►
From YouTube: Cluster API Azure - Release - v0.2.1
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
Found
all
right
well,
hello,
everyone
who
may
be
watching
this
later.
What
we
wanted
to
do
today
was
take
an
opportunity
to
people
have
been
curious
about
cluster
API
Azure
and
how
it
works,
and
all
those
good
things
and
Timmy
had
made
a
great
suggestion
of
checking
out
what
it
would
look
like
to
like
actually
release
cluster
API
measure.
So
we
have
a
rare
opportunity
and
rare
opportunities,
and
there
is
a
bug
fix
that
needs
to
happen.
A
So
I'm
gonna
take
that
opportunity
to
record
a
quick
session
about
how
to
do
this
right
so
without
further
ado,
I,
just
clear
the
screen
and
get
started
all
right.
So
first,
first
I'm
just
going
to
open
up
the
repo
in
code
and
we've
written
some
instructions
on
how
to
do
the
release
process.
The
instructions
will
need
to
be
updated
because
I
have
like
a
little
crib
sheet
of
what
needs
to
happen.
Instead
in
certain
steps,
there
is
a
pre-flight
manual
which
is
kind
of
like
hey.
A
If
there
are
any
modifications
to
the
release
policy,
you
should
take
a
look
at
this.
You
should
create
a
release
commit.
This
is
no
longer
true.
I'm
not
going
to
do
the
the
release
commits
I
think
that's
kind
of
a
waste,
because
we
then
have
to
wait
for
someone
to
approve
the
release
commit
as
opposed
to
just
cutting
on
cutting
on
any
one
specific
sha
write
the
release
tool
that
is
in
the
repo
right
now
is
kind
of
a
copypasta
mess.
A
I,
don't
personally
use
it
I,
just
so
we're
going
to
go
through
the
we're
going
to
go
through
the
manual
steps
right
and
I'm
gonna
I'm
gonna
do
some
improv
jazz
on
the
manual
steps
to
get
this
done.
So
the
first
thing
that
we
want
to
do
is
tag
the
repo
and
push
the
tag,
but
before
we
do
that,
we
should
make
sure
that
we
are
one
checking
out
master
to
me.
Can
you
see
this
okay?
It
should
a
I'm
gonna
pop
it
1-1,
bigger,
that's,
perfect!
A
A
Alright,
the
reason
for
that
was
a
bug
that
was
introduced
in
the
zero
one,
two
version
that
I
think
Timmy
and
I
simultaneously
discovered,
as
well
as
some
people
upstream,
that
was
quickly
fixed,
so
had
tip
to
Justin
Santa
Barbara
for
pushing
the
new
image.
For
that.
If
you
want
to
read
more
about
it,
it
is
cluster
API,
$9.99
controller
Mitch
fails
to
run.
You
can
see.
We
popped
in
here,
so
part
of
it
had
to
do
with
some
cash
image
layers
that
were
on
Justin's
local
as
he
was
creating
the
image.
A
A
A
So
let's
get
started
right.
We
are
going
to
what
did
we
say
we
were
going
to
do
we're
going
to
tag
it
right,
so
get
tagged
s
alright,
yeah
I'm,
going
to
tag
this
as
zero
to
one?
That's!
What
we're
releasing
today
and
I
don't
go
crazy
with
these
messages,
because
they
don't
get
seen
just
adding
zero
to
one
there
right
right.
A
A
Spell
me
my
username
correctly
and
I'm
going
to
stop
sharing
for
a
second,
so
I
can
pull
up
my
password
manager
and
copy
my
github
token.
So
if
you
have
ever
used
github
on
the
command
line
and
you
have
to
factor
authentication
enabled
as
you
should,
if
you're
a
part
of
a
kubernetes
org,
the
requirement
is
that
you
also
generate
a
token
right
to
do
releases
all
so
can
to
do
github
operations
on
the
command
line
right.
A
B
A
Okay,
so
you
can
see
that
that
tag
is
pushed
and
if
we
switch
back
over
to
the
cluster
API
edger
repo,
we
can
see
that
as
well.
Release
is
right
and
zero
to
one.
That's
pushed
a
alright,
so
next
step
is
create
a
draft
release
and
github
and
associated
with
the
tag
that
was
just
created
right,
so
draft
new
release
the
zero
to
one
loop
and
we're
going
to
call
the
release
the
same
thing
I'm
going
to
leave
this
saved
as
a
draft
for
now
and
we'll
get
back
to
that
in
a
little
bit
right.
