►
From YouTube: 20200508 Cluster API Office Hours
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
Hello,
everyone
and
welcome
to
the
august
5th
cluster
api
meeting
called
strapia
is
a
project
of
safe
place
for
life
cycle.
We
have
a
meeting
etiquette.
If
you
would
like
to
speak
up,
please
use
the
raise
hand
feature.
I
post
the
link
to
the
agenda
on
chat,
feel
free
to
add
your
name
to
the
attending
list.
A
A
Going
once
twice
all
right:
let's
keep
going.
I
have
two
psa
for
today
with
release
zero.
Three
eight
zero
three
is
just
mainly
bug
fixes.
We
have
a
few
for
our
cube
adm
control,
plane
that
you
will
see
a
list
here
and
one
for
crs
gloss,
resource
set
and
there's
like
a
bunch
of
documentation
and
other
things
that
we
have
updated.
A
The
other
reminder
that
I
have
is
that
we
started
planning
for
vue
off
before
monday,
and
last
friday
we
did
backlog
grooming
with
some
of
the
materials
and
community
members.
If
you
would
like
to
see
any
features,
definitely
feel
free
to
open
on
rfe
issue,
and
we
can
try
ash
and
put
it
in
the
milestone.
A
The
link
here,
like
links
to
the
v04
milestone
there
is
like
quite
a
bunch
of
things
open,
so
probably
have
to
make
sure
to
kind
of
see
what
we
can
actually
do
for
zero.
Four.
Some
of
these
rfe
features
will
go
through
a
proposal
like
review
and
they
might
need
a
proposal,
but
we're
trying
to
focus
for
alpha
four
on
stability
and
testing
so
yeah.
If
you
have
any
features
that
you
would
like
to
see,
definitely
open
an
issue
and
we'll
try
out
and
put
them
in
the
milestone.
B
All
right,
let's
keep
going
with
discussion
topics.
I've
just
requested
edit
access
to
the
document,
so
I
can
add
my
name.
A
C
A
All
right
so
the
first
discussion
topic-
and
this
is
much
the
call
out.
We
have
the
sig
star
survey
for
kubecon
2020..
I
post
a
link
in
in
the
agenda.
A
If
you
have
any
questions
that
you
would
like
to
propose
for
the
for
the
survey,
we
have
a
few
for
a
cluster
api
but
feel
free
to
either
like
add
a
comment
here
or
add,
more
questions
that
you
would
like,
and
so
this
is
like
a
good
way
to
get
feedback
from
members
of
the
community
that,
like
might
not
join
these
meetings,
or
you
know
like
just
to
dictate
like
actually
the
the
future
of
this
project.
Yeah,
it's
usually
good
feedback.
E
Yeah
just
wondering
how
how
are
we
gonna
get?
Unfortunately,
this
is
after
we've
recorded
our
talks
for
the
maintainer
track.
How
are
we
going
to
get
the
word
out
during
kubecon
eu.
A
I
think
it's
going
to
be
either
a
blog
post
part
of
the
of
the
of
the
blog
post
119,
but
I
don't
know
if
you
look
at
me
is
on
the
call
probably
speak
to
that.
But
look
here,
do
you
know
how
we're
going
to
get
the
word.
A
A
One
thing
I'm
muting
brian
go
ahead.
F
Hi,
so
I
got
the
instructions
for
moderating
sessions,
so
if
you
can
find
who's
moderating
the
session,
they
can
ask
the
technical
audio
visual
engineer
to
put
a
broadcast
message
at
the
beginning
of
the.
A
All
right,
let's
keep
going
fabrizio.
G
Yeah
three
weeks
ago,
I
I
opened
up
a
pr
about
defining
testing
guidance
for
copy,
and
I
already
received
some
some
comments
and
thank
you
all
for
for
for
the
comments.
So
I'm
wondering
if
we,
if
it
makes
sense
to
define
a
deadline
for
those
pr,
so
we
can
start
acting
on
it.
I
saw
there
are
already
pr
pendings
trying
to
implement
what
is
written
there
and
so.
