►
From YouTube: Kubernetes SIG Cluster Lifecycle 20201028 - cluster api
Description
agenda doc https://docs.google.com/document/d/1fQNlqsDkvEggWFi51GVxOglL2P1Bvo2JhZlMhm2d-Co/edit
A
Hello
today
is
the
28th
of
october,
and
this
is
a
cluster
api
office
hour
meeting
and
before
starting.
Let
me
remind
you
something
about
our
meeting
etiquette.
So,
first
of
all,
this
is
a
meeting
in
this
meeting.
We
are
respecting
the
the
cncf
code
of
conduct,
so
please
be
kind
with
each
other
and
second,
we,
we
kind
of
suggest
you
to
use
the
raise
end
feature
before
speaking.
I
will
do
my
best
to
follow
and
call
speakers
as
soon
as
they
called.
A
But
unfortunately,
with
this
clarion
I
cannot
see
the
the
list
of
participants,
so
maybe
andy
you
can
help
me
in
this.
Thank
you
and
and
next
so
there
is
a
meeting
document.
I
I'm
sharing
the
document
in
the
screen.
I've
put
the
link
of
the
document
in
in
the
zoom
chat
and,
if
you
don't
have
access
to
this
document,
you
have
to
subscribe
to
the
c
cluster
recycle
mailing
list,
and
I
kindly
add
you
to
add
your
name
on
the
attended
list.
A
On
the
meeting,
I
see
that
someone
is
already
doing
this
great.
Thank
you
and
also,
I
kind
of
suggest
you
to
add
your
topic
down
in
in
the
agenda
and
before
starting
with
the
agenda
today,
I
would
like
to
add
to
ask
if
there
is
some
new
member
or
some
new
participant
in
today's
meeting
that
want
to
introduce
themselves,
we
will.
We
would
like
to
to
know
you
and
you.
Why
are
you
hear
about
what
are
your
interests
in
this
project?
So
I
give
a
little
bit
of
time
to
see
someone
to.
A
A
A
A
We
will
because
next
week
there
is
also
u.s
election,
so
we
will
decide
it
will
be
probably
at
the
end
of
the
next
week
or
the
beginning
of
the
week
after,
but
we
will
follow
up
on
this
also.
I
want
to
give
a
quick
summary
of
what
is
in
in
the
release.
So
the
main
themes
are
external
remediation
and
also
better
visibility,
and
we
have-
and
we
will
have
a
machine,
not
those
conditions.
A
So,
on
the
machine
we
will
have
a
condition
which
surface
the
status
of
the
node
and
also
we
will
have
a
kcp
condition
that
are
going
to
surface
the
status
of
the
api
server
controller
manager
and
the
tct
static
port
in
into
again
into
machines.
A
Okay,
great
next
topic
is
it's
just
a
reminder.
We
are
starting
to
work
on
kype
for
the
next
v1
alpha
4
release.
I
put
the
the
link
of
two
pr
which
are
which
I
called
that
are
basically
those
pr
are
moving.
The
google
docs
that
we
discussed
about
the
management
caster
operator
and
about
kcp
scaling
into
a
factor
proposal.
So
I
kind
of
ask
to
everyone
to
to
kick
in
and
review
and
provide
comments.
B
Yeah,
hey,
I
sent
out
an
email
on
the
same
subject
to
the
cluster
life
cycle,
but
I
was
looking
to
contribute
the
cue
mark
provider
that
I
created
to
the
kubernetes
to
make
to
the
kubernetes
six
org
and
have
it
be
under
the
cluster
lifecycle
umbrella.
B
Yeah,
I
think,
in
the
past
I
was
looking
at
what
the
most
recent
one
was:
the
nested
cluster
api
provided
nested
and
they
just
like
linked
it
in
the
they
linked
to
the
notes
and
the
video
of
the
sid
call
the
cluster
api
sig
call
as
like
the
evidence.
You
know
like
approval,
so
I
guess
I
wanted
to
announce
it
and
if
anyone
had
any
objections
you
can
hear
them
out,
but
otherwise
I
think
it's
enough
to
just
say
we're
going
to
do
it.
C
D
Handy,
hey
everybody.
I
just
wanted
to
mention
that
we
had
some
folks
who
were
testing
some
v1
alpha
2
to
v1
alpha
3
upgrades
for
kappa,
and
we
ran
into
some
problems
with
some
of
the
api
conversion
logic
in
the
aws
provider
and
the
first
link
that
I
have
there
is
the
pull
request.
That's
going
to
ideally
fix
it.
The
second
one
which
I
want
to
call
attention
to
is
a
pr
that
sadaf
has
open
to
expand
our
conversion
testing.
D
So
hub
just
spoke
and
back
and
spoke
to
hub
and
back,
and
the
smoke
spoke
to
hub
and
back
is
the
new
copy
of
the
test
or
the
new
test.
That's
added,
and
it
makes
sure
that
we
take
the
the
hub
version,
get
the
data
stored
as
an
annotation
like
we're
doing.
We
go
modify
the
spoke
version
and
then
we
do
a
full
round
trip
to
make
sure
that
we
aren't
overwriting
anything
which
is
the
the
bug
that
we
encountered.
D
So
I
would
encourage
you
to
take
a
look
at
the
cluster
api
pull
request.
You
potentially
can
update
your
go
mod
to
pull
it
in
locally
and
just
see
if
you
have
any
conversion
issues,
hopefully
not,
but
I
did
just
want
to
mention
that
this
is
a
particularly
difficult
area
to
get
right.
The
code
can
be
really
confusing
and
trying
to
rationalize
and
understand.