A
So
now,
I'm
going
to
actually
check
out
the
tag
like
I'm
technically,
where
I
need
to
be
already.
If
we
just
see
there
right
but
for
sanity
sake,
we're
going
to
check
out
that
tag
and
detach
head
right,
so
heads
attached
we're
actually
on
the
tag.
Now
it's
the
same
commit
doesn't
really
matter,
but
you
know
so
now
we're
going
to
run,
make
release
artifacts
right.
So
if
you
want
to
see
what
make
release
artifacts
is
doing.
A
We're
tearing
up
this
information
right.
So
any
questions
to
me
so
far:
alright,
cool
cool
cool
cool
right
so
to
look
at
a
little
bit
more
of
what
this
is
doing,
right
we're
doing
cluster
CTL.
So
this
there's
a
version.
There's
a
version
info
package
here
that
has
some
information
about.
What's
built,
how
it's
built
point
in
time,
information
for
when
you're
cutting
that.
A
So
nothing
too
crazy.
I
know
not.
Everyone
speaks
basil
and
I
I
barely
do
as
well,
so
I
won't
go
to
go
into
any
of
the
basil,
isms
I.
Don't
think
they're
required
for
this
specifically,
but
alright
so
back
to
our
instructions.
What
were
we
doing
right?
We're
going
to
run
that
make
make
release
artifacts,
and
this
will
generate
that
output
folder,
that
we
were
talking
about
and
I'm
just
going
to
do
a
time
on
a.
A
So
Basil's
off
doing
its
thing,
doing
a
compile
of
the
cluster
CTL
binary,
which
is
one
of
the
assets
that
we
publish
in
the
repo
when
we
cut
it.
Release
and
I
mentioned
that
I'm
going
to
do
some
improvisational
jazz
on
what
the
instructions
are
written
for
the
release
process.
I
do
owe
an
update
for
some
of
for
some
of
the
instructions.
So
I
know
the
things
that
shouldn't
happen
so
I'm
going
to
skip
the
things
that
will
go
bad
by
following
like
a
following
it
step
by
step.
C
A
Some
of
the
tagging
process-
and
you
know,
we've
done
a
we've-
done
a
draft
release
after
pushing
the
tag
and
we're
actually
generating
the
binaries
right
now,
which
will
attach
to
the
draft
release
and
then
we're
gonna
write.
Some
release
notes
some
janky
ones.
There
I
mean
there
are
only
like
I
think
two
commits
that
we've
done,
since
let's
check
it
out
that
we've
done
since
low
uh-oh,
oh
yeah.
D
A
We
try
to
have
something
of
a
support
policy
and
this
is
kind
of
evolving
as
upstream
works
through
designs
for
a
cluster
API
right.
So
you
know
some
of
this
support
policy
also
needs
to
be
updated.
I
have
a
draft
PR
to
do
some
of
that,
but
essentially,
what
we're
trying
to
do
is
a
sure
providers
hero
to
zero
supports
cluster
API
zero
one
right.
So
this
is
currently
incorrect.
The
PR
fixes
that
right,
but
so
essentially
everything
for
Azure
minor
version.
One
was
everything
to
bring
us
up
to
that.
A
So
our
so
our
minor
two
matches
their
minor
one
right
and
then
also
this
chart
will
probably
expand
its
currently
incorrect
right
now,
because
when
114
has
been
released
but
zero,
zero
should
support
both
114
and
115
as
it
lands.
So
115
is
a
schedule,
tool
tool
and
tentatively
for
next
week,
Monday
right,
Nick,
yes,
okay,.
D
A
We
have
stuff
right.
If
we
look
at
this
output
directory,
we
can
see
that
we
have
the
examples
at
tar,
the
Darwin
and
the
Linux
binaries
right.
So
these
are
the
ones
we
want.
We
are
going
to
assume
that
they
work
for
the
time
being
until
we
test
them
and
I'm
going
to
figure
out
where
my
I
go
directory
is
verse.
B
A
A
A
So
I
didn't
want
to
hold
up
the
opportunity
to
actually
like
stabilize
on
an
upstream
cluster
API
version.
So
I
cut
this
one
right.
