►
From YouTube: Meshery Development Meeting (April 28th, 2021)
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).
B
Yes,
hey
good
morning,
I've
been
I've
been
practicing,
something
which
is
rude
rocks.
A
A
B
C
C
B
A
A
Yes,
so
I
guess
we
will
start
get
started
with
the
first
agenda,
which
is
air
gap
deployments
iso
code?
Do
you
want
to
start
us
off.
D
D
D
B
Partially,
it's
been
going
in
and
out
some
but
asuko.
This
is
to
to
help
with
the
to
help
overcome
the
choppiness
of
your
voice
and
to
help
give
everyone
context
for
the
design
that
you're
putting
together.
B
Maybe
I'll
speak
to
it
a
little
bit
as
as
well
so
asuka.
One
of
the
things
that
you
just
mentioned
was
the
the
is
related
to
another
design
specification
that
husseina
is
putting
together
with
respect
to
eliminating
dependency
external
dependencies.
That
measure
ctl
has
so.
The
one
specifically
is
on
docker
compose
as
an
external
binary
and
actually
the
docker
cli
as
an
external
live
binary
as
well.
B
B
That
there's
been
a
project
that
the
docker
community
has
led
to
rewrite
that
in
golang
and
then
potentially
you
know
so.
For
projects
like
ours,
we
can
potentially
incorporate
docker
compose
as
a
golang
library
and
have
that
a
bit
more
of
a
tight-knit
interaction
and
also,
at
the
same
time
eliminate
the
need
for
that
external
dependency,
which
is
nice
so
kind
of
related
to
this.
But
it's
separate.
B
B
Okay-
okay,
he's
not
on
that,
so
I'm
talking
talking
to
myself
in
front
of
a
bunch
of
people.
This
is
embarrassing,
so
cool!
So
then
let
me,
let
me
say
these
these
things
so
that
to
help
speed
this
along
and
and
hopefully
to
help
level
set
with
asuko
as
well.
So
the
genesis
of
this
particular
effort
is
well
not
quite
what's
described
here.
So
the
title
of
the
design
is
air
gapped
installation
of
measuring
on
kubernetes
yeah
that
we
need
instructions
for
that.
B
Yes,
we
also
need
instructions
for
air
gapped
installation
of
meshri
in
docker
containers
running
in
docker.
We
basically
just
need
instructions
for
air-gapped
deployments
of
measuring.
If
any
of
you
aren't
familiar
with
the
term
air
gapped,
some
half
of
you
are
and
half
of
you
are
wanting
to
jump
in
and
explain
it.
Does
anyone
want
to
explain
it
by
the
way.
B
Wow,
all
of
you
are
showing
exercising
strength
today
so,
but
air
gap
deployments
are
like
it
is
just
that.
It's
that
you
would
deploy
some
software
in
an
environment
that
doesn't
have
a
network
connection
or
rather
doesn't
have
network
connection
to
external
systems,
and
so
suco
is
still
off
the
call.
So
I
was
just
checking
and
that's
kind
of
a
problem
for
meshri,
so
so
when
measuring
decides
to
when
the
user
tells
mesri
to
deploy
a
service
mesh.
B
Well,
one
of
the
first
things
meshrey
will
do
is
reach
out
to
the
internet,
grab
the
installation
files
the
before
that
specific
service
mesh,
pull
down
those
files,
and
then
you
know
it
automates
the
deployment
of
that
service,
mesh,
its
configuration
and
so
on
and
okay.
Well,
that's
a
problem
because
it
tries
to
reach
out
to
the
internet.
Okay!
B
Well,
that's
one
issue
and
then
there's
you
know,
as
you
walk
through
this
there's
a
few
different
there's,
a
few
different
considerations
that
need
to
be
accounted
for
and
for
users
that
want
to
you
know,
run
mesh
re
air
gap
that
air
gap
by
the
way.
B
You
know
that
that
lack
of
network
connectivity
that
can
be
done
in
on
premise
and
is
most
often
done-
on-premise
like
in
an
environment
I
should
say
most
often
is
most
often
done
on
on-premise
in
an
environment
where
people
have
downloaded
software
downloaded
the
bits
and
then
by
one
way
or
the
other
usb
stick
or
the.
What
do
they
used
to
call
it?
The
the
foot
network
or
the
but
anyway,
one
way
or
the
next
they
push
over
the
bits
so
that
those
systems
that
are
disconnected
from
all
other
systems
can
run
software.
B
That
can
also
be
done
in
a
public
cloud.
So
you
can,
you
know,
spin
up
services
in
an
environment
in
a
public
cloud
and
but
restrict
that
particular
you
know
segment
and
coordinate
that
particular
section
of
your
public
cloud
deployment
to
disallow
any
other
network
connectivity.
So
it
you
know
often
the
air
gap
deployments
are
cons.
You
know
thought
of
in
terms
of
being
on-prem,
but
but
it
can
happen
off-prem
as
well.
