►
From YouTube: Kubernetes Federation WG sync 20180801
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
C
C
C
C
Currently,
you
feel
run
the
tap
ensure,
and
then
we
were
generated.
The
lender
I've
currently
Devender.
How
back?
Because
we
need
to
up
may
need
to
make
the
we
need
to
change.
This
felt
make
sure
that
it's
past
the
build
and
it
currently
is
this
bar-
has
been
fixed.
The
increment
is
one
point
eleven,
so
so
I
think
that
maybe
we
need
to
make
some
updates
you
to
the
community
spender
as
well.
So
that's
why
I
want
to
update
the
computers
with
one
point
arrow
and
then
we
can
upgrade.
C
The
current
is
111
to
make
sure
that
we
will
not
miss
a
hack
way
yeah.
This
is
the
first
things
that
I
want
to
discuss.
But
here
are
we
also
have
an
issue
here.
If
we
upgrade
the
chromatids
to
one
on
11
and
then
we
will
also
have
some
other
issues,
so
you
need
to
add
country
increment,
Istanbul,
11.
C
There
are
already
in
current
goals.
There
are
some.
There
are
some.
There
are
some
debris.
There
are
the
kitchen
for
some
of
the
function,
so
we
also
need
to
update
the
khottabych
initially
some
detail,
so
you
mean
that
the
client
gold
there
are
some
deeper
occasion
about
some
code.
So
we
also
need
to
update
the
code
a
bit
such
as
we
may
need
to
update,
leave
Jason
a
theater
when
his
ability
communities
parts,
the
API
server
path.
So
this
is
just
a
summer
mock-up
code
that
we
need
to
update
yeah
I.
D
C
It's
not
probably
about
budget,
because
if
people
make
some
change
of
the
code
template-
and
here
you
need
to
run
the
command
that
that
ensure
to
generated
code
aritaum
people
need
to
update
this
file.
If
we
don't
upgrade
chromatids
to
1.11,
see
yeah
see
see
that
this
is.
D
A
problem
that,
like
we've
dealt
with
before
we
just
will
manually
fix
it
until
we
get
oh
yeah
and
it'll
come
down.
So
it's
like
it's
not
like
we're
broken
and
we
somehow
need
to
go
through
all
this
work.
To
get
unbroken
the
fix
is
simple:
it's
a
one-line
change
and
it's
already
been
done
to
the
tree,
which
is
why
we're
working?
It's
not
a
new
problem.
Yeah
right
so
I
mean
yet
I
mean
I,
admire
your
initiative
and
trying
to
pursue
this.
D
D
D
Like
we
just,
we
use
it
based
on
what's
distributed
by
kubernetes
upstream,
so
like
a
lot
of
this
is
like
it
made
more
sense
to
worry
about
this
back
when
we
were
still
using
our
API
server
builder,
because
we
were
actually
building
an
API
server,
but
right
now
it's
like
the
other
controller,
really
doesn't
have
any
dependencies
on
cue
111
that
are
particularly
important
and
we
have
CR
DS,
which
required
cue
111,
but
only
in
binary
form.
They
don't
really
vendor
anything
that
depended
on
111.
D
C
Right
and
I
also
did
some
tasks
so
from
here
you
meet
you
that
if
I
upgraded
chromatids
to
an
11
and
then
if
I
run,
go
build,
and
then
we
will
see
that
there
are
some
different
kitchen
code
such
as,
if
that
make
landfill,
so
that
removed.
So
we,
so
we
also
need
to
update
the
Federation
YouTube
Federation
controller
YouTube
as
well
to
make
sure
it
can
work.
We're
always
come
to
swamp
on
the
island.
D
So
I
guess
my
point
is
that
this
this
upgrade
is
not
like
a
simple.
Let's
just
change
versions:
yeah
actually
has
to
be
work
on
it.
The
covers
any
there's,
no
urgency
in
it
happening
like
this
is
expected.