We're
cutting
this
one
now
to
fix
this
bug
fix
right
what
we
should
be
able
to
do
what
we
should
feel
feel
comfortable
and
doing
is
exercising
this
process,
as
often
as
needs
needs
to
be
right.
There's
no.
We
don't
currently
have
a
any
specific
release.
Cadence
for
for
cluster
API,
a
sure
right
will
that
change
in
the
future
potentially,
but
right
now
we're
pretty
fine,
alright,
so
the
next
piece.
A
A
Our
logs
okay,
this
is
the
one
right,
so
I
was
so
I
think
the
last
build
I
did
I
was
writing
notes
about
like
okay?
What
are
the
things
that
are
missing
in
in
our
documentation
and
one
of
the
thing,
because
I
built
a
release
and
totally
worked
it
as
I
go
cool?
That's
something.
I
have
to
write
down
now,
so
I
need
to
commit
this
to
the
repo.
So
what
we're
doing
so?
That
I
mean
the
mean
aha,
so
I
should
have
read
this
before
I.
A
Actually,
I
should
have
read
this
before
I
actually
fished
it
which
made
the
artifacts,
because
these
are
no
bueno.
The
reason
for
that
is
that,
during
during
the
release
artifacts
process,
it
does
some.
It
does
some
variable
expansions
and
reads
in
some
of
the
information
about
the
registry,
though
the
pull
policy
and
the
manager
image
right.
So
we're
gonna
do
this
again,
it's
great
that
I
have
that
recorded.
It's
always
fun
to
blunder
on
camera.
A
All
right
so
I'm
swapping
those
version,
numbers
and
I'm
going
to
rerun
this
now
and
the
reason
that
I'm
not
the
reason
that
I'm
not
going
to
fix
those
docs
immediately,
is
because,
if
I
start
making
changes
on
the
docs,
the
git
branch
will
no
longer
be
clean
and
messy
things
will
start
to
happen.
Do
like
the
release
process,
so
I'm
going
to
preempt
that.
A
All
right
so
they're
re-uploading
you
and
while
that's
happening,
we
are
going
to
one
again
using
a
lot
of
the
same
environment
variables
we're
going
to
build
the
docker
image
right.
So
we
have
to.
We
have
to
container
registries
that
we
use
right.
One
is
kate,
staging
slash
cluster
api,
Azure
controller
right,
and
you
can
see
that
these
are
kind
of
funny,
duddy
tags
that
we're
that
we're
looking
at
and
these
this
repo
kate
staging
is
used
for
use
for
development
for
the
main,
maintainer
x'
right,
so
development
of
images.
A
C
A
A
Right
right,
not
specifically
this
one
but
we're
gonna,
sidestep
to
one
more
repo
right,
and
so
these
are
the
official
images
right
mapping
to
each
of
the
version
tags
you
can
see
from
here
from
the
security
scan.
We
went
from
past
unsupported.
The
reason
behind
that
is
that,
right,
a
right
before
we
cut
the
zero-one-zero
release,
I
switched
the
images
to
dis,
realest,
so
dis
realists.
A
There
are
reasons
that
I
don't
know
that
are
unsupported
with
clear
scanning
right.
So
that's
the
reason
that
you're
seeing
that
unsupported
piece,
if
you
are
building
images
yourself,
just
for
people
who
are
listening
to
the
call
I,
probably
should
have
started
the
build
frame
for
people
who
are
listening
to
the
call.
If
you
can
build
on
dis
rellis,
you
should
because
it
better
it's
safer,
lower
attack
surface
for
a
lot
of
things
and
being
able
to
you
know.
For
so.
For
our
sake,
we
really
only
need
a
go
binary
right.
A
A
All
right
so
cig
release
I
was
like
I.
Think
I
approve
this.
Yes
rebase
images
to
destroy
lists
if
anyone's
curious
about
checking
that
out
walks
through
some
of
the
different
types
of
images
that
you
would
encounter
in
kubernetes
and
why
destroy
is
better
and
some
of
the
graduation
criteria,
as
well
as
the
images
that
need
to
be
done
right.
So
all
that
information
is
in
kubernetes
enhancements
the
caps,
repo
and
then
cig
release,
and
then
you
can
find
the
release
images
to
dis,
realest
cap
right,
cool.