So
just
you
know
bear
that
in
mind
as
we
go
through
so
asuko.
D
Okay,
I
can't
I
replaced
my
network
with
my
personal
hospital,
of
course,
okay
and
we
can
go
on,
I
I
totally
decided
we
can
use
double
contacts,
a
lot
of
the
other
image
between
the
kubernetes
worker
node,
but
first
I
think
we
need
to
create
a
ripple
to
sell
our
latest
release.
The
worst
image
with,
for
example,
battery
dot
table
and
the
mercury
linker
to
measuring
is
still
trouble
to
do
this
and
use
the
repo.
D
B
Ctl
yeah
there's
a
number
of
interesting
ideas
to
explore
there
about,
literally
and-
and
we
should
just
as
soon
as
we
say
this
first
thing,
which
is
actually
the
very
first
step
of.
What's
what's
actually
just
I
wouldn't
say
critically,
but
what
is
needed
here
from
a
few
different
users
is
they're,
just
looking
for
a
set
of
manual
instructions
to
go
and
perform
these
tasks
to
download
the
right
software
and
have
it
in
place
so
that
they
can
run
meshri
and
whatever
service
meshes
that
they
want
to
sep.
B
B
Second
to
that,
as
we
do
that
the
the
second
consideration
can
be
okay.
Well,
hey.
How
do
we
is?
Can
we
make
this
more
convenient
for
users?
Is
there
bundling
of
you
know
pre-packaging
of
some
of
that?
Those
external
dependencies?
B
Okay,
okay,
but
I
think
actually,
the
exercise
of
writing
down
the
manual
instructions
has
two
effects
one
is
it
immediately
satisfies
a
couple
of
users
and
then
two?
We
get
clarity
on
the
description
of
the
problem
or
the
set
of
problems
rather,
and
so
that's
not
the
end
of
the
conversation.
But
rather,
if
that's
okay
with
you
as
our
first
goal,
that
will
help
flush
out
and
reinforce
or
not
some
of
the
suggestions
about
what
what
additional
code
or
what
additional
changes
we
might
make
in
the
project.
B
It's
an,
I
think,
it's
an
open
question
like
is
in
par.
Let's
do
put
a
pin
in
that
thought
if
you
would
like.
Maybe
that's
the
right
thing
to
do.
I
I
I
don't
know
I'm
not
suggesting
one
way
or
the
next.
If
you
would,
can
you
scroll
toward
the
bottom
of
the
the
doc
there's
a
comment
that
has
a
hyperlink
to
a
github
issue?
It's
probably
that,
second
to
last
comment.
B
Actually
it's
j
yeah,
it's
not
that
one,
but
it's
in
the
comment
on
the
right
hand,
side
it's
just
above
your
a
comment
right
yeah
that
one
okay.
So
this
is
the
genesis
of
this
particular
topic.
This
is
the
user.
Mr,
mr
polberg,
I
forget
his
first
name.
You
know
nice
gentleman,
he
works
at
a
financial
institution
at
a
bank
and
they
have
policy
that
says
you
know.
Security
is
what
most
important
to
us,
and
so
we
do
like
this
project,
we'd
like
to
run
it
in
our
internal
systems.
B
But
you
know
we
don't
let
those
systems
talk
to
anything
else.
So
so
hey.
How
do
we
do
that?
Can
you
tell
us
so
it's
like?
Oh
okay,
yeah.
We
look
it's
a
great
question,
we'd
love
to
tell
you
there's.
We
need
to
write
down
those
instructions
for
you,
and
so
they
have
you
know
so.
There's
there's
some
commonplace
things
that
enterprises
like
this
or
users
like
that
organizations
like
this
will
have
they'll,
usually
a
lot
of
times.
B
But
the
point
is:
they've
got
their
own
non-public
registry
container
image
registry
such
that
they're
in
control
of
the
supply
chain
for
image
for
software
in
this
case
for
container
images,
and
they
can
control
one
what
images
go
in
there
and
how
they're
sanctified,
how
they're
scanned
all
these
things,
the
provenance
of
them,
whatever
they
want
to
do,
and
then
they
facilitate
the
internet
connectivity,
interconnectivity
between
that
registry
and
their
deployment.
So
if
you
scroll
down
a
little
bit
asuka,
there
was
just
you
know.
B
I
was
trying
to
help
this
gentleman
kind
of
think
through
this.
These
are
like
top
of
mind,
considerations
for
different
measuring
components
that
we
have
and
how
we
might
you
know
and
what
instructions
we
may
need
manual
instructions
we
may
need
to
document
and
deliver,
and
it's
through
the
process
of
writing
all
that
down
that
we
might
arrive
at
some.
B
Some
of
the
suggestions
that
you
have
in
that
design
spec
one
of
the
ones
that
that,
for
my
part,
I
consider
has
a
lot
of
merit,
be
in
part
because
we've
already
put
forth
effort
on
this,
and
that
is
the
bundling
the
potential
potentially
producing
alternative
copies
of
the
mesh
readapter
container
images.
