►
From YouTube: Kubernetes SIG Multicluster 14 June 2022
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).
B
Maybe
you
put
the
proposal
a
little
bit
late.
When
did
you
add
the
discussion.
A
Maybe
we
can
wait
two
more
minutes
and
if
the
house
is
not
present,
we
we
can
talk
about
it
and
at
the
next
time.
B
I
also
message
you,
the
cluster
id
proposal
that
laura
has
been
working
on
for
some
time
now.
So
maybe,
if
there's
some
overlapping
as
well.
D
C
A
E
E
Oh
sorry,
folks,
I'm
on
the
web,
client
and
everything
is
different.
I'm
not
sure
I
can.
E
C
Yeah,
I
think
it's
up
to
you,
I'm
I'm
happy
to
hear
and
talk
about
it
today,
but
I
think
it
you
may
want
paul
and
paul
and
or
jeremy
here,
both
of
whom
can't
make
it
today.
So
I
think
it
depends
what
type
of
feedback
you
want.
A
E
A
Here
I'm
I
request
the
re
review.
The
proposal
from
canada
project
is
about
to
adopt
the
client
side
id
api.
A
Yeah
short
introduction
about
the
background.
Camera
is
the
kubernetes
cluster
measurement
system.
It
enables
you
to
run
your
workload
across
multi
clusters
and
you
can
register
your
cluster
by
the
drawing
command.
A
A
We
put
the
token
to
secret
here
is
just
a
reference
to
to
to
it
and
after
that,
people
can't
get
the
clusters
status
from
like
this,
and
you
can
see
the
member
faster
basic
information.
A
The
problem
is
is
here
in
the
progress
of
our
digital
character.
We
didn't
identify
if
the
cluster
has
been
joined
before
so
people
might
join
the
same
cluster
repeatedly
with
different
arrays
by
mistake.
A
For
example,
people
join
the
cluster
with
ls
number
one
and
join
again
with
yeah,
so
we
we.
We
need
a
way
to
identify
the
clusters
joining
in
the
commander.
King
plan.
A
C
Yeah
you're
intimately
cutting
out.
I
think
we
are
getting
the
gist,
though.
A
Oh,
it's
pretty
good.
B
A
A
So
we
are
going
going
to
propose
a
feature
in
kamada
products,
and
here
I
want
to.
A
If
cluster
has
trust
the
property,
we
just
take
it
as
a
cluster
identifier,
but
if
classroom
no
cluster
property,
we
we
will
take
the
group
system
name,
space
uid
as
unfair
and
at
the
meantime
we
may
create
a
cluster
property
on
the
cluster.
C
Yeah,
so
I
think
the
part
that
the
only
part
that's
different
to
me-
I
guess,
is
the
purpose
or
value
of
the
id
and
the
cluster
api
as
opposed
to
the
id.cades.io
value
in
the
cluster
property
api,
because
I
think
they're
trying
to
be
the
same
thing
and
also
just
part
of
the
reason
the
about
api
is
like
out
of
tree
is
because
it
was
very
difficult
to
agree
to
what
an
id
field
in
well
this.
I
guess
this
is
the.
C
Sorry,
can
you
hear
me
okay,
yeah,
all
right,
so
I
think
the
id
field
you
want
to
put
in
the
cluster
api,
which
I
understand
to
be
a
cluster
kind
in
the
k,
armada,
api
group,
carramada.io,
api
group
and
the
cluster
property
of
the
name
id.k.I.o.
C
I
think
they're,
like
the
same
information,
and
I
think
the
intention
of
the
about
api
and
cluster
property
is
that
that
id.case.io,
the
value
that's
attached
to
the
the
cluster
property
with
that
name
is
like
the
definitive
source
of
truth
for
the
cluster's
name
identification.
C
So
I
think
just
one
point
is
is:
do
we
need
to
put
the
id
in
this
cluster
spec,
as
you
described
under
the
cluster
api
heading?
If
you're
going
to
be
mirroring
it
and
using
it
from
the
cluster
property
kind,
yeah
does.
C
A
I
think
the
the
the
class
the
api
change
is
just
a
draft
proposal.
We.
F
G
A
I
want
to
know,
if
is
if
this
use
case
is.