A
If
anyone
is
curious
about
what
is
happening
in
that
dockerfile,
really
nothing
crazy,
we're
making
a
work
directory,
we're
copying
in
the
so
when
we're
using
going
image-
and
we
should
switch
this
to
just
going
1:12
so
we're
getting
the
latest
but
copying
in
the
package
command
and
vendor
directory
is
doing
a
build
copying
the
binary
from
the
resultant
build
into
the
image
and
setting
the
entry
point
to
manager
right,
nothing,
crazy
right.
So
that
is
done
and
now
we're
going
to
do
a
push
right-
and
you
see
we've
done
this
before.
B
A
A
Direction
somewhere
right
right
in
front
of
me,
so
we
have
the
image
we
have
the
binaries
right.
We
haven't
written
the
release
notes,
but
because
we
have
the
image
and
we
have
the
binaries,
that
means
we
can
actually
start
to
exercise
the
getting
started
Doc's
right.
So
this
is
a
part
where
we
go
off
and
we
build
build
us
a
kubernetes
on
cluster
API
Azure
right.
So
first
things.
First,
it
says:
get
their
latest
release
right.
A
We
already
have
the
latest
release
locally
and
we
haven't
published
the
binaries
yet
right,
so
we're
going
to
use
the
local
stuff
alright,
so
I
am
going
to
switch
over
to
yet
another
terminal
to
make
that
bigger
right,
so
I'm
just
going
to
copy
out
so
one
I'll
do
a
make
directory
of
projects.
Kate's
cap
c,
0,
2,
0,
right
and
then
I'm
going
to
copy
a
note
star
to
that
last
thing
right
and
then.
B
A
That
last
thing,
so
here
I,
am
all
of
our
fancy.
Stuff
is
here,
I'm
also
going
to
actually
what
do
I
want
to
do?
I
want
to
do
this
going
to
go
back
where
I
was
for
a
second,
because
I
have
something
very
important
there
right,
so
I
have
an
environment
file
here
which
includes
my
credentials
so
I
don't
expose
them
on
screen,
as
I
am
known
to
do
we're
gonna
switch
back,
alright,
alright!
So
let's
get
back
to
those
instructions
right.
A
It
says:
let's
set
up
an
environment,
let's
make
sure
that
the
environment
variables
are
set.
We
already
did
that
now
we're
going
to
take
that
tar
and
untie
it
Boop
all
right
and
we're
gonna
set
some
we're.
Gonna
set
some
settings
not
too
so
I'm
just
gonna
take
this,
and
this
basically
means
like
it's
me.
It's
me
I'm
testing
zero
to
zero,
and
this
is
a
second
time
I'm
doing
it
right.
C
A
A
A
Okay,
so
I've
got
most
of
the
things
that
I
need
in
place
now,
right
and
I
can
copy.
This
and
so
hopefully
anyone
who's
used
these
instructions,
like
it's
simple
enough
like
there
are
no
dollar
signs
or
anything
in
the
in
the
in
the
code
blocks
like
it's
simple
enough
to
just
be
able
to
copy
and
paste
a
lot
of
these
things,
and
that's
kind
of
that
was
my
intent
there.
A
A
Right
and
make
sure
that
I
grab
the
cube
config
from
kind
which
should
be
up
at
this
point,
yep
up
and
then
I
also
want
a
time
of
how
long
it
takes
for
the
pods
that
I
care
about
to
be
up
all
right,
so
you've
got
the
azure.
We've
got
the
cluster
API
controller
manager.
These
are
the
two
ones
I
care
about
really
right.
This
is
the
one
I
really
care
about,
because
this
is
the
the
one
with
all
the
important
cap.
C
blogs
right.
A
All
right
and
and
in
the
background
you
can
see
that
you
know
it's
just
waiting
for
the
machine
to
be
ready
all
right
and
it's
gonna
keep
spitting
this
out
until
we're
we're
done
really
all
right.
So
Azure,
that's
the
one
I
care
about
all
right,
so
it
looks
like
it
took
about
a
minute
and
19
seconds.
That's
cool!
That's
about
average
and
now
we're
gonna
tail.
The
logs.
A
All
right
so
people
who
are
watching
this
and
are
worried
about
these
errors
that
pop
up
first,
it's
checking
for
a
resource
group
that
doesn't
exist.
Yet.
