►
From YouTube: Kubernetes SIG Multicluster 2021 Jul 06
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
B
A
All
right,
it
is
three
after
the
well
30..
So
let's
get
started.
This
is
the
july
6
2021
sig,
multi-cluster
weekly
meeting
thanks
everyone
for
coming.
A
I
have
a
feeling
that
it's
going
to
be
a
relatively
light
attendance,
given
that
in
the
us
at
least
it's
the
it's
the
july
4th
week-
and
I
know
some
folks
are
still
on
holiday,
but
we
have
a
pretty
light
agenda
anyway,
so
this
might
be
a
relatively
short
meeting
unless
anybody
has
anything
they
want
to
add
at
the
last
minute.
A
The
first
one,
I
think,
is
really
more
encouraging
people
to
to
take
a
look
at
this
and
share
thoughts,
but
some
of
the
istio
folks
have
created
a
pr
for
client
go
leader
election
to
provide
a
way
to
add
precedence
to
a
leader,
and
one
of
the
one
of
the
motivators
here
is
is
to
basically
make
it
so
that
if
you
have
you
know,
controllers
spread
across
multiple
clusters,
you
can
give
precedence
to
controllers
running
in
a
cluster,
for
actions
running
in
that
cluster,
basically
giving
some
building
blocks
for
for
multi-cluster
failover.
A
For
controllers-
and
I
thought
this
one
was
pretty
interesting
and
it
kind
of
comes
to
a
broader
point
that
I
think
we
should
probably
you
know
in
this
form
start
talking
about,
is
how
do
you
actually,
how?
How
do
we
think.
A
This
is
this
is
kind
of
a
new
thing
I
think
for
for
kubernetes.
You
know
all
of
the
leader
election
constructs.
Everything
are
assuming
that
all
of
your
controllers
are
in
or
acting
on
a
cluster.
But
what
happens
you
know
when,
when
you
have
controller
spread
across
clusters
that
you
know
you're,
maybe
coordinating
some
global
state?
I
think
the
the
example
highlighted
in
this
in
this
pr
linked
in
the
meeting
notes
for
anyone
who
doesn't
see
it
is
basically
you
know,
I've
got
a
cross
region
controller.
A
I've
got
a
replica
in
the
cluster
that
I
care
about,
and
I've
got
a
replica
in
another
region
and
I
want
to
give
precedence
to
the
local
region.
I
think
that
makes
a
lot
of
sense,
but
what
happens
when
you
know
you're,
really
controlling
global
state
or
something
outside
of
the
clusters?
A
How
what
would
precedence
look
like,
then?
How
do
we
want
to
coordinate
that?
So
I
don't
have
any
good
answers
right
now.
I
think
there's
some
ideas
that
have
been
kicked
around,
but
I
wanted
to
throw
that
out
as
as
food
for
thought
and
maybe
something
we
can
dive
into
more
in
the
next
week
or
so.
But
I
think
it's
probably
worth
taking
a
look
at
this
pr
linked
in
the
notes
here.
If
anyone
hasn't
thought,
because
this
is
an
interesting
space.
A
So
yeah
and
I
think,
as
we've
seen
more
folks
start
to
adopt
multicultural
services.
The
need
for
cross-cluster
coordination
is
is
growing
for
sure.
A
So
that's
all
I
wanted
to
say
on
that.
So
again,
this
might
be
a
very,
very
short
meeting.
Is
this
a
situation
where
we
want.
A
Yeah,
I
was
thinking
that
next
week,
we'd
kind
of
maybe
dive
into
some
more
ideas,
so
I'd
also
like
to
solicit
anyone
here,
who's
got
ideas
and
want
to
just
wants
to
discuss
them
next
week.
I
think
it
would
be
pretty
interesting
to.
A
To
see
what
we
can
kind
of
come
up
with
one
one
idea
that
I've
had
and
I've
heard
from
other
people
that
I
think
might
be
interesting,
is
you
know
what?
If
what
if
there
was
a
library
that
let
you
easily
use
like
the
cd
raft
algorithm
across
across
clusters,
maybe
with
crd
as
as
backing
storage
or
something
could
be
interesting.
But
you
know
I'd
like
to
see
what
we
can
kind
of
come
up
with
as
a
group.
A
I
think
the
the
other
topic,
I
think,
we've
we've
had
a
little
bit
of
sporadic
meetings
for
the
last
little
while-
and
I
think
part
of
this
is
we
are.
We
are
one
of
the
few
sigs
actually
doing
weekly
meetings,
so
I
think
you
know
we
meet
maybe
two
regularly
and
it
might
be
easier
to
to
fill
the
agenda
more
consistently
if
we,
if
we
meet
bi-weekly
as
do
most
of
most
other
sigs.
I
think
paul
paul
suggested
this
in
when
canceling
last
week's
meeting.
A
A
A
All
right
well
again,
super
short
meeting,
so
thanks
everyone
I'll
see
you
next
week
and
yeah.
If
please
any
anyone
who
can
who
has
any
ideas
around
how
we
should
be
coordinating
workloads
across
clusters,
I
think
it'd
be
good
to
write
them
down
and
share
them
next
week
and
and
see
if
we
can
kind
of
get
something
started:
cool
awesome!