B
So
as
mesh
kit
and
the
mastery
adapter
library
was
was
written
and
kind
of
as
work
went,
went
through
there
and
so
asuku
you're.
You
know
this,
I'm
not
telling
you
something
that
you
don't
know,
but
is
that
I
don't
think
we've
leveraged
it
much,
but
but
there
are
the
underpinnings
of
the
ability
to-
and
this
was
this
was
this
functionality
was
actually
in
the
adapters
before
mesh
kit,
and
so
I
think
it
it
hasn't.
B
This
bundle
of
install
files
varies
from
mesh
to
mesh.
Those
can
be
outside
of
the
adapter
on
a
file
system.
That's
or
an
http
server.
That's
readily
available
internal
to
that
deployment.
There's
like
a
bunch
of
options
here
for
how
this
can
get
accomplished
like
what
people
can
do
today
and
that's
what
we
want
to
do
is
like
write
down
the
instructions
to
say
this
is
how
the
adapter
behaves
today,
so
it
will
by
default.
Look
out.
You
know,
reach
out
to
github
reach
out
to
this
place.
B
For
these
you
know,
public-facing,
publicly
downloadable
artifacts
for
a
given
mesh.
Here's
what
you
user
can
do
to
circumvent
that
you
can
go
into
the
adapter
config
and
then
just
tell
it
to
use
a
different
url.
You
can
do
dns
redirect
you
can
write.
You
know,
build
your
own
copy
of
the
messy
adapter
image
that
has
locally
bundled
stuff,
like
there's
a
bunch
of
ways
to
tell
them
to
solve
it.
B
After
we're
all
done
doing
all
that,
I'm
hopeful
that
we
can
then
come
back
and
as
a
community
as
a
project,
we
can
say:
well,
hey
you
know.
Hopefully
it
would
be
something
like
this
like
hey.
We
think
we
could
do
that.
There's
probably
merit
in
taking
that
additional
step
to
assist
and
make
this
a
little
easier
on
those
users.
B
This
this
one
has
the
highest
impact.
It's
the
quickest
thing
to
do.
Let's,
let's
do
whatever
that.
One
is
one
of
those
things
that
you're
pointing
out
is
like
has
to
do
with
measuring
ctl
and
eradicating
external
dependencies.
That
measure
ctl
has
that's
a
good
idea.
Hussein
has
a
separate
design
spec
on
it
and
actually
it's
a
lesser
concern.
B
Measuring
ctl
is
a
client
side
utility
that,
like
often
you
would
find
that
the
user,
the
administrator,
has
downloaded
mastery
ctl
as
a
client
to
like
their
laptop
per
se,
maybe
just
as
just
and
then
maybe
they're,
using
that
laptop
to
connect
to
that
disconnected
system.
You
know
the
air
gap
system
and
so
mesri
ctl.
It
might
be,
have
the
ability
to
kind
of
break.
You
know
reach
out
to
the
internet
because
that
laptop
might
be
able
to
do
it.
B
Maybe
maybe
maybe
not
depends
on
the
environment,
but
but
some
of
those
localized
concerns
some
of
those
concerns
from
s3ctl,
not
as
significant
as
the
concerns
and
the
heartache
that
goes
into
an
air
gap
measuring
server
and
adapter
deployment,
so
yeah.
So
as
you
consider
like
even
measuring
when
it
deploys
itself
and
it
wants
to
deploy
the
meshi
operator
onto
kubernetes
or
the
adapters
on
the
kubernetes.
Those
containers
are
those
pre-cached.
Do
those
need
to
be
pre-pulled
into
the
kubernetes
cluster
nodes,
or
should
those
be
in
a
private
repository?
B
And
if
so,
how
do
people
configure
a
private
repo?
Is
that
something
that
you
can
do
from
sri
ctl,
or
how
does
it,
and
so
really
it's
through
walking
through
all
that
that
we
can
intelligibly
come
to
a
place
where
we
would
have
a
good
suggestion
on,
or
my
perspective
is
it's
after
having
done
that
work
that
we
would
be
able
to
step
back
and
assess
and
say
here's
a
couple
of
things
that
we
want
to
help
automate.
B
But
asuka
I
didn't
mean
to
you
know
what
what
do
you?
What
are
your
thoughts?
B
B
Nice,
nice,
okay,
yeah,
did
anyone
else,
get
a
chance
to
read
the
or
have
does
anyone
else
have
thoughts
on
this
topic
for
asuko?
It's,
it's
really
important.
There's
there's
some
significant
users
that
get
blocked
by
this,
and
so
this
is
a
great.
This
is
by
the
way
for
anyone
who
sometimes
people
come
to
the
community
and
they
say
I'd
like
to
do
devops
things
which,
which
I
don't
know
what
that
means,
since
that's
a
cultural
thing,
but
they
I
think
they
mean
like.
B
D
We
can
we
can
find
some
examples.