That's
absolutely
ok,
remember
that
you
know
in
kubernetes
everything
is
a,
hopefully
everything
as
a
controller
right
in
some
way,
so
so
we're
doing
reconciliation
loops
over
this
information.
A
So
the
second
time
around
we
do
get.
We
do
it
does
a
check
for
the
resource
group.
It
finds
that
it's
not
there
does
the
create
of
the
resource
group
right.
So
the
reason
for
this
is
this:
actuator
that's
running
is
actually
the
machine
actuator
right,
so
the
machine
actuator
handles
information
about
the
machine
right.
So
how
big
should
this
machine
be?
What
is
the
image
type
that
I'm
using
for
the
machine?
What
is
the
resource
group?
The
security
groups
need
to
be
attached
to
it.
A
Things
like
that
right,
the
cluster
actuator,
these
things
both
start
at
the
same
time
right,
so
the
cluster
actuator
is
actually
the
one
that
is
responsible
for
creating
the
resource
group,
the
v-net,
the
the
security
group
you
know
and
and
the
the
route
table
for
the
nodes
right.
So
because
they're
kind
of
running
at
the
same
time,
it's
you
know
the
first
condition
that
the
machine
hits
is
the
fact
that
the
resource
group
is
missing
the
second
time
around.
A
It
will
find
the
resource
group,
because
the
machine
because
the
cluster
actuator
is
running
alongside
and
mul,
will
kind
of
play
catch-up
there
right
right.
So
you
can
see
it's
creating
a
load
balancer
of
subnet
internal
load,
balancer,
bla,
bla,
bla,
bla,
bla
and
all
the
certificates
were
happy
right.
A
E
A
I
had
mentioned
before
that
we
kind
of
used
this,
for
you
can
see
we
kind
of
use
this
for
the
cube
tests,
testing
for
a
right,
and
then
these
are
mine
right.
So
this
is
the
one
we
were
we
care
about
and
we
can
see
one
we've
created
the
security
group
we've
created.
The
resource
group.
Excuse
me
and
we
have
created
a
bunch
of
resources
right
and
so
what's
happening
now
is
the
control
plane
is
probably
our.
We
have
to
startup
script
now,
it's
still
creating
the
control
plane.
The
next
thing
to
happen
is
the.
A
Where
are,
we
is
the
the
deployment
script
that
will
run
to
configure
all
of
the
the
cube
ATM
bits,
the
dependencies
incubating
ATM
bits
right
so
create
creating
a
VM
extension
startup
script,
an
additional
thing.
If
anyone
is
interested
in
taking
this
on,
we
also
need
a
successfully
created
blah
right,
because,
right
now
you
kind
of
have
no
idea
what
happens
here
until
it
fails
right.
A
A
That's
a
that's
a
different
one:
sorry
yeah!
No,
no!
No!
That's
fine
and
essentially,
like
I
I,
think
we
need
to
do
a
better
job
of
closing
the
loop
on
on
different
set
like
we
have
a
loop
right,
so
you
know
there's
a
start
and
an
end
kind
of
where
it's
the
same
place,
but
we
should
say
when
we're
at
the
beginning
of
it
and
when
we're
at
the
end
of
it
right.
A
A
A
A
A
B
A
A
The
error
that
was
outputted
to
the
screen
for
the
portal
exposes
the
private
key
right.
Something
we
need
to
fix
so
make
the
script
a
little
less
brittle.
So
the
what
happened
is
I
think
the
pull
of
CRI
container
D
that
tar
that
download
failed
and
that
download
failure
caused
the
entire
thing
to
fall
apart
right.
So
we
should
do
like
a
curl
with
a
retry
max
5
or
something
like
that
right.
A
We
proceed
like
a
fake
node
name
in
the
sir
just
to
make
sure
that
the
SIRT
thing
passes.
So
we
need
to
do
a
little
edit
of
the
the
timing
for
the
certs
right,
because
the
cert
contains
information
about
the
the
public
IP
fqdn,
and
we
need
that,
for
we
need
that
to
write
the
cert
and
we
won't
have
that
information
until
after
the
public.
Ip
is
actually
generated
right.
So
we
need
the
search
generation
process
to
happen
after
that,
and
then
we
can
remove
this
fake
cert
information.
A
A
So
we
have
wire
wire
contact,
space,
logging
and
scopes
and
sorry
there's
a
dog
barking
outside.
