►
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
Hey
folks,
this
is
cluster
API
provider
AWS
office
hours
and
we
are
today,
is
July,
10th,
2023
and
so
a
couple
of
things
too.
Before
we
start
so
we,
if
you
want
to
speak,
you
want
to
add
any
agenda.
You
can
add
into
the
doc
which
is
shared
in
the
chat,
and
if
you
want
to
speak,
you
can
raise
your
hand
and.
B
A
I
will
pick
who
can
speak
and
it's
good
that
we
abide
by
the
cncf
code
of
conduct
and
also
is
there
anyone
who
is
new
and
want
to
introduce
themselves.
They
can
unmute
themselves.
C
D
Hello,
yeah,
here's
one
new:
we
are
here
yeah
and
you
want
to
I
come
from
so
right
now
in.
E
D
The
Berlin
then
Hawaii
Joy's
meeting
is
the
our
my
our
King
right
right
now,
it's
investigator
because
API
is
our
container
illustration
tool
yeah
to
manage
our
kubernetes
install
we
jumped
into
the
clouds
the
API.
It
will
be
provided
to
try
to
use
some
try
to
use
the
AWS
management
Pro.
For
this,
then
we
encountered
some
issue
then
hoping
we
can
get
some
help
in
this
meeting.
A
Great
great
nice
to
see
you
Julian,
okay,
so
we'll
start
with
the
PSA
for
this
week.
So,
oh
I'm,
sorry
is:
does
anyone
else
want
to
introduce
I,
miss
them?
No?
Okay,
okay,
so
we'll
start
with
PSA.
So
this
release
we
had
so
this
week
we
had
release
for
Kappa
version
2.2.0
and
you
can
go
ahead
and
see
the
release.
Link
I
will
put
the
link
here
and
you
can
go
ahead
and
see
what
all
new
features
have
been
released.
B
Yeah
sure
I
can
that's
basically
a
follow-up
of
what
we
had
in
the
last
few
weeks.
I
can
recap
of
this
managed
subnets
broken
feature.
What
it
means
is
copper
can
either
create
the
subnet
for
you,
that's
the
managed
one
or
you
can
print
your
own
subnet,
which
means
you
define
the
ID
and
it
just
gets
taken
for
the
cluster.
B
B
We
discussed
it
at
length
in
this
draft
here
which
I'm
working,
which
is
a
figure
scope,
so
that
the
whole
network
encoding
couple
and
there's
also
the
issue
linked
here
with
many
other
network
use
cases.
The
issue
is
called
API
Evolution
for
VPC
networking
topologies.
B
It
makes
us
more
flexible
and
we
all
had
the
we
were
I
think
three
or
four
people
in
the
meeting,
and
we
all
thought
that
this
kind
of
makes
sense
as
a
solution
rather
than
extending
the
existing
API
types
and
trying
to
fit
everything
in
make
it
separate
and
there's
some
Main
benefits
which
I
wrote
down
here.
As
a
summary,
so
you
can
create
the
network
separately,
which
is
good.
People
said
that
it's
good
for
end-to-end
tests,
because
they're
super
slow
and
pre-creating
a
network
might
make
sense.
B
B
So
this
might
just
make
the
whole
thing
more
stable
and
less
flaky
regarding
box
yeah,
and
then
we
ended
up
saying
that
first
of
all,
there's
rumor
that
copy
might
introduce
a
load
balances
here.
C
Yeah
yeah
a
couple
of
couple
of
things:
yeah
yeah,
just
to
reiterate
what
Andrea
said,
so
we
we've
known
for
a
long
time
that
the
network
inside
of
Kappa
is
in
need
of
a
overhaul
for
you
for
at
least
two
two
and
a
half
years,
especially
because
people
want
different
flavors
of
of
network
yeah.
So
this
is.
C
This
is
really
good
if
we
can
start
to
work
on
this
and
allow
that
customization
without
having
to
really
bring
your
own
network
or
you're
you're,
doing
your
infrastructure,
I
guess,
from
a
purely
practical
point,
I
I
yeah
I,
like
the
idea
of
a
proof
of
concept,
I
think
that
would
be
really
really
helpful
and
help
shape
usability
and
and
stuff
like
that,
I
guess.
The
other
question
is:
when
would
we
what
time
frame
would
we
do
this?
C
Would
we
need
a
API
version,
bump
I,
suspect
so
and
another
question
and
another
statement
really
is.
This
would
also
help
with.
We
also
have
a
desire
to
move
our
managed,
so
our
eks
implementation,
so
that
it's
more
in
line
with
the
managed
kubernetes
proposal,
which
basically
means
we're
going
to
have
to
strip
out
the
networking
of
the
manage
control
plane
anyway.
At
some
point,
so
this
may
help
with
that
migration.
So
it's
lots
of
I
just
waffled
there,
but
hopefully
some
of
it
made
sense.
B
Yeah
I
think
the
word
of
migration
is
kind
of
important
here
and
that's
those
critical
thing
which
I
want
to
find
out
in
a
prototype
implementation
that
there
will
be
a
separate
controller
reconciling
the
network.
The
old
networking
code
will
also
still
exist
for
a
certain
time
frame
and
then
we
have
to
see
how
do
we
migrate.
