
so
we
have
a
separate
repository
for
the
helm,
charts
there
and
we
just
linked
that
from
helm,
go
harder
than
IO,
or
rather
we
have
them
published
there,
because
that's
a
requirement
for
the
new,
the
new
structure.
So.
G
I
F
Yes,
I
can
do
that
so
I
think
you
know
part
of
the
reason
is
that
there's
a
quite
a
bit
of
effort
to
move
things
into
the
into
that
specific
repo.
But,
more
importantly,
our
goal
is
eventually
to
donate
contour
to
the
scene.
Cf
project
umbrellas
so
essentially,
since
CF
will
immediately
become
its
own.
Its
own
project
hosted
separately
from
everything.
Vmware
will
be
driven
by
the
community.
F
It
will
be
governed
by
the
CN
CF
policies,
code
of
contact,
governance
and
and
and
that
made
sense
for
us
to
kind
of
delay
the
merge,
because
we're
trying
to
go
in
different
direction.
I
think
that's
gonna
be
a
great
thing
for
the
community
and
a
great
thing
for
contour
itself.
Just
given
the
amplification
of
visibility
icon
to
will
get
with
that
donation,
that's.
D
And
that's
the
key
reason
for
moving
into
a
new
top-level
or
to
begin
with.
Also
VMware
ton
Zoo
didn't
exist
yet
when
this
decision
was
made
so
that
wasn't
even
an
option
on
the
table
and
even
if
we
had
decided,
we
wanted
to
be
in
dream,
we're
done
so
we
didn't
want
to
do
that
to
our
users.
You
know
bounced
around
twice
in
the
period
of
a
couple
months.