►
From YouTube: SIG Kube-Proxy Bi-Weekly Meeting for 20220907 (APAC)
Description
SIG Kube-Proxy Bi-Weekly Meeting for 20220907 (APAC)
A
So
we
are
at
the
copying
apac
meeting
on
september,
7
2022,
okay.
So
what
I'm
trying
to
do
is
basically
watch
sunscreen
through
cupping
and
see
get
him
started
with
contributing
to
popping.
A
A
All
right,
so,
basically,
what
we're
trying
to
do,
as
you
must
have
gathered,
is
try
to
add
back
ends
to
the
cupping
brain
right,
wherein
we
are
trying
to
decouple
the
architecture
of
your
proxy.
So
these
are
a
bunch
of
back
ends
that
we
have
in
case
now.
User
space
is
moving
away
from
kk
code
base.
There's
a
pir
in
flight
to
remove
the
code
for
user
space.
One
of
the
issues
that
you
could
contribute
to
is
see
how
we
can
add
user
space
to
the
cpi.
A
So
we
do
have
most
of
the
back-ends
wired
in
to
the
ci
when
we're
running
kubernetes
and
performance
tests,
but
user
space
is
not
one
of
them.
As
of
now
this
user
space
is
the
support
of
user.
Space
is
heavily
dependent
on
the
one
in
queue
proxy
and
we
you've
tried
to
make
it
work
with
cup
thing.
A
So
there
are
a
bunch
of
things
that
are
remaining
things
like
adding
cgi.
I
don't
know
if
we
should
have
a
release
or
things
like
that,
but
and
and
then
there
are
things
like
you
know,
rewriting
the
implementation
using
a
different
approach.
That's
mentioned
in
copying.
A
A
quick
win
over
here
would
be
adding
ci
and
then
looking
at
diff
store
implementation
to
see
how
it
works.
There
are
a
couple
of
tests
which
are
going
to
fail.
I
tried
to
run
this
locally
a
while
ago
or
a
couple
of
things
which
are
failing.
That's
okay,
but
we
can
at
least
you
know,
plug
it
into
this
ti
and
see,
get
the
results
in
right.
So
we
are
at
a
stage
in
cupping.
As
of
now
wherein
we
are.
A
Our.
Our
main
goal
is
to
merge,
gets
you
know,
get
working
on
the
camp
and
see
towards
that.
The
cap
is
merged
or
you
will
see.
A
The
caps
a
little
bit
here-
I
can't
yeah
here
all
right,
so
so
we
we
are
working
on
this,
but
apart
from
this
we
are
trying
to
see
if
we
can
harden
all
the
back
ends
that
we
have
right
so
does.
Does
it
make
sense
to
you
the
the.
A
A
Working
group
is
to
make
sure
that
we'll
be
hard
on
the
ci,
and
my
unit
is
how
to
make
sure
that
eg
does
stop
passing
for
all
the
vacuums
and
also
get
the
camera
moving.
So
yeah.
A
First
step
is
to
make
sure
that
all
the
conformance
tests
pass
right.
That's
like
the
basic
first
steps
where
we
are
at
for
most
of
the
back
ends.
I
haven't
taken
a
look
at
iptvs,
but
most
of
the
tests
have
been
passing
and
people
are
working
on
making.
You
know
making
sure
that
the
test
passes,
fixing
the
issues.
A
So
a
quick
win,
if
you
want
to
go
through
the
code
base
of
cupping,
would
be
going
through
this
user
space
trying
to
see
trying
to
make
user
space
run,
and
you
know
you
can
then,
maybe
probably
I
had
a
read
me
over
here.
I
can
pair
with
you
on
this.
I
had
to
read
me
on
how
to
run
cupping
in
user
space
mode.
A
Then,
basically,
finding
your
way
through
this
thing.,
just
this
basically
acts
as
a
main
program
to
you
know
to
get
into
the
base
of
cupping.
Through
this,
you
can
see
how
things
are
working.
As
of
I
think
I
think
as
of
now
this
is
this
only
works
so
ipv4.