B
This
whole
thing
right
because
cluster
creation
depends
on
network,
so
the
controllers
have
to
kind
of
depend
on
each
other
in
some
way,
so
yeah
I
had
to
have
to
see
if
that
then
creates
a
dependency
and
there's
a
new
API
type
version
yeah.
Let's
just
see
what
happens
to
start
implementing
now
we'll
see
that.
A
Yeah
I
think
when
we
introduce
the
new
crd,
we
might
need
new
API
version
so
yeah,
but
but
that
should
be
okay
like
we
can
have
that.
Maybe
a
next
major
release
or
something
and
then
we
can
introduce
and
that
that
should
not
be
an
issue
so
and
we
already
have
an
issue
I
think
for
P1
beta
3.
If
I'm
not
wrong
so
yeah
we
can.
We
can
clap
this
all
together
in
one
issue
and
then
maybe
we
can
release
a
new
API
version,
so
that
should
not
be
the
issue.
A
My
concern
only
is
this
about
the
load
balancer
crd,
which
is
getting
introduced
in
Cappy.
So
does
anyone
has
any
like
concrete
information
on
this?
Yes,
project
go
ahead.
C
It's
not
it's
not
concrete.
So
there's
just
been
mentioned
a
bit
on
a
few
issues
recently
that
perhaps
we
you
know
perhaps
Cathy
could
do
with
a
a
load,
balancer
resource
kind,
so
yeah
I,
don't
think
it's
actually
concrete.
As
far
as
I'm
aware
of
yet.
A
A
Great
okay,
moving
on
to
the
next
topic,
Alberto.
E
Hey
there,
thank
you.
Yeah
I
just
wanted
to
hear,
hear
that
PR
that
we
started
for
water
support.
For
those
who
don't
know,
Russia
is
basically
the
open
shift
managed
offering
that
we
provide
to
only
with
AWS,
so
implementation
wise.
That
would
look
very
similar
to
other
managed
kubernetes.
E
Via
copy
so
be
similar
to
The
eks
implementation,
for
example,
so
yeah
I
started
that
PR,
like
that's,
providing
like
the
very,
very
basic
support
for
it
like
you
can
create
a
cluster
and
pretty
much
that's
it.
So
you
can
expect
to
see
folks
like
myself,
Mulan
Vines,
to
follow
up
on
this
work
and
I
just
wanted
to
make
sure
that
we
are
all
aligned
and
we
love
to
to
get
this
in
the
Kappa
repo
and
to
be
able
to
collaborate
there.
C
E
C
Yeah,
sorry,
just
on
that
last
topic,
so
I
guess
we
could
try
and
get
this
reviewed
and
merge.
We
could
get
that
into
2.3
release
eventually
in
the
future.
Yeah
should
be
good,
yeah,
cool,
yeah.
Sorry,
sorry!
So
the
next
point
was
I'm.
Gonna
make
a
change
this
week
to
change
the
release,
notes
generation
back
to
using
the
the
prowl
Plug-In
or
prowl
whatever
they
call
them
extension,
mainly
because
the
release
notes
generation
that
we
currently
have
has
two
faders
doesn't
work
very
well
across
branches.
C
So
if
you're
trying
to
do
a
diff
between
the
two
branches
from
say
the
2.1
release
and
the
2.2,
it
doesn't
work
well
and
there's
duplicates
and
also
there
is
no
way
to
flag
currently
the
action
needed
which
is
the
equivalent
in
in
the
prow
world.
So
if
you
have
the
action
needed,
it
will
automatically
put
up
hey.
C
E
A
I
have
one
question
so
I
think.
Last
year
we
tried
using
this
right,
the
pro
plugin
and
we
removed
the
release
node
generation,
and
then
we
rewarded
back
because
we
Face
some
issue
but
I,
don't
remember
exactly
what
we
face.
So
do
you
remember
it,
sir.
C
So
we
did
have
the
prowl
plugin
for
quite
a
long
time,
and
then
we
changed
it
to
the
GitHub
release,
notes
generation,
just
because
it
had
a
few
nice
features
in
like
marking,
people
that
had
their
first
contribution
to
the
repo
which
was
really
nice
to
call
those
contributors
out.
C
So
that
was
a
that
was
a
nice
to
have
and
that's
originally
why
we
moved
to
that
so
and
I
think
when
we
initially
tried
to
move
back.
There
was
a
few
issues
with
just
identifying
the
right
commit,
but
I
think
we
can
pretty
much
roll
back
to
what
we
had,
which
was
working
yeah.
A
C
A
C
Yeah
I
think
last
time
it
failed.
It
was
that,
but
we
need
to
fix
that
as
well.
Don't
we
try
that
was
that
unscheduled,
a
gender
item.
A
A
C
C
So
just
a
question:
if
anyone's
seen
any
issues
with
the
current
proud
jobs,
since
we
migrated
across
to
the
eks
cluster
but
I've.
C
A
I
think
only
issue
that
I
see
is
the
GitHub
token
expiry.
I
guess.
E
A
A
A
Related
to
the
e-gas
cluster
that
we
moved.
A
No,
then
thanks
a
lot.
There
will
be
an
after
hours
where
we
could
discuss
some
help
with
the
issues
or
Buck
triaging
or
anything
else.
So,
if
folks
have
anything
regarding
that
agenda
because
they
could
stay
back
and
for
rest
of
the
folks,
I
think
see
you
all,
maybe
in
the
next
office.