►
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
All
right,
hi,
everyone
today
is
wednesday
january
27th.
This
is
the
cluster
api
office
hours.
Cluster
api
is
the
subproject
of
sig
cluster
lifecycle.
Please
make
sure
to
follow
the
cncf
code
of
conduct
and
the
meeting
etiquette
during
this
meeting.
A
B
Here
I
am
hi
everyone,
so
I
have
two
psa
from
the
c
crystallized
cycle
meeting.
So
one
one
psa
from
the
secretary
cycle
meeting,
basically
forks
from
ks
infra
working
group
requested
that
everyone
should
migrate,
end-to-end
job
which,
which
are
reliant
relying
on
google.
B
B
A
C
Yeah,
I
was
just
going
to
say
I
was
looking
at
that
earlier
and
I
think
the
the
rendering
issue
that
you've
highlighted
is
actually
already
merged
into
master.
I
don't
think
that's
caused
by
the
cbe.
I
had
a
look
at
the
the
master
netlify
and
it's
already
there.
Someone
has
added-
or
at
some
point
we've
added
enough
providers
to
the
tabs
so
that
it's
overflowed
on
the
line.
So
I'm
not
sure
what
we
can
do
there.
B
I
think
thank
you
for
the
for
investigating
the
problem
joel.
So
if
this
is
the
case,
we
have
to
fix
so
I
agree
we
have.
We
can
merge
the
pr
as
it
is
and
to
fix
the
rendering
problem
we
have
to
to
change
a
css
file
and-
and
if
we
look
at
some
old
pr,
I
will
search
for
for
the
change
that
we
have
to.
E
Sorry,
yes,
I
just
saw
this
on
twitter
like
half
an
hour
ago,
so
there's
a
talk
from
equinix
on
tinkerbell
and
cluster
api
at
12pm
pst
today,
with
cncs.
A
A
All
right,
if
not,
let's
move
on
discussion
topics.
We
have
laura
from
sick
cluster
sig
multi-cluster,
hello.
I
did
make
some
slides
to
try
and
facilitate
a
little
bit.
Let
me
see
if
I
can.
I
don't
think
I
can
share
my
screen.
Yes,.
A
C
A
Can
everybody
see
okay,
great
all
right,
so
my
name
is
laura
and
I'm
visiting
from
sig
multi-cluster
and
I'll
kind
of
get
right
into
it
here
to
sort
of
explain
why
I'm
here,
and
also
what
signal
to
cluster
is
so
sig
multicluster
is
focused
on
trying
to
figure
out
how
to
solve
issues
related
to
connecting
and
networking
multiple
kubernetes
clusters.
A
Together
we
have
a
mcs
api
that
is
in
cap
form
that
has
been
implemented
a
couple
different
places
and
part
of
our
work
is
also
anything
that's
sort
of
blocking
that
kep,
one
of
which
is
the
concept
of
a
cluster
id
where
a
cluster
knows
itself
what
its
name
is
for
our
purposes.
We
also
kind
of
needed
to
know
what
membership
like
what
membership
of
multi-clusters
of
a
multi-cluster
group
that
it's
actually
in
so
that's
kind
of
the
the
story
from
us.
A
We've
been
working
on
this
for
directly
as
a
cap
for
the
past,
like
two
months
or
so
right
now
we
have
it
as
status.
Provisional
insider
sig,
multi
kep
list
and
we're
mostly
focused
on
defining
the
properties
of
a
cluster
local
crd,
is
the
implementation
we
want
that
stores
the
cluster
name
and
also
again,
information
about
you
know
what
am
I
a
member
of
me
this
cluster?
A
What
other
clusters
am
I
friends
with
and
as
we're
trying
to
move
this
into
implementable
and
also
just
keep
moving
it
down
the
kep
pipeline,
we're
considering
when
should
we
go
alpha?
Should
we
try
and
do
that
in
1.21
and
also
this
camp?
This
started
the
conversation
of.
Should
we
be
targeting
this
in
a
sigs
repo
or
in
tree,
and
for
the
latter
point
we
think
that's
only
necessarily
relevant.
A
If
you
know
this
is
useful
to
other
cigs
and
other
and
other
people
and
I'll
get
into
it
a
little
bit
of
why.
I
think
this
may
be
relevant
to
cluster
api
specifically,
but
that's
kind
of
why
we're
going
around
and
seeing
what
other
groups
are
are
about
just
to
highlight
this
a
little
bit
more
of
what
the
kep
itself
is
about.
From
the
perspective
of
our
sig.
A
We
basically
wanted
to
find
a
standard
for
how
clusters
should
refer
to
each
other
when
they're
in
the
same
multi-cluster
environment,
and
we
want
it
to
be
useful
and
strict
enough
to
unblock
some
known
use
cases.