So
apologies
for
that
and
use
logger
everywhere
right.
So
you
can
see
that
there
were
some
updates
done
here
to
the
scope
and
the
updates
that
we're
done
here,
we're
kind
of
just
to
pull
in
this
log
or
information
right
and
clean
up
the
way
that
that's
listed
right
and
then
also
to
switch
to
to
switch
to
use
the
logger
instead
right
instead
of
K
log
right.
So
the
reason
that
we're
doing
this
is.
A
Here
right,
we're
gonna
fix
some
of
the
errors
to
log
our
lager
right.
So
what
we're?
What
what's
changing
here
is
we're
we're
basically
adding
or
adding
some
new
information
to
the
struct
right.
The
the
parameters
for
the
scope
right
now
by
adding
the
logger
that
means
I
can
pass
I
can
pass
a
context
for
the
logs
across
the
entire
scope
right.
So
every
function
or
you
know,
every
method
that
uses
that
uses
a
scope
in
some
way
means
I,
can
pull
out
a
log
or
I
can
pull
out
a
context
for
so.
A
So
the
whole
point
is
to
be
able
to
apply
some
context
across
these
logs
to
let
you
know
what's
happening
also
consider
the
instance
where
you
have
a
management
cluster
that
is
dealing
with
more
than
one
cluster
right.
It
would
be
helpful
if
we
have
like
the
cluster
name
or
something
right
across
the
context
of
the
law.
Great,
so
that's
kind
of
the
purpose
for
adding
some
of
the
stuff
in
we
don't
wire.
A
A
So
we
kind
of
we
continue
that
along
the
cluster
scope
code
right
so
logger
pram
with
name
right.
It's
giving
me
the
API
version,
it's
giving
me
the
name
space
and
it's
giving
me
the
cluster
name
right
so
now
we're
starting
to
get
a
little
bit
more
useful
information
out
of
the
logs
right.
But
the
functions
are
the
methods
that
use
these
scopes
need
to
also
be
wired
to
know
that
they
need
to
access
that
in
additional
information
or
provide
that
additional
information
alright.
A
So
that
is
work
to
be
done
and
part
of
the
reason
that
this
was
this.
This
was
worked
up
was,
for
the
sake
of
using,
for
the
sake
of
adding
the
end-to-end
tests
right.
So
the
reason
that
the
the
end-to-end
tests
are
critical
outside
of
the
obvious
reason
that
they're
critical
in
general
is
that,
with
a
project
with
a
project,
that's
starting
to
gain
traction
and
the
the
smaller
amount
of
maintainer
x'
that
we
have
when
a
big
change
comes
in
the
requirement
around.
A
That
is
essentially
for
me
to
do
exactly
what
we're
doing
right
now
that
the
tests
do
human
smoke
testing
of
of
the
cluster
before
accepting
the
change
right.
So
because
of
that,
it
would
be
a
lot
easier
if
I
could
add
an
end-to-end
test
to
our
test
framework
and
get
signal
out
of
that.
Instead
of
having
to
dedicate
time
to
do
that
right.
So
every
you
know
every
bit
of
time.
I
have
to
dedicate
to
that
means.
We're
not
doing
you
know
we're
not
doing
prioritization
of
the
backlog
or
we're
not
writing
actual
code
right.
A
So
the
hope
is
that
spend
some
time
on
writing
the
intent
tests
so
that
we
can
enable
more
people
to
contribute
frequently.
So
this
is
something
not
to
be
concerned
about
right.
This
end
right
here.
What
this
means
is
that
it's
saying
that
the
container
that
doesn't
exist,
the
reason
the
container
doesn't
exist.
It's
because
if
we
go
over
here,
so
one
we're
updating.
So
one
we've.
A
Finally
gotten
to
the
point
where
the
control
plane
is
up
right
and
it's
saying
I'm
going
to
I'm
going
to
update
the
information
about
that
right
and
it's
you
know
now
creating
the
target
cluster
right.
So
the
target
cluster
is
actually
the
cluster
that
is
out
in
Asia
right.
So
remember
that
we
started
off
there,
two
phases
right.
A
So,
where
created,
it
was
a
sure
right
and
now
that
we
have
the
cluster
that
we
need
we're
now
pivoting
the
assets
that
exist
in
the
kind
cluster
over
to
Azure
right.