I
remember
ranger
aircraft
and
maybe
maybe
I
can
copy
some
ideas
from
renter
air
gap
deployment
and
there
are
made
two
states.
I
think
we
first
we
can
use
the
simple
script
scripts
to
achieve
this
goal
and
some
like
a
load
image
to
the
work,
node
and
apply
deployments
of
measure
and
material
adapter
and
the
stage
tool.
We
can
implement
those
features
with
column
code
and
add
this
feature
to
the
measuring
cdl.
B
It
sounds
great
to
your
point,
the
pullberg,
the
the
user
that
had
requested
these
instructions.
He
had
done
exactly
what
you're
saying
he
had.
He
had
pointed
to
the
longhorn
project,
which
is
actually
from
a
rancher
sort
of
a
rancher
stewarded
project
and
he'd,
given
a
link
to
their
air
gap,
instructions
and
so
yeah.
B
Okay,
I
see
you,
you
see
it
already
so
yeah
that
sounds
perfect
and
part
of
this,
I
guess
is,
since
mr
gefeller
is
with
us
as
well
I'll
ask
on
that
one
item
of
potentially
taking
action
toward
bundling
service
mesh
installation
files
in
an
image
in
in
adapter
container
images.
B
Is
that
a
far
fetched
is
that
do
we
know
if
those
installation
files
are
generally
way
too
thick,
but
hey?
Are
we
gonna?
We're
gonna
have
to
each
of
those
service
meshes.
Have
any
number
of
other
container
images
anyway.
So
are
we
talking
about
bundling
in
their
container
images
into
our
image,
like
probably
not,
and
so.
F
F
I
don't
have
exact
figures,
but
I
don't
think
that
would
be
too
big
a
bit
more
sort
of
like
and
the
interesting
challenge
would
be.
What
do
we
do
about
docker
images.
F
F
How
do
we
define
out
this
aircraft
environment
and
what
are
the
possibilities
like
usually
my
the
developer
pc
has
access
to
the
internet
and
then
the
developer
pc
is
allowed
to
copy
over
to
the
network
that
has
not
access
to
the
internet
and
so
that
we
describe
this
in
maybe
a
bit
more
detail
and
and
also
a
couple
of
other
prerequisites
or
things
we
think
should
be
in
place
like
you
should
maybe
have
a
docker
registry
in
that
network.
F
F
B
Totally
like,
like
hey,
you
know
that
we
assume
that
the
developer
laptop
or
the
engineer
the
laptop
has
there's
connectivity
here.
We
assume
that
there's
a
private
registry
or
like
and
and
yeah,
I
think
that
it's
a
large
enough
consideration
that
and
really
we
don't
need
to
and
there's
enough
work
for
us
to
do,
that.
We
don't
need
to
be
embarrassed
about
the
litany
of
manual
steps
or
whatever
those
are
that
people
need
to
account
like
and
they're
they're
accustomed
to
it.
B
You
know
this
gentleman
pullberg
was
very
forthright
saying:
hey,
I
that's
my
cross
to
bear
I'll,
do
whatever
the
instructions
say,
I'm
totally
used
to
that.
I
just
need
you
to
tell
me
what
those
instructions
are
and
so,
and
so
it
in
it
yeah
and
undoubtedly
it
is.
It
is
something
like
yeah
anyway.
Let's
not
so,
let's
not
solve
it,
but
but
that's
the
yeah
diagram
can
be
enlightening
and
then
maybe
there's
two
scenario.
Maybe
we'll
write
up
instructions
for
two
scenarios-
and
you
know.
D
For
the
doctor
image,
could
we
use
this
repo?
I
have
a
demo
and
I'll
send
the
link
to
the
chat.
As
you
can
see
from
the
readme.
I
send
three
sensory
image
to
the
ripple
and
we
can
configure
the
repo
ip
at
the
remote
machines,
type
dot,
tl
tst
and
run
this
with
ansible.
But
I
think
this
is
the
early
stage,
because
I
do
not
want
to
import
ncbo
script
to
the
measure
cdl
and
I
I
said
before
we
can
use
maybe.
D
B
Yep
yeah
make
it
make
sense.
Yeah,
providing
people
with
tools
that,
like
immediately,
providing
or
suggesting
or
providing
people
with
external
tools
like
this
one
is,
is
good,
like
there's
a
few
different
tools
that
help
people,
pre-load
images
and
pull
automate
the
pulling
of
images
and
even
in
the
dissemination
of
those
images
across
clusters.
Or
what
have
you
like,
there's
yeah,
so
so
I'm
sensitive
to
the
the
time.
I
think
we
characterize
what
it
is
that
we
need
to
see.
A
C
All
right
two
seconds:
can
you
see
that
yeah
yeah
right
so
for
context?
Configuration
wizard
is
a
project
that
myself
vijay
and
austin's
been
working
on
for
the
past
few
months,
or
so
we've
been
making
progress.
