►
From YouTube: Kubernetes Federation WG sync 20180709
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
A
We
don't
necessarily
have
owners
against
each
of
them.
I
think
there
isn't
much
that
we
need
to
further
do,
except
for
figuring
out
the
image
build.
That
is
a
religious
image,
build
and
version
II
think
is
to
aya
and
the
third
one
is
the
documentation.
These
three
are
the
pending
items,
which
we
need
to
complete.
A
A
A
B
A
A
D
Yeah,
that's
been
updated,
there's
a
PR
for
that
open,
that's
ready
to
be
reviewed.
I
was
able
to
test
the
I
was
able
to
test
with
a
couple
of
mini
Q
clusters,
but
it
does
require
building
with
the
latest
master
of
mini
cube,
there's
no
actual
released
version
of
mini
cube
that
supports
version
111
and
profiles.
D
So
I
provided
some
instructions
for
that
and
updated
it
to
use
q
builder,
updated
the
development
Docs
to
use
q
builder,
as
well
as
provides
some
instructions
on
building
your
own
container
image
and
your
own
deployment
configuration
using
q
builder
and
then
verify
these.
The
scenario
kind
of
the
example
scenario
on
validating
just
some
basic
validation
of
the
Federation
deployment
by
creating
some
resources
and
being
able
to
move
them
from
one
cluster
to
the
other
kind
of
added
those
example
scripts
or
those
example,
resources
and
verify
those
work.
D
D
D
Yeah
sure
I
think
once
the
user
guide
and
development
guide
is
reviewed,
we
can
get
that
committed.
I'm,
probably
gonna.
Look
at
reviewing
the
unjoin
stuff,
don't
think
it's
required
for
alpha
release.
It
would
simplify
some
things
so,
probably
getting
that
reviewed
will
be
my
next
priority
and
or
getting
a
test
script.
If
we
all
think
it
would
be
valuable
to
have
a
test
script.
That
runs
through
the
example
scenario.
In
the
user
guide
to
validate
the
Federation
v2
deployment
works
to
some
degree.
B
So
I
think
that's
something
good
to
do
as
a
next
step
in
that
area.
I
then,
and
I
would
start
probably
with
a
single
cluster
version
of
it,
because
that
seems
like
going
from
testing
with
one
cluster
to
multiple
clusters
is
a
pretty
big
feat
and
I
think
we'll
definitely
be
able
to
get
a
cluster
from
brow.
B
C
A
Yeah
I
also
think
the
same,
like
crime
would
put
the
CI
set
up
with
multiple
pressures
and
all
that
has
anything
to
do
with
yourself
enemies.
Sure
she
I
have
a
question
for
you.
I
remember.
You
mentioned
that
you
were
able
to
set
up
for
the
latest
mini
tube
with
the
profiles
like
multiple
questions
and
1.11
done.
You
did
from
a
branch,
oh
yeah
I
think
I
did
with
one
of
her
release
from
the
PR.
It
is
not
at
official
release.
A
Oh
oh,
so
I
just
use
that,
but
it
takes
up
quite
a
long
time
to
come
up
and
I
dig
in
dig
into
it
and
it
seems
like
the
systemctl
is
restarting
the
coop
city.
Oh
sorry,
coop
like
every
ten
seconds,
and
it
takes
quite
a
little
long,
but
eventually
it
is
coming
out
I.
Think,
probably
once
they
release
the
new
mini
cube
version,
they
would
definitely
fix
that.
Until
then
we
might
need
to
go
with
this
temporary
mini
cube.
That's
what
I
one
was
talking
about
in
the
beginning,
yeah,
so
yeah.
D
The
user
guide
has
a
link
to
that
version
of
mini
cube.
It's
a
CI
built
version,
not
an
official
release
version
and
also
suggest
as
another
alternative
to
just
build
the
latest
from
master
which
is
pretty
easy
using
their
make
file.
So
one
of
those
two
approaches
should
work.
I
did
run
into
some
other
strange
issues,
though,
that
I'll
probably
follow
us
issue
about
as
a
result
of
testing
some
of
this
stuff.
D
D
A
D
A
D
Definitely
possible
I
know
that
ete
stuff
doesn't
wait
for
the
namespace
to
be
deleted
anymore.
So
if
it
isn't
being
properly
deleted,
it's
possible
that
it's
it's
not
being
detected
because
it
just
attempts
to
delete
it.
Otherwise,
it
slows
down
the
ete
test
quite
a
bit,
but
yeah
I
have
noticed
yeah
I
think
it
might
be
the
same
issue
or
sometimes
it's
deleted,
and
sometimes
it
isn't.
C
Please
file
an
issue.
The
issue
like
the
reason,
I
disable
detection
of
deletion
on
e
to
e.
Is
it's
just
slows
things
down,
and
usually
you
don't
necessarily
want
you
don't
want
to
check
for
deletion.
Success
serially!
You
want
be
able
to
do
it
in
parallel,
so
my
idea
would
be,
is
you
would
collect
all
of
the
namespaces
that
you
and
tentatively
you
delete
them
and
the
tasks,
but
then
you
check
for
deletion
at
the
very
end
and
hopefully
by
then
you
know
everything
would
have
been
deleted
or
at
least
you've
maximized.