This
has
previously
been
kind
of
like
trapped
in
a
abstraction
before,
but
we
have
some
specific
use
cases
in
multi-cluster
that
we
know
we
can
try
and
address,
and
specifically,
we
know
that
we
want
people
to
be
able
to
identify
which
cluster
is
which,
in
a
multi-cluster
setup
in
their
logs
for
debugging
purposes.
A
So
a
name
would
be
helpful
there,
just
a
disambiguate,
pods
multi-cluster
headless
service
to
make
dns
names
to
track
when
new
clusters
have
joined
the
environment.
So
that's
what
the
q
two
clusters
down
here
are
talking
about
a
couple.
People
from
cluster
api
have
actually
said,
hey,
and
I
think
y'all
are
in
this
meeting.
So
hi
to
everybody,
who's
commented
on
the
pr
or
got
connected
somehow
on
github,
so
vince
commented
a
little
bit.
A
Cecile
had
a
issue
that
got
linked
at
one
point,
so
it
seems
that
there's
this
independent
interest
in
cluster
api
to
have
something
similar.
I
think,
if
I
understood
the
issue
properly,
there
is
some
talk
about
having
it
as
an
annotation,
we're
interested
in
having
sort
of
like
this
separate
crd
as
an
implementation.
A
So
I
think
we'd
like
to
figure
out
how
we
can
collaborate
or,
if
they're,
separate
issues
or
or
what
so
fundamentally,
I
think
what
we
want
to
know
from
you
all
and
in
general,
from
other
cities
and
sub-projects
is:
do
you
have
a
need
for
your
clusters
to
know
their
own
name
via
close
cluster
local
crd,
that
has
some
of
these
kept
standard,
specific
characteristics
related
to
their
uniqueness,
mutability
and
lifetime?
A
What
are
the
use
cases
that
you
have
do?
They
differ
from
the
ones
that
we've
already
included
in
the
kep
in
the
first
place,
and
I
put
a
little
link,
which
is
basically
the
same
as
what
I
showed
on
the
prior
slide
with
the
two
clusters
talking
to
each
other
and
if
you're
interested
in
collaborating
or
getting
updates
like
this
style.
What
is
the
best
way
for
me
to
work
with
you
and
as
a
sort
of
little
aside?
A
Possibly
the
hardest
thing
of
any
open
source
project
is
naming
things
and
we're
still
naming
our
crd.
So
if
you
are
a
person
who
likes
to
brainstorm
names,
I
also
have
a
google
sheet.
That's
collecting
them,
and
if
you
like,
to
make
your
voice
known
in
the
democracy,
I
can.
E
A
Send
you
the
survey
when
it
comes
down
to
it,
but
that's
that's
for
later.
So
that's
that's
my
my
spiel.
A
F
Hi
and
welcome
lord
to
this
meeting.
Thank
you
so
much
for
presenting
this.
I
know
that
we've
been
kind
of
like
silent
on
that
pr
for
a
while,
so
really
appreciate.
You
come
join
this
meeting,
so
I
think
this
has
been
asked
like
quite
like
a
bit
to
like
understand
the
name
of
a
cluster
from
the
cluster
itself.
F
There
have
been
like
a
different
things
that
have
been
asked,
but,
like
our
current,
I
think,
like
idea,
was
to
just
annotate
the
the
node
where
ico
like
with
the
with
an
annotation
specific
to
cluster
api,
but
I
mean,
I
think,
like
I
don't
want
to
speak
for
everybody
here,
but
like
we're,
probably
willing
to
kind
of
adopt
like
something
that's
like
in
kubernetes
itself,
actually
like.
F
We
actually
welcome
that,
like
a
lot
of
like
we
don't
go
our
own
way
so,
like
huge,
plus
one
from
from
my
side
on
that,
and
definitely
like
interested
in
collaborating
on
this
a
little
bit
more.
We
can
bring
use
cases
to
you
folks
and
bring
something
back,
and
we
could
also
like
tie
some
things
up
together,
a
little
more
as
far
as
naming
like.
I
would
love
to
to
see
the
spreadsheet
personally
I
like
to
black
chat
on
that
a
bit,
and
hopefully
we
can
help
each
other.
A
Anyone
else
one
thing
that'd
be
great
is
if
you
can
share
those
slides,
so
we
have
the
links
and
everything
in
the
meeting
notes.
Awesome,
yeah
they're
in
the
meeting
notes.
Next
to
my
name,
perfect
too,
so
you
can
grab
it
and
yeah.
I
think
if
it's
best
to
collaborate
on
this
online,
should
I
just
find
you
all
on
slack
yeah,
we're
all
in
flag,
hashtag
cluster
api.