How
we've
been
working
on
it
is
austin's
been
providing
us
the
designs
in
figma
which
are
here,
and
then
I've
been
building
the
ui
in
react
and
updating
it
in
the
measuring
itself.
The
reason
I've
built
the
prototype
is
the
development
server
takes
quite
a
while
to
look
to
reload.
C
C
C
And
then
the
service
meshes
we're
not
sure
about
the
advanced
settings.
Just
here.
I've
added
the
button
there
as
it's
on
the
design,
but
we're
still
in
discussions
whether
we'll
need
the
advanced
settings
or
not,
and
then
we've
got
the
external
screen
and
then
the
user
will
be
able
to
navigate
either
to
the
dashboard
start
over
if
they
want
to
or
not.
C
There
we
are
so
our
next
steps
for
us
are
connecting
it
all
together,
obviously,
updating
design
and
connecting
it
also.
So
it
works.
The
only
question
I
have
is:
how
do
you
get
the
react
application
to
talk
to
go?
Vijay
put
this
question:
for
me,
my
assumption
is
somehow
through
redux,
as
I've
noticed
a
few
redux
files,
but
I
mean
I'm
not
entirely
sure.
I
don't
know
if
you'd
be
able
to
help
us
with
that.
E
All
of
its
requirements,
but
I
suppose
for
most
of
them
we
already
have
yes,
for
example,
toggling
operator
or
connecting
to
basically
define
service
meshes,
I'm
not
sure
if
those
toggle
buttons
service,
which
is
always
for
connecting
or
disconnecting
with
adapters,
because
I
don't
think
that's
it
so
in
shorter,
it's
like
the
half
industry.
Server
exposes
a
lot
of
apis,
and
this
spotted.
B
And
this
is
one
of
the
steps
that
doesn't
have
an
advanced
setting
so
like
the
only
thing
for
the
user
to
do
here
really
is
flip
on
if
the
switch
isn't
on
like
flip
on
the
switch
and
sort
of
wait
for
confirmation
that
mastery
operator
is
good
to
go
and
they
can
advance
to
the
next
step
so
that,
in
what
I
just
said,
implies
a
couple
of
things
that
the
first
thing
that
it
implies
is
that
when
the
switch
moves
from
off
to
on
there's
a
rest
api
invocation
there
and
that
can
literally
be
the
same
invocation
that
occurs
bart.
B
B
C
So
we
could
almost
kind
of
go
off
the
normal
settings
page
and
then
kind
of
use
that
as
a
guide
to
connect
the
configuration
wizards.
B
Yeah,
absolutely
in
a
lot
of
respect,
while
some
of
what
I'm
about
to
say
is
true
today,
it
will
be
even
more
true
in
the
future,
where
I
think
it's
the
next
measure
we
release
where
it
might
have
been
this
one,
but
where
we're
trying
to
expound
on
the
concept
of
an
environment,
to
the
extent
that
a
user
could
load
up,
one
cube,
config
and
another
one
and
another
one
and
like
and
get
kind
of
complicated
with
how
many
kubernetes
clusters
they're
talking
to
and
what
context
and
like,
and
they
can
shoot
their
themselves
on
the
foot
here,
we'll
try
not
to
let
them
do
that,
but
sometimes
users
want
their
cake
and
want
to
eat
it
too,
and
so
they
have
to
you
know
so,
we'll
let
them
shoot
themselves
in
the
foot
potentially
we'll.
B
Let
them
get
themselves
all
tangled
up
inside
this
advanced
area
in
the
config
wizard.
However,
we're
going
to
continue
to
say
something
like
well,
no
look
hey!
This
is
just
to
get
you
up
and
going
quickly
to
help
you
help
the
project
achieve
quick
time
to
value
and
to
help
you
help
like
guarantee
a
user.
B
If
you're
having
problems
with
measuring
and
and
one
of
you
in
the
slack
says,
did
you
run
the
configuration
wizard
like
guaranteed
if
they
step
through
each
of
those
four
steps,
they're
good
to
go
like
like
that's
the
kind
of
thing
that
we're
so
this
configuration
wizard
is
to
guarantee
that
they're
successful
is
to
not
let
them
shoot
themselves
in
the
foot,
it's
to
be
very
simple
and
not
like
some
people
may
say
the
measure
you
what
the
mystery
operator.
So
it
was.
B
I
don't
know
what
that
is,
but
there's
just
one
button
to
click
anyway,
so
it
doesn't
really
matter
I'll,
just
click
it
and
wait
until
the
next
button
to
light
up
and
great.
This
is
like
that
very
simple
you're
installing
a
piece
of
software
and
you're
you're
clicking
next
next
next,
next
you're
agreeing
to
conditions
and
terms
you're,
not
reading
them,
you're
just
like
next
next
thing:
that's
what
you
want
to
do
in
the
config
wizard,
and
so
yes,
very
much
so
what
you
said.
B
C
C
Yeah,
I
didn't
really
think
about
to
use
the
settings,
but
like
now
now
you've
mentioned
it
makes
me
feel
about
80
better.