A
Thanks
for
bringing
this
up
so
for
folks
that
they
haven't
had
time
to
read
this
new
testing
guidelines,
so
today
we're
using
genko
and
gamaka
in
a
lot
of
places
and
some
other
places
were
using
straight
go
tests
in
the
past,
we
used
genko
to
spin
up
and
test
and
test
is
just
a
way
to
get
like
a
local
minimum,
kubernetes
cluster,
so
that
you
can
run
your
test
against
it.
A
A
It
might
not
be
useful
for
developers
that
want
to
run
quick
tests
because,
like,
for
example,
like
editors
today
like
offer
a
way
to
run
an
individual
test
with
code
lenses,
for
example,
that
would
work
fine
with
go
test,
but
get
good
doesn't
have
any
extensions
for
the
major
editors
so
we're
trying
to
solve.
For
that
and
the
this.
This
proposal
that
fabricio
put
together
and
kind
of
like
works
in
that
direction.
So
we
will
move
to
go
test
and
use
and
using
m
test
from
controller.
A
A
A
H
Hello,
yes,
I've
got
the
cap,
that's
basically
waiting
on.
H
I
just
wanted
to
ping
everyone
here.
If
there's
any
last
questions
or
comments,
I'm
happy
to
answer
them,
for
you
there's
also
a
pr
that
implements
the
idea
as
well.
It's
linked
somewhere,
I
think
in
that
proposal.
H
But
so,
if
anybody
has
any
questions,
let
me
know.
I
Indeed,
thanks
for
bringing
it
up,
michael
gtm
did
last
week
or
the
week
before
so
really,
just
looking
for
an
approval,
like
michael,
was
mentioning
this
one
is
kind
of
funny.
I
think
not
that
what
you're
proposing
is
funny,
but
the
situation
that
we're
in
where
the
pull
request
that
implements
this
is
basically
uncontroversial
and
we're
spending
a
lot
of
time
debating
user
stories
and
other
aspects
of
the
cup.
I
So
I
did
go
through
and
make
a
bunch
of
resolutions
to
the
outstanding
comments,
because
I
think
they
were
pretty
much
all
addressed.
I
know
cecile
was
the
last
open
one
which
you
responded
to
michael,
so
I
would
hope
we
could
just
go
ahead
and
merge
this,
given
that
the
there's
more
controversy
in
the
cap
than
there
is
in
the
implementation.
A
I
get
my
approval
if
you're
ready
to
emerge
it
michael
just
hold,
cancel
and
yeah.
I
can
merge
this.
A
In
and
you
still
have
your
entries,
if
you
want
to
see
something
else,.
A
So
then,
let's
just
hold
cancel
because
it
has
been
a
while.
So
I
do
it,
we
can
just
merge
this
and
have
it
in
three
nine
all
right,
perfect
anything
else
on
the
the
leaf
deletion
lifecycle,
hooks
proposal.
D
I
Thanks,
sorry,
I'm
eating,
so
I've
got
my
camera
off
so
during
our
backlog
grooming.
Earlier
this
week
and
last
week,
one
of
the
topics
that
came
up
was
when
do
we
stop
adding
features
or
accepting
features
into
the
current
branches
for
v1
alpha
3,
and
when
do
we
open
up
the
main
branches
for
breaking
changes?
I
Essentially
v1,
alpha
4
or
api
changes,
and
one
possible
consideration
that
was
proposed
was
that
we
accept
things
into
cluster
api,
0.3
or
v1
output
3
for
non-breaking
compatible
changes
through
the
end
of
this
quarter,
which
means
the
end
of
september,
and
I
think,
additionally,
we
probably
want
to
set
the
bar
fairly
high
around
what
goes
in
versus
what
we
wait
to
do
in
v1
alpha
4.
I
But
I
wanted
to
open
it
up
to
see
if
anybody
had
any
thoughts
around
this
topic
and
if
there's
any
concerns
with
what
I
just
said
with
the
timeline
through
september
for
v1,
alpha
3.