Ivan
I
think
he
was
called
a
bug
on
the
fact
that
the
existing
stuff
that
the
resource
plan
Pentagon
was
getting
deprecated,
but
it
only
becomes
a
problem
if
you
want
to
upgrade
right
now
and
like
I
said,
there's
just
there's
no
reason
to
do
that
right
now.
D
B
C
C
And
another
topic
that
I
want
to
discuss
is
currently
out.
I
was
not
working
with
Alex
from
IBM
and
we
are
planning
to
read
a
spec
for
fat
failure
grid
and
Couric
and
I
think
that
we
are
going
to
open
the
document
in
this
week
and
st.
will
be
ready
for
review,
and
we
hope
that
we
can
get
some
comments
from
you
guys
so
that
we
can
move
for
all
the
fauces
product
as
well.
C
B
C
Kurt
I
think
that
we
already
have
some
discussion
in
the
github
issue
here
and
currently.
The
comments
from
Mauro
is
that
we
should
have
a
back
to
to
see
to
see
like
here.
We
have
some
discount
here.
Are
we
need
okay
back
here
to
see
what
is
the
use
case
in
motivation
and
also
of
who
is
going
to
use
it
and
also
what
why
they
won't
use
it
and
also
some
design
as
well?
So
we
are
working
on
this
now.
C
I
think
that
it
will
be
ready
for
a
very
next
week:
yeah,
okay,
okay,
so
the
last
time
I
get
out
because
Carla
I
found
out
there
are
the
pending
some
PR,
so
I
I'm,
not
sure
if
we
can
speed
up
the
process.
The
new
process
I
think
that
note
there
there
are
some
of
small
fixes,
such
as
update
the
user
guide
or
update
the
deployment
script,
and
also
some
other.
Oh
sorry,
okay,
okay,
yeah.
F
B
C
B
Okay,
apart
from
this,
somebody
else
has
it's
another
topic.
If
they
want
to
talk
about
this
last
meeting,
I
think
I
did
mention
about
each
of
us
individually,
having
a
look
at
the
items
that
we
did,
talk
about
and
clarifies
couple
of
meetings
earlier,
I
did
not
see
any
update
against
those
I
mean
that
wasn't
really
expected
also,
but
if
you
already
have
seen
those
and
are
willing
to
give
a
look
at
that
again,
then
we
can
need
the
best
in
them.
In
this
meeting.
B
Yeah
see
if
the
point
was
that
those
topics
are
actually
captured
in
these
meeting
notes
and
share
this
thing:
oh
okay,
and
why
I
am
requesting
this
is
so
that
either
I
mean
I
can
I
can
have
understanding
of
how
to
go
about
creating
the
issues
for
them.
I
don't
want
to
create
issues
for
each
and
every
one
of
them.
That's
the
whole
point.
C
D
B
C
Sure
I
think
that
I
can
append
some
detail
in
the
meeting
later,
as
some
use
cases
from
IBM
yeah.
Currently
I
I
can
do
that.
If
failure
read
and
the
queries
are
one
of
the
important
you
use
cases
for
us,
because
unity
that
you
know
speak
the
Google
cloud
next,
that
the
announcer
that
we
have
the
from
G
key
and
they
also
did
a
demo
for
the
Premji
key
to
will
work
together
with
the
chickee,
and
they
also
create
a
much
closer
on
much
cloud.
B
D
D
As
a
server-side
application,
that's
not
in
kubernetes
like
it's
actually
something
above
that
and
thus
far
we
haven't
really
thought
about
user
experience
beyond
I
wouldn't
say
we
haven't
thought
about
it,
but
I
mean
we've
kind
of
been
focused
on
my
command
line,
and
maybe
this
is
a
juncture
where
we
could
consider
like.
Is
that
always
that
the
only
way
that
we
want
to
present
Federation
functionality?
Do
we
want
to
consider
providing
some
sort
of
user
experience
beyond
CLI.
C
Yeah
currently
I'm
planning
to
do
either
we
want
to
press
on
function
using
the
control
command,
can
get
out.
The
resources
cross,
Costner's,
yeah
right.