So
it
makes
things
much
easier.
B
And
most
often,
then,
what
would
be
the
case
for
the
advanced
button
in
the
configuration
wizard
would
be
that
I
don't
know
if
it's
quick
enough
for
you
to
give
an
ex
like
if
you
jump
back
into
the
third
step
with
the
advanced
button,
the
advanced
buttons.
Will
you
it'll
be
more
or
less
like
a
one-to-one
mapping,
almost
from
yeah
like
if
the
user
were
to
click
that,
where
would
they
go,
they
would
go
to
settings
and
they
would
go
to
that
second
tab
where
they
can
type
in
their
own
like
here.
B
B
Vj
and
austin,
whom,
like
those
those
things,
these
are
not
configurable.
These
are
just
switches,
that's
it
and
you
can
turn
it
on
and
you
get
confirmation
and
and
that's
it
and
if
that
won't
turn
on
for
you
and
you
don't
get
confirmation,
then
you
need
to
go
to
advanced
settings
to
go
figure
out.
Something
else
like
this.
Isn't
this
isn't
the
place
to
enter
in
an
advanced
like
different
url
to
your
machines,
but
point
for
bart
being
that
the
advanced
settings
it'll
always
just
take
you
into
settings
to
the
respective
page.
G
Yes,
we
actually,
we
spoke
about
these
things
today
and
we
also
stepped
through
that
code.
You
know
in
business
studio
code
and
then
you
know
we
went
to
that
particular
place
where
this
is
happening
in
in
in
mishri
right
now,
in
the
code,
where
it's
doing
these
things
and
and
so
we
are
actually
planning
to
scavenge-
or
you
know
at
least
a
call,
some
things
where
possible-
or
you
know-
or
you
know.
G
Use
modified
versions
of
the
of
that
code
if
we
can't
reuse
the
same
thing.
B
Yeah
totally
both
for
the
components
as
well
as
if
we
can
not
write
any
golang
here
that
would
be
beautiful.
Like
that's
ideal,
there
will
only
be
risk
in
writing
any
new
goaling.
All
the
logic
should
be
accounted
for
elsewhere,
and
if
it
isn't
then-
and
we
do
need
to
write
golang-
that
probably
needs
to
go
into
the
existing
endpoints.
G
Right
I
identified
some
of
the
places.
The
the
whole
thing
is
is
not
is
not
ready
yet,
but
what
what
the
the
next
step
is?
You
know
there
are
little
things
you
know
like,
like
you
said,
you
know
no
url
or
context,
and
the
other
thing
is
to
how
do
you
highlight
when
something
is
connected
or
not
connected?
I
would
say
connected,
or
are
you
going
to
put
a
green
button
or
a
red
button,
and
I
I'm
not
instead
of
having
that
url
in
the
context.
G
If
the
switch
is
you
know,
flipped,
it's
possible,
it
may
connect
properly
or
it
may
not
connect
connect
properly
right
so
either
you
want
to
say
that
it
instead
of
context
there
you're
going
to
take
out
the
context,
because
that's
not
relevant
anyway,
but
see
those
two
green
buttons
are
there.
G
So
if,
if
the
control
plane
two
is
say
you,
they
would
know
that
it
connected
right,
but
if
you,
if
in
the
case
of
grafana
and
prometheus
you're,
not
going
to
get
any
output
from
that,
so
you
have
to
have
a
visual
thing
to
to
actually.
G
B
Yeah
yeah,
I
think,
there's
a
graphql,
endpoint
or
there's
a
set
of
emerging
graphql
endpoints.
That
will
be
quite
helpful
in
understanding
the
state.
B
You
know
understanding
whether
or
not
the
thing
that
switched
on
green
is,
in
fact,
green
and
so
yeah
as
you
switch
it
on
green,
and
you
show
the
confirmation
details
which,
by
the
way,
austin
there's-
and
I
know
part
of
this-
is
between
the
design
and
the
implementation,
but
but
for
now,
there's
only
there's
really
only
like
two
things
that
we're
trying
to
create
that
are
kind
of
new
one.
Is
the
stepper
design
at
the
top?
B
The
other
one
is
the
card
design
for
switching
things
on
and
off,
but
otherwise
the
rest
of
it,
I
think
is
about,
is
about
the
same.
There
is
that
label
for
context
and
really
like
anyway,
we're
trying
to
do
it
more
the
same
now
than
we
are
different.
We
do
need
an
uplift
of
all
of
mesher's
ui,
but
we
want
to
try
to
do
it
uniformly
as
we
go.
B
The
thing
I
was
going
to
say
about
the
graphql
endpoints
is
that
those
are
our
friend
in
terms
of
subscribing
for,
and
a
notification
back
that
like
went
so
when
bart,
when
you're
demoing
the
you're
walking
through
this
and
you're
clicking
you're
turning
things
on
and
clicking
next,
it's
like
well
eventually
you'll
have
lot
and
it
can
be
client-side
and
that's
fine
but
logic
client-side.