So
the
assets
is
in
not
the
quote-unquote
physical,
our
cloud
resources
right,
but
more
so
the
idea
and
the
information
about
them
that
are
contained
within
the
cluster
spec
status
and
the
MANET
and
the
Machine
spec
and
status
right.
So
it's
telling
you
right
here,
I'm
going
to
go.
Look
for
the
list
of
things
that
I
need
to
go.
A
Do
stuff
too
right,
I'm,
going
to
scale
down
the
controller's
that
exist
right
now
right!
So
when
you
saw
that
this
container
is
no
longer
able
to
be
found,
that's
what
was
happening
right.
That
container
is
getting
spun
down
right
and
if
I
was
to,
if
I
was
try
to
do
anything
here
we
can
see.
Connection
to
the
server
is
refused
right.
This
kind
cluster
has
now
been
destroyed
right,
so
the
speech
type
cluster
no
longer
exists
right.
So
it's
gathering
a
list
of
all
of
the
things
that
it
needs
to
move.
A
I'm
gonna,
look
for
the
machine
classes,
nothing
to
do
there,
I'm
going
to
look
for
the
clusters.
I
found
one
cluster
I
have
to
move
that
it's
going
to
read
out
the
stuff.
That's
in
that
cluster
right,
get
the
machine
deployments
right.
No
machine
deployments,
no
Machine
sets
I
found
a
machine,
though
right
so
I
need
to
move
that
machine
over
right.
A
So
now
it's
waiting
for
that
machine
to
become
ready
and
then
it's
doing
the
same
checks
that
same
reconciliation
loop
against
that
machine
that
it
was
doing
before
right
when
it
was
on
the
bootstrap
side.
So
now
that
that's
all
done
right!
So
it's!
It's!
Writing!
It's
writing
the
status.
It's
writing
the
status
to
the
new
target
cluster
right
and
now
it
has
one
more
job
to
do
and
that
job
is
to
create
the
node
right.
A
So
if
we
wanted
to
see
what's
happening
there
at
this
point,
you
have
information
in
your
repo
lit
because
I
can't
spell
you
have
information
in
your
repo
to
get
access
to
that
cluster
right.
So
let's
go
back
to
the
instructions
and
see
if
we
also
see
that
right
all
right
so
waiting
for
that
to
become
ready.
A
You
can
now
access
your
cluster
blah
blah
blah
blah
blah
right
and
then
we
said
if
we
do
the
export
we'll
do
this
in
a
separate
screen,
and
it
will
make
sure
that
this
is
lowercase
right
and
then
I
can
do
a
cluster
CT.
Oh,
oh,
you
know
just
a
few
commands
to
show
you
that
this
thing
is
working
right.
That's
the
information
yay
all
right!
A
A
E
A
A
Right-
and
you
know
there
are
some
there's
some
key
information
in
there,
but
this
will
be
destroyed
in
quick
order
right.
We
don't
have
a
back-end
pool
or
anything
that's
going
on.
There
are
actually
we
might
not
be
writing
the
information
back
to
this
right.
So
this
is
another
potential
for
contribution
right.
This
is
not
super
critical.
We
don't
need
to
pass
this
to
anything
right.
This
is
just
kind
of
information
for
the
user
right
you
can
see.
We
have
some
key
material,
not
a
big
deal.
A
A
A
A
C
A
So
there
I
mean
so
there
are
definitely
some
so
yeah.
So
the
the
thing
that
took
the
longest
really
was
spinning
up
the
cluster
validating
that
that
word
exact
yeah.
So
you
know
if
we
have
and
and
really
what's
you
know
really,
what
part
of
this
is
is
exercising
the
tools
right.
So
we
need
to
figure
out
what
we
can
do
to
exercise
the
tools
right.
So
part
of
the
end-to-end
test
should
be
so
we're
gonna
start
off
like
can
we
build
a
cluster
right?
Can
I
build
a
cluster
object?
A
Can
I
build
a
machine
within
that
cluster
object
right
and
once
we
solve
the
base
case
for
the
the
intent
test,
I
think
that
we
should
expand
it
to
exercising
the
tools
like
exercising
cluster
CTL
right
and
and
that
way,
if
we're
consistently
validating
that
we're
building
the
binaries
and
and
we
can
exercise,
we
can
exercise
the
the
tower
unpacking
turn
and
walking
through
that
entire
process.