C
Yeah
yeah,
so
I
think
if
I
like,
separate
from
whether
you're
going
to
put
the
id
in
this
kr
moda
cluster
api,
then
the
proposal
here
related
to
cluster
property
is
correct,
so
definitely
using
it
to
store
the
name
and
defaulting
to
applying
the
cube
system
name.
Space
uid
is
completely
in
line
with
the
cluster
property
kep
and
what
what
we
want
to
go
into
this
field
for
anything,
that's
authoritative,
to
give
the
cluster
a
name.
A
C
So
we
can
talk
about
that
a
different
time,
and
I
understand
that
that's
a
that's
a
separate
proposal
to
take
to
k
armada
to
you
know
see
if.
C
There
or
if
it's
valuable
enough
to
be
in
the
cluster
property
api
for
care,
modest
purposes,.
B
Yeah
yeah,
hong
kong,
I
have
a
question
so
the
cluster
properly
lives
in
the
member
cluster
right.
C
E
B
A
Think
we
will
make
an
flag
or
an
option
for
user
to
decide
if
we,
if,
if
it
allowed
commander
to
create
the
object
on
their
cluster.
C
Yeah
I
see
if,
if
the
kr
moda
cluster
api
is
for
the
purposes
of
a
config
cluster,
maintaining
like
centralized
information
for
all
clusters
like
yeah,
then
yeah.
I
see
that
that's
like
a
totally
separate
use
case
as
cluster
property.
C
D
A
A
So
I
want
to
why
and
what's
the
difference.
C
Yeah
good
catch
that
should
just
be
metadata.name,
so
I
think
we
just
typoed
it
around
at
some
point,
so
we
can
clean
that
up.
A
Okay,
now,
oh,
if
we
want
to
find
the
cluster
added
fire,
I
need
to
consult
to
the
spec
name
or
the
metadata
name.
B
G
Yeah,
I
have
a
question
laura.
C
C
Keep
definitely
keep
us
posted
or
if
you,
if
we
want
to
sync
up
more,
I'm
happy
to
connect
to
so
just
let
me
know
you
can
hit
me
up
in
slack
or
dm
me
if
I
can
be
useful
to
that
conversation
too,.
A
I
just
remember
I
have
another
question
why
this
question
is
also
in
and
we
see
there
is
an.
C
Yeah,
I
think
this
is,
I
mean
fundamentally,
the
layout
was
from
qbuilder.
So
if
we
have
some
beef
with
how
it's
laid
out,
we
can
we
can.
We
can
change
it
and
also
maybe
we
want
to.
A
I
will
try
to
to
see
if
any
programmer
here,
maybe
a
yeah.
If
I
even
maybe
I
can
request
a
change,
maybe.
F
C
Okay,
gotcha
we'll
come
back
to
you
but
hong,
say
yeah.
I
think
if
what
was
I
trying
to
complete
the
thought
of
okay.
D
G
C
If
we
can
def
we're,
definitely
open
to
changing
this,
we're
just
using
the
cube
builder
boilerplate
and
then
if
it
also
means
that
there's
something
valuable,
that
we
can
merge
up
stream
to
cube
builder
so
that
it's
like
more
correct
or
a
better
option.
Then
I'll
have
into
doing
that
as
well.
Because
that's
where
the
organization
came
from.
C
All
right,
thank
you,
so
much
hongsai
again
happy
to
connect
later
with
the
krmana
proposal
too.
If
I
can
help
and
then
thanks
for
pointing
out
the
inconsistency
with
the
about
api
implementation
and
the
cap,
so
we
can
definitely
clean
that
up.
Sanskar.
You
can
go
ahead
with
with
your
questions.
F
Okay,
so
hi
everyone.
I
have
a
few
questions
in
general
regarding
sick
multi-cluster,
like
I
was
exploring
some
of
the
six
where
I
can
fit
myself
in
and
a
little
bit
about
me.
I
am
currently
a
you
know:
beginner,
I'm
not
actually
a
beginner
programmer.
F
It's
been
like
one
year
since
I'm
in
programming
at
google
summer
of
code
2022
participants
with
osu,
and
currently
I'm
trying
to
contribute
in
kubernetes
in
some
field
that
aligns
with
my
core
interest
and
the
thing
that
aligns
something
with
my
interest
is
like
day
one
and
two
cluster
based
operations
like
western
apis
and
sigmund
cluster
is
the
place
where
I
want
to
contribute,
and
I
read
about
cube,
fed
and
how
it
is
fundamentally
different
from
karmada
and
open
cluster
management.