That
says
the
next
button
is
not
lit
up
and
you
actually
it's
disabled.
B
You
can't
click
it
until
I've
received
back
that
I've
either
pulled
the
get
the
the
rest
endpoint
for
a
particular
component
and
it
responding
back
that
like
oh,
yes,
this
component
is
alive
and
ready
and
there
are
small
ping
commands
that
are
linked
in
the
chat.
The
there's
a
spreadsheet
that
looks.
F
G
Another
question
is:
should
all
the
service
measures
should,
should
they
try
to
connect
all
of
them
or
is
it?
Is
it
okay
for
some
of
them
fail?
How
does
that
at
the
end
of
the
configuration
wizard?
Does
everything
have
to
be
perfect.
B
No,
the
last
two
steps
are
optional,
so
only
the
first
two
steps
are
required:
okay,.
B
And
really
like
there's
one
we
should
probably
make-
and
I
won't
do
it
now,
but
just
probably
make
I'll
probably
make
clear
the
point
of
probably
confusion
over
the
third
step.
Is
that
connecting
to
measuring
adapters,
or
is
that
deploying
a
service
mesh?
And
so
we
should
probably
follow
up
on
on
that,
because,
right
now
it
looks
like
it's
connecting
to
adapters
and
that's
fine.
B
It's
it's
a.
We
should
start
there.
It's
going
to
be
harder
for
you
guys
to
get
to
the
point
where
you're
deploying
a
mesh,
and
so
but
what
I
don't
see.
I
don't
think
I've
seen
in
the
designs
either.
Austin
is
like
a
good
design
on
how
it
is
that
there's
either
a
carousel
or
a
grid
or
something
to
how
it
is
that
people
are
navigating
through
what
would
be
a
lot
of
like
10
different
service
mission
cards.
B
Also
the
confirmation
details
are,
I
don't
know
that
they're
consistently
and
statically
placed
in
the
same
position,
but
we
just
want
uniformity
across
each
of
the
steps,
to
the
extent
that
we
can
there's
an
area
for
turning
things
on
turning
things
off
an
area
for
inputting
something
like
specifically
there
in
the
kubernetes.
You
can
input
something
and
then
the
other
part
is
really
consistently
placing
these
confirmation
details
and
that's
it
like
it's
very
what
we
do
this
is.
B
We
want
to
take
all
the
all
the
thinking
out
of
it
for
the
users
that
are
using
this.
We
want
for
advanced
users
to
look
at
this
and
like
scoff
at
it
and
say,
I'm
gonna
go
use
the
the
advanced
thing
and
we
expect
a
lot
of
traction
and
people
using
this.
Quite
a
bit
and
it'll
be
great
in
demos
and
great
and
like
really
it'll
really
really
help.
People
really
well,
and
we
really
will
ask
people.
Did
you
run
through
the
configuration
wizard?
Okay.
C
B
Other
stuff
for
you,
as
you
guys,
are
going
through
and
connecting
to
endpoints,
eventually
bart
you'll,
see
that
you'll
get
notifications
back
in
the
notification
center
and
there'll
be
a
little
bit
of
that
notification.
Center
itself
is
a
whole,
a
whole
nice
project
to
uplift,
it's
separate
from
the
config
wizard,
but
so
cool,
so
we
have
10
minutes
left
left
or
less
than
that.
I
love
hearing.
You
guys
are
your
meeting
you're
advancing!
Thank
you
for
demoing
the
progress.
This
is
great.
B
I'm
excited
about
this,
but
we
only
have
eight
minutes
left
for
push.
B
H
H
Okay,
so
that's
how
it
works
just
to
check
that
it
doesn't
work
when
there
isn't.
There
is
no
cluster.
So
let
me
delete
the
cluster
right.
H
H
By
the
way,
just
two
like
before,
like
this,
I
was
doing
it
like
in
a
very
simple,
simpler
way.
Then
lee
and
amanda
told
me
to
do
it
with
client
client
go.
So
I
took
some
inspiration
from
linkard
and
that's
how
I
ended
up
with
this
new
revamped
code.
H
A
It
feedback
for
ph,
I
I
I
have
a
question
like
is-
is
is
necessary
for
like
for
this
check
to
pass,
or
is
it
optional.
H
It's
optional.
You
mentioned
in
the
issue
that
it
should
be
an
optional.
It
should
be
just
a
warning
that
docker
is
not
installed
here.
So
let
me
show
you
the
code
very
close,
yeah
yeah,
so
I'm
just
logging
in
case
of
docker,
but
I'm
throwing
errors
in
all
the
other
cases.
A
D
D
H
D
So,
for
the
no
permission
situation,
what
should
we
do.
H
B
Let
me
let
me
bridge
the
english
divide,
if
I
could,
which
is
to
say
I
think
asuko
is
being
quite
polite
and
actually
what
he's?
I
I
think
we
can
like
change
the
this
warning
here.
Then.