C
D
If
I
have
a
you
know
a
web
application
that
is
querying
clusters
in
the
back,
that's
a
totally
different
implementation,
and
so
far
we've
been
really
focused
on
on
the
first
one,
where
we're
just
implementing
via
the
kubernetes
api
and
be
honest.
That
is
restrictive.
I'm
not
saying
that
we
don't
want
to
continue
doing
that,
but
I
think
it
might
be
worth
considering
the
cost
for
pursuing
that
strategy
to
the
exclusion
of
all
of
their
options.
Yeah,
okay,.
D
Yeah
I
guess
I
agree
I'm,
just
having
looked
at
some
of
the
proposals
for
implementing
like
read
access
across
clusters
via
an
API
I'm,
not
sure
that
I
think
that
there's
going
to
be
compromises
than
that
a
lot
and
I
think
we
haven't
really
been
considering
that
what
are
the
trade-offs
versus
alternatives?
You
just
been
saying:
what
is
what
we
have
to
do?
I,
I,
don't
know
there
was
stripping
ourselves
to
just
thinking
about
one
way
of
doing
it
is
gonna,
be
particularly
constructive.
D
G
D
B
Is
it
something
parallel
to
what
what
Vanya
would
be
this
thing
out
like?
He
also
mentioned
that
you
have
something
ready
by
next
week,
yeah.
D
B
C
B
E
E
Suggestion
I
mean
if
I
don't
want
somebody
to
spend
a
lot
of
time
and
effort
writing
up
a
complicated
document.
If
others
perhaps
already
decided,
they're
gonna
shoot
it
down.
So
so
I
think
it's
useful
to
I
mean
if
you
have
alternatives
that
you're
suggesting
I
mean
you
don't
I,
don't
I'm,
not
sure
that
I
agree
as
much
as
Paul
I.
Don't
think
we
need
too
much
detail.
Are
you
talking
about
a
graphical
user
interface,
so
you're
talking
about
a
command-line
tool,
work.
D
D
C
D
E
I
think
it
would
actually
be
useful
to
explore
at
least
the
surface
area
of
you
know.
If
we,
if
we
I
mean
this,
has
been
a
questions
being
sort
of
plaguing
this
project.
Since
the
beginning
is,
if
we
provide
a
kubernetes
api,
then
it
has
a
bunch
of
upsides
which
is
like
other
tools
exist.
The
existing
tools
work
with
it.
E
You
get
free,
gooeys
and
a
whole
bunch
of
other
things,
but
you,
then,
you
know
kind
of
shoehorn
yourself
into
an
API
that
wasn't
designed
to
do
multi
clusters
and
if,
on
the
other
hand,
you
provide
a
multi
cluster
specific
API,
then
you're
immediately
incompatible
with
all
of
everybody's
understanding
of
kubernetes
and
all
the
tools
and
weighing
up
you
know.
The
pros
and
cons
of
those
approaches
is
kind
of
useful
and
we've
tried
both
in
the
past
and
so
I
think
currents
during
those
two
up
front.
E
D
D
E
B
Yes,
so
that's
what
I
actually
wanted
to
talk
about,
which
hasn't
sort
of
concluded?
What
I
want
to
do
is
these
are
the
items
that
we
listed
and
we
gave
some
rough
priorities
to
them,
but
I
could
figure
out
that
majority
of
the
items
did
not
get
both
the
stigma.
Major
stakeholders
names.
So
if
you
see
here,
many
of
the
items
have
guava
only
listed
at
least
here.
B
So
what
I
wanted
to
do
was
get
some
indication
from
reddit
also
as
their
their
most
priority
items
and
the
real
priority
on
them,
and
let
me
go
ahead,
creating
issues
with
respect
to
the
number
of
people
that
we
have
currently
working,
say
five
or
six
and
then
move
ahead.
But
for
us
we
were
already
working
on
these
like
these
for
pending.
So
this
is
what
I
am
working
on,
which
is
job
shadowing
xph.
B
It
will
be,
and
what
I
currently
was
doing,
I
figured
out
that
good
amount,
of
course,
would
be
reused
in
the
other
shed
user.