C
A
A
A
A
Okay,
so
yeah
on
Thursday,
we
had
this
list
like
we
did,
try
to
brainstorm
and
figure
out.
If
there
is
anything
else
that
we
need
want,
I
mean
what
will
be
born
part
as
part
of
the
release.
This
is
the
miss
tightness
being
shared
right
now.
Is
there
anything
else,
Eva,
not
the
intestinal
issues
and
anything
else
you
might
want
to
add
to
this
or
if
we
just
ran
into
completion.
B
A
C
C
A
C
C
A
B
A
B
B
B
A
A
A
We
can
deliberate
more
on
what
versioning
we
want
to
follow,
but
mice,
mice,
edit,
is
something
like
this.
So
what
you
mentioned
fall
is
like
progression
of
for
the
versions
in
due
course
of
time,
and
that
is
something
that
we
can
follow,
but
at
any
given
point
of
time,
each
of
the
releases
also
might
like.
A
For
example,
we
are
talking
about
Oh,
dot
or
dot
one
right
now
that
one
release
itself
might
be
something
that
we'd
leaves
in
say
two
months
span
of
time
or
one
and
a
half
six
weeks
span
of
time
or
something,
but
we
might
have
intermediate
releases
in
between
every
week.
So
once
you
have
gone
through
all
those
in
the
middle
tulisa's
after
six
weeks,
you
might
have
no
suffix
alpha
or
beta
to
it
and
you
might
have
a
key
in
the
zero
zero
one
version
release.
That
sounds.
B
I
think
it
sounds,
it
sounds
even
more
as
reasonable
is
what
I
suggested,
but
wait
link
different
I'm
happy
to
I'm
happy
to
talk
like
like.
Let's
do
this,
let's,
let's
birth
right
up,
like
our
specific
ideas
on
Terman
versioning
scheme
number,
one
five
one
and
we
can
hash
it
out
there
and
I
think
what
what
is
in
your
poll
is
fine
to
the
verge
now.
I
think
it
says
the
right
thing
at
this
point,
so
we're
fun.
Were
you
going
to
as
a
follow-up
to
this
one?
D
A
Is
one
more
I
think
there
is
a
issue
here?
It's
a
version
for
all
yep,
this
115
version
into
all
binaries,
so
she
mentioned
that
he
might
be
doing
something
to
eat.
So
you
haven't
started
and.
D
A
Haven't
started,
but
it's
pretty
simple
I
believe
yeah.
You
can
eat
from
other
tools
like
you,
barium
yeah
yeah.
So
my
intention
was
a
just:
have
some
common
implementation
of
picking
up
the
version
and
all
the
binaries
can
just
whenever
they
are
build?
If
they
can
pick
that
question
so
I'll
take
it
up.
C
A
A
A
B
Make
sure,
since
we're
basically
acting,
it's
like
one
group
right
now
like
we
want
to
just
make
sure
everybody
knows
what
everyone
else
was
working
on.
I
think
that,
let's
see
we've
talked
about
you
doing
the
tag
builds
and
we're
fun,
you're,
gonna,
add
to
get
version
and
shashi
you're
gonna,
add
the
same
type
of
thing
for
the
controller,
Ivan
I
think
you're
working
on
test
script
for
the
user
guide
that
accurate.
B
D
A
D
D
A
A
Yeah
I
don't
see
how
that
can
be
useful,
see
if
something
like,
for
example,
earlier
in
common
tests,
are
there
used
to
be
a
remote
shell
which
was
provided
to
a
cluster?
The
user
can
play
with
it,
so
that
kind
of
a
thing
for
Federation
is
there?
I
I
feel
it
might
be
useful
to
the
user,
but
a
test
script,
which
just
runs
some
example.
I,
don't
know
if
it
is
reflecting
something
or
providing
some
information
to
the
user
user
might
want
to
just
run.
A
A
D
User
can
get
operated,
yeah
and
I
think
the
idea
at
the
time
that
was
discussed
is
to
have
a
script
that
would
kind
of
automate
that
so
that
we
could
all
have
some
basic
form
of
validation.
That
says
the
user
guide
is
working
and
the
Federation
control
plane
deployment
works
and
potentially
even
add
that
to
a
CI
job.
At
some
point
and
I
guess,
I
didn't
hear
whether
we
agreed
whether
a
script
is
valuable
enough
for
an
alpha
release
or
not.
D
A
D
B
A
C
Just
catching
up
on
all
the
all
that's
been
probably
reviewing
the
docks,
I
mean
I.
Think
the
the
main
thing
from
last
week
was
getting
cute
fed
reading
or
whatever
is
renamed
cute
pad,
and
the
dogs
updated
to
be
consistent
with
the
way
we're
doing
things
I
mean
other
than
but
but
I've
been
kind
of
half
following
this
and
half
reading,
you
know
but
like
what
else,
yeah
yeah.
A
B
A
I
was
saying
is
why
I'm
trying
to
emphasize
on
listing
what
else
needs
to
be
done
as
issues
over
here,
so
that
we
know
don't
need
to
have
daily
sync
on
status?
We
can
just
post
whatever
what
we
got
a
person
tilled
in
that
slack
group
that
we
have
and
distract
these
issues
to
completion.
So
once
these
are
completed,
we
have
time
for
this
right.
Yes,.