F
So
few
questions
that
I
wanted
to
ask
regarding
the
scope
of
this
sick,
like
first
thing
that
I
came
to
realize,
is
that
that
we
are
definitely
trying
to
make
some
tools
that
might
be
usable
for
multi-cluster
management
for
sure,
but
do
we
have,
but
I
saw
that
there
are
various
projects
that
are
fundamentally
being
built
to.
You
know
add
the
capabilities
and
kubernetes
to
have
to
have
multi-cluster
management
tools
like
karmada
to
help
these
tools
like
calmada
and
open
cluster
management.
F
C
Yeah,
so
the
actual
chairs
of
sigma
take
cluster
are
paul
and
jeremy.
Neither
of
you
are
here,
but
I'm
happy
to
give
my
take
on
it
and
then
will
defer
to
them
as
well
for
maybe
more
color
in
a
future
meeting,
but
in
general
the
like
multi-cluster
space,
while
there
are
many
projects
in
the
ecosystem
remains
kind
of
new,
especially
from
the
perspective
of
what
the
role
is
of
upstream
kubernetes.
To
like
fully
manage
that
process,
and
part
of
that,
I
think,
is
you
know.
C
Other
stuff
is
going
on
and
even
just
single
cluster
world
and
then
I
think
like,
for
example,
working
group
multi-tenancy,
there's
some
like
intersection
there
too,
just
about
like
what
are
the
different
options
for
people
who
want
bigger
deployments
or
different
deployments
than
just
like
isolating
single
clusters
to
one
thing.
C
So
I
think
the
role
of
sig
multicluster
there
a
couple
years
ago
which
predates
me,
and
maybe
some
folks
here-
were
there
but
was
tackling
a
lot
of
projects
like
cluster
registry.
C
The
idea
of
how
do
we
get
different
workloads
onto
different
clusters,
which
is
like
what
work
api
and
cube
fed
previously
were
working
on,
and
some
some
of
those
projects
had
more
ability
to
like
be
something
that
could
be
agreed
upon
upstream,
some
of
them
less
that,
like
the
latter
case,
I
believe
cluster
registry
kind
of
stopped
being
a
project
that,
like
say,
multicluster,
wanted
to
solve
in
a
general
way
and
upstream,
because
the
like
use
cases
were
too
numerous
and
two
and
two
diverse
and
then
most
recently
the
flagship
project
has
been
the
mcs
api
and
I
think
interesting
thing
about
that
and
other
multi-cluster
related
apis
like
the
gateway
api,
which
is
owned
by
sig
network
that
has
like
a
multi-cluster
gateway
component
or
network
policy
which,
like
multi
multi-cluster
network
policies,
now
something
under
discussion.
C
The
role
of
sig
multi-cluster
is
to
try
and
make
sure
that
everything
that's
still
happening
in
single
cluster
works
for
or
doesn't
like
super
conflict
with
what
we
have
going
for
multi-cluster
in
the
upstream
ecosystem
and
then
try
and
figure
out
best
practices
patterns
or
suggestions
for
how
other
projects
can
fit.
In.
C
With
a
standard
that
leaves
enough
room
for
implementation
outside
of
upstream,
but
that
you
know
doesn't
make
like
this
big
smorgasbord
of
stuff
that,
like
never
works
together
so
like,
if
you
read
the
mcs
api,
kef,
there's
a
lot
of
stuff
in
there
and
a
lot
of
history
too,
about
keeping
it
really
implementation
light
and
just
describing
like
the
bare
bones
of
the
api
and
the
standardization
we
all
have
to
agree
to
and
then
implementations
are
have
can
be
happy
to
plug
in.
However,
they
want
so
going
forward.
C
At
least
the
position
of
the
sig
right
now
is
that
there
will
be
lots
of
ecosystem
projects
that
are
separate
from
like
core
like
in
core
and
like
hates,
like
kk
repo
that
implement
the
standards
that
sig
multi-cluster,
proposes,
debates
or
otherwise
organizes
around
and
then
sigma.
The
other
side
of
it
is
that
sig
multi-cluster
needs
to
keep
tabs
on
what's
happening
in
single
cluster
upstream
kubernetes,
to
make
sure
that
those
don't
conflict
wildly
with
this
multi-cluster
feature
we're
trying
to
maintain
and
provide
for,
like
an
ecosystem
of
implementations.