I
assume
there's
a
multi-cluster
channel
as
well.
A
There
is
feel
free
to
to
hang
out
with
us,
too
yeah
cool.
I
should
probably
join
that
yeah.
So
from
my
sides,
the
annotation
use
case
was
more
like
well,
there's
two
sides
of
it
like.
If
you're
it
was
first
of
all,
like
just
a
nice
convenience
like
if
you're,
a
user
who's
creating
a
bunch
of
cluster
api
clusters-
and
you
just
want
to
be
able
to
quickly
tell
like,
oh,
which
one
am
I
on
right,.
E
A
Just
to
be
able
to
like
see
that
information
and
then
also
like,
if
you're
a
platform
or
something
building
on
top
of
cluster
api,
and
you
have
different
cluster
api
clusters
from
different
providers
right
now,
there's
no,
nothing
that
identifies
the
kubernetes
workload
cluster
as
being
a
cluster
api
cluster
like
it
knows
nothing
about
its
management
cluster,
because
it's
all
the
other
way
around.
A
A
Yeah
sounds
like
there's
a
relationship
between
the
workload
and
the
management
cluster.
In
the
same
way,
we
envisioned
just
like
general
multi-cluster,
so
I
think
there
is
actually
even
overlap
on
that
point,
which
I
hadn't
considered
in
the
first
place.
So.
F
One
thing
to
notice
like
if
this
comes
in
121
as
an
alpha,
we
can
probably
like
line
up
of
first
experimental
implementation
for
our
v1
alpha
4..
So
I
think
I
think
it
should
line
up.
I
mean
we're
kind
of
a
few
months
away,
probably
from
like
a
0
release,
but
but
yeah
we
can
try.
We
can
definitely
try
something
out
like
it's
mostly
a
matter
of
like
opening
an
issue
and
like
see
if
someone
wants
to
work
on
it
and
yeah
create
these
resources.
A
In
terms
of
the
kep,
what's
the
timeline
on
it,
because
I
know
it's
been
open
for
a
while,
like?
Are
you
nearing
consensus
or
yeah?
I
think
internally,
it's
like
multi-cluster.
We
have
a
lot
of
consensus
and
we're
trying
to
fill
in
all
the
details
for
implementable,
which
is
also
why
we
thought
it.
A
We
had
a
chance
to
hit
alpha
for
one
two,
one,
because
we've
already
kind
of
talked
it
out
a
lot
in
our
in
our
meetings
and
just
need
to
like
get
it
down
on
the
paper,
but
definitely
wanted
to
see.
If
there
was
you
know,
broader
utility
and
then
also
any
like
specific
pushback
from
other
group
affiliated
groups
before
we,
you
know
really
make
that
timeline
commitment.
A
I
guess,
but
that's
that's
my
perspective
and
we're
super
actively
working
on
it
and
if
things
sound
good
from
from
this
group-
and
we
can
keep
talking
about
it
offline
too,
but
I
think
we
would
push
like
full
steam
ahead
on
that.
A
Okay
yeah.
So
I
think
it
yeah
if
anyone's
interested,
go
check
out
the
kep
and
definitely
review
it,
because
there
is
definitely
some
overlap
or
not
overlap
but
similar
interests.
Here.
A
Okay,
I
don't
see
any
hands
raised.
Anyone
else
have
questions
or.
A
A
Thank
you
laura.
Let's
keep
going
to
jen
debug
help
needed.
G
Yes,
hi
everyone.
I
just
wanted
to
ask
help
here.
I've
been
opening
pr
like
week
ago
for
rollout
strategy
for
for
kcp
and
I'm
hitting
the
issue.
Now
it's
a
conversion
related.
I
think
I
hit
this
bootstrap
token
form
error
in
in
ci,
and
I've
been
trying
to
debug
these
like
days
now,
and
I
I
don't
have
any
clue
how
to
how
to
fix
the
conversion.
G
So
if
anyone
have
any
ideas,
I
don't
want,
we
don't
necessarily
need
to
go
through
it
right
now,
but
slap
me
or
or
commenting
into
the
pr
any
help
would
be
great.
A
Sounds
good
yeah,
let's
take
this
offline,
I'm
sure
a
few
of
us
can
take
a
look
and
pitch
in.
B
Yes,
yesterday,
in
secretary
cycle,
we
discussed
it
again
and
which
talk
to
we
want
to
to
bring
to
the
next
kubercon,
which
is
vehicle
again,
and
basically,
the
current
stance
of
of
recycle
is
that
we
are
bringing
one
torque
as
a
sig,
so
the
the
usual
sig
introduction
talk
and-
and
we
are
seeking
for
volunteer
for
for
this
talk.