So
all
the
ipv6
tests
will
change
so
you
maybe
you,
can
try
adding
ipv6
as
well
over
here.
There
are
like
a
bunch
of
ideas.
B
We
had
a
couple
of
questions
about
user.
Space.
First
is,
doesn't
use
these
being
deprecated
like
like
so
like,
so
then
you
know,
why
is
cupping
supporting
it.
A
One
of
the
main
things
that
one
of
the
main
reasons
why
we
added
using
the
space
back
engine
coupling
is
because
we
were
trying
to
add
all
the
packages
to
their
proxy
to
compete.
We
were
trying
to
record
all
the
bathrooms
using
space
was
one
of
them,
and
user
spaces
is
being
depleted,
it's
being,
as
I
mentioned,
it's
being
removed
from
the
kkk
repository.
So
we
have.
A
A
One
of
the
things
over
here
is
to
rewrite
it
using
diff
store
implementation.
So
this
diff
store
implementation
is
very
specific
to
cupping
right
wherein
you
get
the
entire
state
of
the
kubernetes
networking
model
at
the
back
end
and
then
the
back
end
uses
a
b3
sort
of
a
thing
to
you
know,
get
a
diff
of
what's
updated.
A
We
can
try
to
see
if
we
can
use
this
in
the
user's
basement
so
use
the
whole
idea
of
user
space
is
to
try
out
things.
You
know
for
new
contributors
who
want
to
try
to
view
the
cupping
try
to
experiment.
That's
that's
like
the
way
in
and
then
you
can
look
at
ip
tables,
ibvc
tv.
You
know
it
goes
on
and
on.
A
Absolutely
absolutely,
but
you
may
have
to
play
around
with
all
these
kit
of
action
styles-
and
you
know,
on
your
way
through,
you
can
probably
add
a
couple
of
readmes
to
use
space,
and
you
know
whatever
you've
been
experimenting
on
whatever
you've
been
learning.
I
can
pair
with
you
on
that.
If
you
need
any
help,
gi
hangout
on
big
network
kpng
channel
on
slack
so
feel
free
to.
B
A
Anytime
and
once,
and
maybe
your
next
goal
would
be
to
you
know,
experiment
with
the
other
back
ends,
you
know,
then
all
of
these
issues
would
make
much
more
sense
to
you
and
then
you
can
pick
up
any
of
the
other
issues
and
then
start
working
on
that
or
help
us
with
the
cap.
You
know.
B
But
yeah,
this
is
the
thing
where
all
can
I
continue
to
copy?
Okay?
This
is
a
good
start
and
then
I'll
I'm
hopefully.
Ultimately,
this
will
allow
me
to
like
contribute
to
like
the
actual
logic
as
well
like
the
actual
buttons.
A
One
thing
that
I
would
like
to
point
out
over
here
is:
if,
if
I
remember
you
collect
correctly,
I
p
tables
when
you
use
the
space
share,
a
similar
mechanism,
the
way
they're
trying
to
use
cupping
the
way
they're
trying
to
create
iptables
rules.
So
if
we
are
able
to
get
diff
store
implementation
in
user
space,
we
can
try
to
port
over
to
ip
tables
as
well.
So
we've
got
hana
month
and
we
wake
who
help
us
with
iv
tables
and
itvs
backends.
A
They
also
hang
out
on
the
slack
channel,
so
they
sort
of
own
iptables
and
ipvs,
so
you
can
discuss
tip
store
with
them
as
well.
At
some
point,.
B
All
right,
okay,
oh,
that
sounds
great.
I
will
look
into
this
as
soon
as
possible
to
probably
today
or
tomorrow,
and
adding
the
thanks
a
lot
for
bringing
me
at
this.
A
Awesome
thanks
thanks
for
joining
and
I
don't
see
any
of
anyone
else
joining
this
meeting,
but
you,
if
you
have
any
other
questions,
comments,
discussion,
time,
stem
cpg.
B
A
Cool
and
if
it
works
out,
maybe
we
can
pair
on
this
next
week
as
well.
You
know
if
you're,
okay,
if
you're
not
following
yeah,.