F
No,
it
was
great
a
few
things
that
I
would
like
to
ask
like,
so
it
really
makes
sense
that
sig
multi-cluster
is
actually
trying
to
manage
all
the
stuffs
that
is
going
around
and
it
is
being
as
a
advocate
between
upstream
kubernetes
and
those
projects
that
are
trying
to
you
know
make
strides
in
multi-cluster
management
capabilities.
F
C
Yeah,
I
think,
cube,
fed
itself
has
kind
of
like
the
the
maintenance
and
the
sort
of
how
to
say,
like
the
pressure
to
like
get
that
towards
beta
slowed
down
a
lot
and
then
the
work
api
is
kind
of
a
new
successor
to
to
the
idea
which
is
being
worked
on
by
a
couple
people
in
here
and
mike
feel
free
to
jump
in.
If
you
want
to
talk
a
little
bit
about
that,
so
that's
definitely
sort
of
the
slot.
C
Where
there's
a
lot
of
activity
about
that
specific
problem
of
workload,
distribution
and
then
there
are
definitely
some
other
spaces.
C
I'll,
actually
ask
you
to
reference
in
the
meeting
notes
from
march
8th,
there
was
sort
of
a
conversation
of
like
what
should
we
do
next,
and
so
I
think
that
there's
some
good
ideas
in
there,
but
I
also
think
that
it's
kind
of
just
a
matter
of
when
someone
has
the
bandwidth
to
sort
of
pull
those
you
know
off
while,
like
some
folks
are
working
on
work,
api
and
some
folks
are
working
on
getting
the
mcs
api
to
ga.
C
F
Okay,
I
will
I
will
reference
them
and
one
more
thing
from
from
kubernetes
contributor
perspective,
like
you,
surely
would
be
a
kubernetes
contributor,
so,
like
you
are
being
you
are
involved
with
sick
multi-cluster.
So
do
you
have
to
work
with
some
other
other?
You
know
sex,
like
maybe
cluster
apis.
That
is
the
part
of
cluster
life
cycle,
like
sometimes.
C
Yeah
totally-
and
I
think
sigma
type
cluster
in
particular,
is
in
a
interesting
spot
where
it
operates
in
some
ways
similar
to
a
working
group,
because
it
like
crosses
a
lot
of
different
other
cigs,
and
you
know,
and
as
we
just
mentioned,
also
like
things
outside
of
the
core
kubernetes
ecosystem.
But
that's
definitely
true.
In
my
experience
in
the
sig.
F
C
Cool
cool,
yeah
and
feel
free
to
also
message
me
on
slack
either
in
the
signal
type
cluster
room
or
you
can
dm
me.
If
you
have
some
more
questions,
I'm
happy
to
to
talk
multi-cluster
all
the
time.
C
Okay,
cool:
there
is
nothing
else
officially
on
the
agenda,
but
were
there
any
other
questions
or
comments
from
anybody
else?.
D
C
Okay,
I
see
why
target
quarters
not
to
find
the
service.
Poor
white
ports
are
part
of
service
for
import
expectancy
status.
Several
questions,
multiple
service,
okay.
I
did
not
see
this
select
message
yet,
but
I
can
definitely
respond
there.
D
D
Yeah
that
way,
maybe
if
there
are
enough
discussion,
we
can
yeah
make
this
agenda
for
next.
Okay,
next
time.
D
Great
yeah,
that's
all
cool.
Oh
one,
more
question
about
recording
I
missed
the
last
last
times
community
meeting.
I
I
assume,
there's
a
recording
right,
yeah,
don't
seem
cannot
seem
to
find
it
yet.
C
Okay,
I'll
ask
paul
paul
is
the
recording
wizard
who
gets
them
onto
the
youtube,
so
I'll
poke
him
to
get
it
up
to
date.
Thank
you
very
much.
Kids.
C
All
right
well,
thank
you
for
coming
to
the
june
14th
meeting
of
signal
type
cluster
and
we'll
chat
in
the
slacks
and
the
dms
and
then
also
catch
up
with
the
rest
of
our
action
items
for
next
meeting.
As
you
were
mentioning
ryan,
we
can
talk
about
this
next
point
and
other
than
that
everybody
have
a
great
day
thanks
for
coming
out.
Thank.