Then
we
can
say
that
we
can.
A
We
can
automate
the
release
process
right,
but
there
are
some
parts
where
and
I
think
that
you
know
overall,
is
a
project
we
haven't
figured
out
where
we
want
to
go
in
terms
of
like
provenance
right,
so
image,
provenance
provenance
of
the
of
the
bits
that
we're
we're
we're
providing
as
assets
right.
You
know
so
like
in
in
an
ideal
world
I'm,
also
providing
you
know,
sha-256
sums
of
each
of
these
things
right
and
or
multiple.
A
You
know,
sha-1
sha-256
shot
shot
512
right
of
these
things
right,
so
people
can
validate
against
them
and
then
we
can
also
have
our
tooling
validate
against
those
things,
but
also
you
know
this
is
this.
This
release
is
verified
right.
This
is
this.
Is
GPG
signed
right
so
solving
for
that
right,
it's
not
something
that
you
know
is
trivial
right
and
that's
that's.
You
know.
A
That's
the
reason
that
there
are
bottlenecks
on
some
of
these
things
kind
of
across
the
kubernetes
repos,
because
if
we
are
doing
image
signing
and
stuff
like
that,
then
we
have
to
be
concerned
with.
Who
has
the
keys
to
do
that
right
and
can
we
have
them
available
to
do
the
thing
that
needs
to
be
done
right
push
the
button
to
turn
the
release.
So
anyhow,
you
should
have
received
an
email
or
you
will
receive
an
email
shortly.
I
just
got
one
for
this
new
release.
Yep.
C
No,
it
was
just
useful
to
see
what
that
process
looked
like
I.
Think
that,
like
at
minimum
there's
areas
where
people
can
help
relieve
some
of
the
burden
here,
right,
like
I
know,
I
can
be
a
test
monkey
for
does
a
cluster
e,
something
up
appropriately
right,
like
so
at
least
maybe
I
can
assist
with
something
like
that
and
like
leave
people
who
are
more
equipped
to
handle
the
architecture
of
this
actual
code
base
to
doing
that.
So
that's
kind
of
where
I'm
gonna
try
to
focus
some
of
my
attention.
Yeah.
A
A
We've
done
the
live
release
session.
We've
we've
already
done
a
few
it
both
in
person
and
on
video,
builds
as
cluster
API,
Azure
right.
So
I
think
what
would
be
cool
next
is
to
touch
on
some
of
the
project
management
stuff
of
it
right
and
do
a
grooming
session
right.
Do
a
public
grooming
session
for
for
our
stuff
to
see
where
we're
at,
and
maybe
we.
C
C
A
So
that's
so
as
we
walk
through
these
words
I'm
going
to
be
walking
through
them
with
people
who
do
this
for
a
job
right,
so
it
won't
be
too
bad
to
it
yeah.
But
there
are.
There
are
places
other
places
in
the
project
that
would
be
want
for
project
managers,
so
I
don't
want
to
I,
don't
want
to
spin
them
off
on
this
if
they
don't
need
to
be
cool,
so
I,
don't
know
if
you
have
seen
this
board
before,
but
I,
occasionally
agreement,
that's
it's!
It's
pretty
up-to-date.
A
All
right,
so
that
is
fairly
up
to
date,
if
you're,
if
you
filter
on,
if
you
click
any
one
of
the
tags,
if
you
click
one
any
one
of
the
priority
tags
right,
you
can
see
that
stuff.
That
is
priority
important
soon,
is
kind
of
either
in
the
T
or
in
progress,
columns,
right
and
then
the
stuff
that
is
priority,
important
long
term.
You
can
see
there
are
mainly
in
the
backlog
and
to
do
right.
So
there's
you
know,
there's
a
fair
amount
of
prioritization.
That's
gone
into
this
work
already.
A
We
just
need
to
make
sure
we
keep
up
the
pace
and
and
start
doling
out
some
of
these
issues
to
people
who
are
interested
in
them.
So
anyone
who
is
taking
a
look
at
this
video
later
can
feel
free
to
hit
the
project
board
our
hit
the
issues
tab
and
go
ahead
and
search
for
issues
that
are
click,
the
issues
that
our
Help
Wanted,
those
were
the
ones
that
we
want
help
on.