B
Not
exactly
yeah,
yes,
and
no
here,
let
me
give
you
a
better
suggestion
and
that
is
so
push.
This
is
exciting,
you're
doing
good.
This
is
great.
It's
nice
to
see
mesh
kit
being
incorporated
as
well
lots
of
lots
of
very
nice
and
positive
and
uplifting
things
to
say
here,
but
also
let
let
let
me
beat
you
up.
I
guess
instead
of
not
beat
you
down,
but
beyond,
there
are
who
have
a
chat
with
lucena.
B
I
think
she'll
probably
highlight
to
you
that
there's
one
of
the
reasons
why
we
need
to
break
out
the
is
it
the
utils
and
the
helper
functions
and
things
that
we've
got
kind
of
ad
hoc
thrown
into
the
kitchen
sink
the
proverbial
kitchen
sink
over
there,
that
there
are
checks
already
that
more
intelligibly
understand
whether
or
not
something
like
whether
or
not
pseudo
is
necessary
or
not,
or
whether
or
not
like,
there's
actually
a
number
of
these
checks
that
are
already
accounted
for
today,
and
that
this
is
in
a
good
way.
C
B
Bite
you
in
the
rear
end
because
they're
each
going
to
have
like
well
what
what
were
you
using
a
barrier
token
to
check
the
api,
kubernetes
api
or
the
cert,
or
what,
if
it
was
on
gcp
and
then
you
didn't,
have
the
blah
blah
blah
like
oh
my
gosh,
these
things
are-
and
this
is
actually
why
I
was
on
the
side
saying.
Maybe
it's
good-
that
you
started
with
this,
I'm
afraid
we
may
have
cracked
open
a
bigger
can
of
worms
than
then
is
the
priority
for
us
yeah.
B
C
B
Is
going
to
be
really
good,
this
is
going
to
be
the
other
thing.
People
are
going
to
say
we're
going
to
say.
If
they're
in
the
ui
we're
going
to
say,
did
you
run
the
config
wizard
and
if
they're
not
in
the
ui,
yet
we're
going
to
say?
Did
you
run
the
system
check?
A
A
B
Anybody
object
to
making
a
dot
release
a
5.3.
E
B
Got
you
okay
yeah,
we
will
hold
hold
until
you
signal
yeah
good
point
about
the
experimental
section.
Oh
speaking
of
experimental
things,
there
was
a
there
was
a
security
briefly,
since
this
is
the
measuring
development
call.
There
was
a
security
vulnerability
that
was
recently
reported.
The
individual's
name
is
lee.
Wang.
B
The
diligence
of
writing
it
up
very
well
and
documenting
it,
and
the
maintainers
here
have
been
aware
of
it
and
actually
utkarsh
corrected
it
so
with
under
a
week's
time
of
like
our
first
vulnerability,
officially
being
reported
and
then
correcting
that
the
vulnerability
doesn't
affect
any
any
user
or
anyone
here,
there's
no
there's
no
personal
information
or
no
there's,
not
none
of
but
it,
but
we
ended
up
creating
a
cve
id,
which
is
to
say,
like
we
ended
up
officiating
the
fact
that
there
was
a
an
issue
of
vulnerability
in
an
experimental
api,
and
so
it's
good
for
us
to
see
that
process
and
practice.
B
B
So
I'm
should
never
say
that
I'm
tickled
to
see
that
we
have
a
security
vulnerability,
but
I
just
tickled
that,
like
you,
know
it's
of
interest
to
to
people-
or
you
know
the
project
in
that
way,
so
a
couple
of
things
as
we
all
go
to
get
off
the
call.
I
want
to
say
that
we're
going
to
talk
about
a
couple
of
things
in
the
community
meeting
this
week.
I
hope
so
one
is
you
know
it's.
B
We
use
the
community
meeting
for
a
lot
of
things
and
some
of
those
are
one
of
those
things
you
can
be
talking
about
individuals
that
have
been
nominated
for
maintainership
there's
a
topic
there,
there's
probably
a
topic
on
roadmap
a
bit
like
and
we've
got
roadmap,
there's
a
publicly
published
one
and
a
markdown
file
we've
also
for
for
measuring
right.
We've.
Also,
I
mean
the
other
projects,
s
p
and
get
direct.
They
have
their
own
roadmaps,
but
then
there's
a
there's,
a
more
expansive
roadmap.
B
B
There
was
one
more
thing
I
forget,
but
but
yeah
yeah,
oh
yeah.
The
other
thing
is
donation
to
the
cncf,
so
there's
been
so
we'll
talk
about
that.
There's
there's
feedback
on
that.
We'll
talk
about
that
at
the
community
call.
A
Okay
over
time,
we
missed
a
couple
of
items,
so
aditya
recently
put
up
a
pr
to
which
will
actually
display
the
measuring
endpoints,
which
may
be
different
specific
to
platform.
A
So
I
guess
we
we
will
wind
up
the
call
tank,
buddy
and
stealing
military.