I
We
just
say:
that's
it
then,
and
I
do
like
I
don't
want
the
main
branch
to
necessarily
be
in
a
middle
or
broken
state
for
a
long
time,
while
we're
developing
v1
alpha
4,
but
at
the
same
time
I
don't
necessarily
want
to
wait
all
that
long
to
start
doing
some
awesome
new
features
so
yeah.
If
you
can
pull
up
the
milestones
for
zero.
Three
nine
and
ten
defense
there's
still
lots
of
things
in
zero.
I
Three
nine
41
open
combined
issues
and
prs,
so
we
may
reevaluate
some
of
that
stuff,
but
yeah.
Unless
there's
no
objections,
let's
just
go
with
end
of
september
for
any
additional
things
going
into
v1
alpha
3-
and
I
guess
does
that
mean
we
want
to
set
october
1st
as
when
we
start
accepting
breaking
changes
on
the
main
branch.
A
Or
is
that
too
late?
I
think
that's
that's
kind
of
late,
so
we'll
keep
the
release
zero
three
branch?
Okay,
so
I
put
august
31st
for
zero,
three
nine.
Hopefully
we
can
get
the
the
external
remediation
and
well
the
one
that
we
just
merged
for
that
we
just
brought
up
for
deletion.
Hooks
should
be
merged
like
within
this
week
or
next
week.
A
So
we
should
be
fine,
then
so
then
those
will
be
the
two
major
features
they're
probably
going
to
are
gonna
go
in
and
then
we
keep
zero
three
ten
just
for
bug
fixes
probably-
and
we
say
but
like
after
reason-
zero.
Three
nine.
I
think
the
main
branch
should
probably
start
accepting
breaking
changes.
So
the
types
and
things
like
that.
A
I
So,
basically,
the
rest
of
this
month.
No
breaking
changes
in
maine
and
starting
in
september
will
open.
G
Okay,
just
just
the
configuration
of
these,
I
checked
the
v039
milestone
and,
in
my
opinion
there
are
the
two
biggest
things
that
we
have
to
still
to
work
on
apart.
The
one
that
that
you
already
mentioned
are
the
search
manager
upgrade,
which
is
already
in
good
state
and
also
the
the
move
for
external
object,
which
is
which
we
have
the
pr
is
nearly
finished.
A
A
A
A
A
B
Further
up
in
the
blue
section
above
it,
I
added
questions
regarding
further
up
again
above
that
there
you
are.
I
added
the
questions
about
the
storage
types
that
are
being
used,
major
storage
concerns
and
whether
or
not
metal
cube
is
being
used
for
bare
metal,
node,
provisioning.
B
B
A
J
A
A
One
thing
that
I
just
remembered
was
kcp
remediation,
nd
or
ben.
Do
you
want
to
give
the
status
of
that
and
just
like
what
we
discuss
and
maybe
also
call
out
that
proposal
that
you
wrote.
A
So
if
you
want
to
like
a
lengthy,
the
purpose
of
your
rook
for
okay,
cp
remediation
and
like
the
status
on
that,
I
think
like
we
we're
delaying
that
to
alpha
four.
Unless
there
is
a
major
concern
to
not
support
control
plane
enough
for
three.
But
if,
if
we
want
to
revisit
that
like,
we
should
probably
do
soon
soonish.
C
A
C
Merge:
okay,
yeah
I'll!
Look
it
up.
A
Yeah,
if
you,
if
you'd
like
to
read
this
proposal
like
that,
that
would
be
great
like
if
you
have
any
concerns
about
kcp
kind
of
heavy,
build
built-in,
remediations
or
a
machine
that
have
failed
or
haven't
passed
the
health
checks.
It
covers
like
a
bunch
of
use
cases
as
well,
so
but
but
yeah
like
this
proposal,
we
can
probably
revisit
if
we
want
to
do
this
within
alpha
3,
because
sandy
pointed
out
like
last
week
it
mhc
today
machine
health
check.
A
A
All
right
anything
else.
A
All
right
well,
thank
you
all
36
minutes
back.