B
Second,
we
we
leave
each
project
the
possibility
to
send
their
own
submission
so
that
technically
we
can
send
a
submission
for
copy
and
and
also
submission
for
each
copy
provider
implementation.
So
I
would
like
to
to
understand
to
hear
if
someone
is
interested,
please
get
in
contact
with
lubomir,
justin
or
team
to
basically,
and
they
don't
want
to
block
or
nothing
that
they
want
to
only
just
to
keep
a
picture
of
what.
How
many
suggestions
are
we
sending,
and
last
thing
is
that
the
deadline
is
the
ninth
of
february.
B
Yeah
exactly
so,
basically,
we
are
not
aware
of
limitation
from
the
cncf,
so
the
idea
is
that,
okay,
don't
don't
scope
down?
Don't
don't
give
limit
to
this
project
if
everyone
want
to
submit,
they
are
free
to
submit
and
then
we
we
will
look.
We
will
basically
listen
to
what
the
cncr
falls
will
with
answer.
D
Okay,
we
have
a
note
in
chat
that
the
deadline
is
the
seventh,
not
the
ninth.
D
F
Yeah,
I
just
wanted
to
say,
like
this
is
a
great
opportunity,
usually
for
new
folks,
for
the
project
to
contribute
and
if
you're
interested
in
might
be
doing
like,
for
example,
like
a
cluster
api,
deep
dive,
or
things
like
that.
F
Please
do
have
a
submission
in
place
and
we
can
definitely
help
you
in
terms
of
like
giving
you
an
idea
of
what
we
submitted
like
in
the
previous
years.
But
I
would
really
suggest
for
new
folks
to
take
this
opportunity
because
that's
that's
a
great
great
way
to
step
up
the
community.
A
A
Also,
one
thing
I
would
add
is
like
this
doesn't
necessarily
have
to
be
like
a
overall
all
of
cluster
api
deep
dive,
because
I
think
that's
what
we've
mostly
done
in
the
past,
and
it
made
sense
because,
like
people
still
didn't,
really
know
what
cluster
api
was.
But
I
think
we're
getting
to
a
point
where
we're
mature
getting
mature
enough,
where
it
would
be
interesting
to
see
some
deep
dives
on
like
more
specific
topics
or
more
specific
features
and
go
maybe
a
bit
deeper
than
just
like
here's.
A
H
H
Well,
thanks
cecilia,
I
just
want
to
say
I
think
your
idea
about
like
branching
out
going
a
little
wider
on
the
topics
is
an
excellent
thought.
My
question
would
be
like.
Is
that
going
to
be
acceptable
to
the
you
know
the
people
kind
of
organ?
I
guess
the
cncf
kind
of
organizing
kubecon
or
whatever,
because
I
know
sometimes
they
get
a
little.
H
A
Yeah,
that's
a
good
question,
I'm
not
sure
honestly,
but
I
think
definitely
you
want
to
like
give
a
bit
of
an
introduction
for
to
be
like
friendly
to
everyone
in
the
room
for
people
who
haven't
really
heard
the
basics.
But
I
think
it's
okay
to
like
go
a
bit
deeper
like
later
on
in
the
talk
on
like
some
of
the
topics,
especially
if
you've
had
like
previous
talks
that
already
went
over
all
the
basics,
but
that's
just
yeah
my
opinion.
H
Go
ahead.
Okay,
I
was
just
going
to
follow
up
like
I
guess
the
question
would
be
like
you
know
in
terms
of
kubecon
like
would
we
ever
have
the
opportunity
to
get
like
two
talks
in
where
we
could
do
like
a
101
and
then
a
201,
or
something
like
that.
You
know
anyways,
that
was
it
go
ahead
for
gracie.
B
So
my
understanding
is
that
those
talkers
are
talk
of
so
they
are
torque
of
the
maintainer
strikes.
So
the
goal
of
the
maintainer
tracks
usually
is
to
basically
tell
people
there.
Is
this
project
and
look
and
tell
to
the
people
what
this
project
is
doing
and
and
then
look
for
new
contributors,
but
I
I
totally
agree
with
cecile
that,
given
that
it
is
the
third
or
the
third
fourth
time
we
are
doing
this
in
kubercon
or
even
more,
it
is
not
required
to
do
a
full
introduction,
high-level
introduction
project.
A
Cool
but
yeah
if
you're
curious
reach
out,
I
don't
see
any
other
topics
for
today.
Are
there
any
last
minute
additions
or
demos
or
anything,
interesting
questions
that
people
want
to
throw
out.
A
There
all
right,
if
not
thanks,
all
we're
going
to
cut
it
short
and
have
a
great
rest
of
your
day,
see
you
all
next
time.