So
I
was
refactoring
a
little
bit
which
is
complete,
so
I'll
be
raising
the
PR
for
that,
probably
today
or
tomorrow
sure
she
was
in
the
process
of
completing
the
DNS
submitted
which,
as
per
our
earlier
discussions,
we
were
because
moving
into
of
of
be
worth
changing
to
reuse
on,
you
can
say,
use
the
multi
cluster
not
multiply
so
the
external
DNS
functionality
yeah.
F
Yeah,
let
me
give
that
update
so
yeah.
Currently,
I'm
almost
done
utilizing
the
external
DNS
to
make
it
work
for
our
scenario
like
service
discovery
and
I'm
working
all
the
way
like
in
external
units,
make
feature
happen
like
some
implementation
details
there.
So
hopefully
you
should
be
like
from
the
predation.
V2
wrapper
perspective.
I
think
the
features
are
complete.
I
think
I
raise
the
PR,
so
I'll
review
comments.
If
any,
then,
let's
see
and
then
external
DNS,
once
that
lands
in
I
think
we
should
be
over.
F
E
B
B
Does
it
make
sense
for
us
to
make
issues
for
all
of
them
and
list
them
against
our
echo
and
maybe
I,
don't
know
when
they
are
going
to
get
picked
or
that
kind
of
stuff,
or
we
can
realistically
have
say
a
short
list
of
items
which
we
say
will
be
anyways
focusing
for
next
and
for
our
next
in
Excel,
for
something
like
that.
But
I
have
to
get
clear
signals.
B
G
I
think
it
would
be
good
if
we
had
issues
to
represent
all
these,
because
that
is
a
good.
That's,
probably
the
most
visible
way
that
people
can
see
that
there's
some
interest
in
doing
them,
so
I
would
definitely
be
supportive
of
like
having
issues
to
represent
these
in
github,
so
that,
if,
if
folks,
you
know,
are
looking
at
the
project
and
going
through
open
issues
that
they
can
see,
there's
some
degree
of
interest,
maybe
even
if
people
aren't
actively
working
on
them
and
identify
hey
I'd
like
to
work
on
that.
B
D
D
So
there
was
a
prelude
to
committing
major
effort
to
anyone.
Saying
I
mean
smaller
stuff,
fine,
but
it
was
like.
Oh
we're.
Gonna
go,
spend
like
weeks
two
months
implementing
something
who'd
be
really
nice
to
know.
It
was
actually
a
priority
for
our
customers
or
members
of
the
community
of
Ida
Huawei.
Now,
it's
not
suggesting
that
you
know
you
don't
have
good
reasons
for
implementing
what
you
want
to
implement,
but
in
the
absence
like
evidence
that
it's
more
than
just
you
it's
hard
to
say
yes,
this
is
absolutely
something
we
need
to
do
that.
G
Yeah
I
I'll,
just
echo
Bru's
comments
that,
like
I,
think
it's
really
important
to
get
validation
from
users,
since
this
is
a
fairly
significant.
You
know
departure
from
v1
in
terms
of
the
API
constructions.
I
am
really
interested
in
in
talking
to
folks
in
the
community
that
may
have
tried
Federation,
v2,
I'm,
I'm
kind
of
curious
to
know.
E
Yeah,
that
was
one
of
the
reasons
why
I
did
the
presentation
at
the
community
meeting.
There
was
quite
a
bit
of
interest,
but
nobody
is
actually
step
four
on
offered
that
information
there,
but
I
think
we'll
just
have
to
you
know
we
need
to
post
blog
posts,
maybe
make
a
video
demo
these
kind
of
things
and
and
distribute
them
more
widely.
G
Yeah
I'm
thinking
about
stuff
like
how
can
we
raise
raise
visibility
into
the
fact
that
there's
something
new
out
there,
like
on
social
media,
for
example,
to
try
to
get
folks?
That
might
say,
hey
I,
didn't
realize
that
that
thing
existed,
but
it
might
solve
a
problem
for
me.
I
want
to
try
it
out
and
use
it.