►
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
Let
me
show
you,
then
here
I
will
well.
Let's
start
with
this:
is
the
meeting
of
the
sick
cloud
provider,
azure
provider,
subproject
meeting
for
october
19
2020
and
as
a
cncf
and
kubernetes
community
meeting,
we
should
all
abide
by
the
cncf
code
of
conduct
and
basically
that
sums
up
to
be
nice
and
respectful
to
each
other.
B
B
B
A
Yeah,
did
you,
what
do
you
like?
Give
a
video.
A
Let
let
me
do
that,
so
the
this
design
is
mtm
support,
multiple
standalone
balancer.
Within
the
theme
clutter.
We
actually
already
support
a
basic
load
balancer
for
multiple
basic
load
balance
in
the
same
cluster,
but
for
standalone
balancer.
We
only
support
one
because
we
added
all
the
notes
in
the
sim
cluster
to
the
same
standalone
balance
back
in
a
dress
pool,
but
there
are
actually
some
scenarios
that
requires
multiple
standalone
balancers,
such
as
custom,
one
to
control
the
different
outbound
ip
for
different
node
poles.
A
So
in
this
case,
standard
balancer,
much
more
than
the
balancer,
is
aims
to
support
this
scenario.
B
A
I
think
it
is
unique
to
ira,
because
on
other
cloud
providers,
the
each
single
load
balance
service
may
be
mapped
to
a
different
load
balancer,
but
on
error
they
are
shared
yeah.
They
are
sharing
the
same
load,
balancer,
yeah
and
also
for
larger
clusters.
There
are
some
hard
limits,
for
example,
the
the
request
body
and
the
the
advanced
row
number
for
a
network
interface.
Then
they
then
hardly
mean
makes
we
couldn't
support
two
larger
clusters
in
a
single,
not
balancer.
A
So
if
we
split
them
in
a
different
balancer,
then
we
could
support
much
more
notes
within
the
same
cluster,
excellent
yeah.
So
for
the
user
experience
we
are
aiming
it
to
be
exactly
same
with
current
basic
load
balancer.
A
That
means
it
also
supports
the
annotation
to
select
which
load
balance
should
be
used
for
the
service
and
because
the
current
the
current
default
behavior
is
adding
all
the
vms
to
the
single
standard
balancer.
We
want
to
make
sure
the
default
behaviors.
They
are
the
same,
so
a
new
cloud
provider.
Config
option
would
be
added
and
it
default
is
false,
but
when
it
is
the
unique
other,
then
a
different
different
nut
pole
would
would
use
different
standard
balancer.
B
Which
is
also
called
vm
flex
right,
the
new
name.
Yes,
the
vmss
flex.
B
A
Yes,
we
also
asked
a
few
internal
reviews.
I
we
didn't
think
see
any
big
issues
with
our
internal
teams.
B
And
are
there
I
mean
I
presume
this
is
being
asked
for
for
specific
customer
scenarios.
Did
this
get
reviewed
by
any
users.
A
Yeah,
so
for
the
next
step,
a
review
and
guide
it
implement.
A
Yeah
she
would
implement
this
yeah
I'll.
Do
that
great.
A
Yeah
so
for
the
feature,
I
think,
there's
a
discussion
about
the
the
the
intrigue
provider
will
not
accept
new
features
since
1
20.
B
B
B
Okay,
great
any
other
things.
To
add
to
that,
I
think
we
are
good
for
the
first
one.
Okay,
let's
look
at
the
network
api.
A
Yeah,
this
is
the
breaking
changes
of
new
network
api
version.
So
when
we
update
our
network
api
version
to
this
new
version,
the
default
behavior
of
available
zone
will
be
changed
a
lot.
So
we
need
to
determine
how
could
could
we
set
the
the
correct
zone
for
for
load,
balancer
public
ips.
A
I
I
think
the
easiest
way
for
us
in
design
is
we
we
will
get
a
list
of
zones.
A
little
region
then
support
the
zone
and
always
set
the
zone
in
this
region,
but
then
we
would
have
an
easier
than
to
win
with
some
new
regions
for
the
support
zone.
The
this
origin
list
needs
to
be
updated,
and
usually
this
update
will
need
a
version
buff
in
cloud
provider
so
that
that
would
be
easier.
B
A
We
didn't
update
the
network
api
yet
so
we
probably
don't
want
to
design
the
game
120..
So
I
think
this
new
api
version
would
only
be
included
in
auto
3
rlg
provider.
A
Yeah,
but
for
all
our
azure
related
projects
such
as
the
cloud
api,
the
aks
engine
would
also
have
the
same
issue.
So
we
we
need
to
check
with
all
these
projects
and
ensure
the
correct
of
the
the
correct
zone
is
set
when
when
exchanging
or
creativity
are
providing
the
load.
Balancer.
B
Is
that
something
that
you
can
do,
can
you
file
issues
against
those
projects?
There
are
already
issues
there.
I.
A
B
A
B
B
A
Okay
and
the
next
one
is
support
public
I'll,
be
sharing
by
multiple
service,
so
the
pr
has
been
merged
in
math
kubernetes
and
also
in
math
collaborate
error.
We
will
have
also
added
the
entry
test,
so
this
would
be
a
part
of
one
120
feature,
part
of
120.
B
B
A
B
Side
is
great
by
the
way
I've
heard.
Multiple
people
praise
it
that
they
have
found
things
that
they
wouldn't
have
found
otherwise,
and
it
solved
problems
for
them.
So
that's
great
news.
B
B
A
Okay,
craig
do
you
have
any
other
topics
want
to
discuss.
A
Yeah,
so
the
zenfia
is
actually
almost
ready
for
mode.
I
didn't
think
I
didn't
check
it
this
week,
but
in
last
week
I
have
checked
the
pr
and
it's
already
ready
to
module,
at
least
from
our
reviews.
A
So
a
new,
a
new
plugin
is
added
to
kubelet
and
also
any
example
of
asia
is
added
together
in
the
in
the
request.
So
I
I
think
after
that
get
merged.
We
we
could
validate
the
enter
in
eks
engine
and
probably
add
some
engine
tests
to
validate
that
feature
and
in
the
future
we
would
also
need
to
part
the
plugin
to
our
external.
B
B
Okay,
so
those
are
those
are
remaining
work
to
be
done.