►
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
Cool
hello,
everyone
welcome
to
the
cluster
api
provider,
aws
aws,
sorry
office
hours
on
monday,
the
10th
of
january
2022
have
a
new
year
to
everyone
on
that
note,
just
a
reminder
that
we
abide
by
the
cncf
code
of
conduct,
so
please
be
kind
to
each
other.
A
A
Please
can
you
add
yourself
to
the
attendee
list
and
add
any
topics
that
you
would
like
to
discuss
to
to
the
meeting,
just
a
reminder
that
we
use
the
the
raise
hand
feature
if
you,
if
you
want
to
talk
at
any
point,
and
if
this
is
your
first
meeting,
please
feel
free
to
unmute
yourself
and
say
hello
and
introduce
yourself
if
you
want
to
start
looking.
You
know
you
don't
have
to
so
I'll.
Just
go
and
mute
for
a
minute
or
so
just
to
see.
If
anyone
wants
to
do
that.
B
Hi
yeah
I'm
just
joining
first
meeting.
My
name
is
shaolin
nice
to
see
you
all
happy
new
year.
A
Cool
yeah,
if
anyone
does
want
to
it
changes
their
mind,
wants
to
introduce
themselves
just
feel
free
to
raise
your
hand,
and
you
can
be
you
can
introduce
yourself
so
starting
with
the
the
psas,
so
a
pretty
big
psa
to
start
with
and
start
the
new
year,
so
nadir
who's
been
a
long-term
maintainer
of
kappa
has
stepped
back
from
kappa
and
will
be
won't,
be
involved
going
forwards,
and
so
he
has
been
moved
to
emeritus
maintainer,
which
is
a
I
guess,
a
sad
loss
for
the
project.
A
As
he's
been,
I
guess
a
mentor
for
for
a
couple
of
years
now
on
this
project,
and
you
know
you
can
look
at
his
contributed
stats
on
the
repo,
but
it's
a
big
loss,
but
he's
moved.
Please
correct
me
if
I'm
wrong
to
a
a
team
that
will
be
consuming
cappy
upstream
within
vmware.
So
I
guess
he's
not
here
but
thanks
to
dear,
especially
for
me.
C
Yeah
I
want
to
thank
nader
as
well
thanks
for
all
his
contributions.
He
did
great
things
for
this
project,
especially
in
committee
building.
He
had
a
lot
of
us
like
ramping
up
and
he
will
definitely
be
missed
so
yeah,
thanks
for
all
his
contributions.
A
Yep
sad
days,
but
almost
not
but
so
yeah.
I
guess
on
on
that,
though,
obviously
the
more
maintainers
and
more
reviewers.
We
have
really
helps
the
project
so
yeah
if
anyone's
new
and
you
know,
needs
help
getting
started,
feel
free
to
reach
out
to
any
of
the
existing
reviewers
or
maintainers.
A
And
you
know
we
can
help
you
through
that
process
of
your
first
commit
or
or
you
know,
moving
to
the
to
the
next
level
so
on
to
the
the
actual
items
from
the
the
meeting
literally
before
the
holiday
break.
So
sadef
yours
was
the
first
one
around
google
summer
of
code
and
linux
foundation,
mentorship.
C
Yes,
so
I
created
a
project
in
the
website
itself,
but
it
is
not
yet
they
they
didn't
open.
Cncf
didn't
open
the
spring
repo
directory.
Yet
so
we
need
to
create
a
project
there
as
well
for
having
a
like
paid
internship.
So
I
am
following
up
on
that.
I
don't
know
when
google
summer
of
code
will
will
open,
but
it
should
open
soon.
So
if
we
have
more
projects
that
we
want
someone
to
work
on,
we
can
apply
that
also,
but
right
now,
I'm
not
planning
to
do
so.
C
Let
me
know
if
you
have
any
projects
that
maybe
that
can
be
doable
in
three
months
period,
especially
that's
all,
and
the
second
one
is
from
me
as
well
so
yeah.
Actually,
since
ignition
pr
is
coming
to
a
good
position,
the
tests
are
passing.
I
want
to
actually
I'm
gonna
build
the
flat
car
images
so
tilio.
Could
you
remind
me
what
I
should
be
careful
about
during
building
those
images.
D
In
in
general,
please
be
careful
regarding
any
issues
and
you
run
into
and
then
bring
us
and
yell
at
us,
because
we're
gonna
keep
this
smooth
and
functioning
there
shouldn't
be
any
any
larger
issues
to
that.
It
should
just
just
work
straight
forward.
If
it
doesn't
it's
on
us.
C
D
It's
it
is
part
of
image
builder,
I
think
so
everything
should
be
in
there.
D
No,
no
there's,
no
there's
no
magic
involved.
I
think
the
only
input
should
be
the
respective
flag
car
version
number.
So
we
have
new
releases
every
now
and
then
they're
an
image-based
distribution
and
at
that
point
either
you
pick
it
up
or
we
just
ping
you
or
we
create
some
kind
of
automation
that
picks
it
up
and
then
yeah
you'll
need
to
issue
a
new
aws
class
api
images
as
well.
D
I
can
put
you
I
can
team
you
up
with
the
with
the
engineer
who
took
care
of
that
in
the
past.
If
that
has
any
help.
B
A
Well,
thanks
to
deaf
and
next
one
is,
is
me
so
we
we
have
started
to
talk
about
ipv6
support
in
kappa,
both
unmanaged
and
managed,
so
that
there
was
a
document
that
has
been
started.
I
did
have
a
task
to
update
that.
I
haven't
done
that
yet
to
be
honest
over
the
over
the
whole
day
break,
however,
we've
been
discussing
internally
and
we've
worked,
so
ipv6
was
recently
added
to
eks
cuttle.
A
So
we've
done
a
lot
of
work
in
that
area
and
I'm
I'm
hoping
some
of
the
engineers
from
the
eks
custom
team
will
be
contributing
to
capra
in
the
future.
I've
been
persuading
them
and
running
a
deep
dive
for
them
this
week,
so
I'm
hoping
to
get
them
to
do
the
proposal
for
the
ipv6.
So
if
anyone's
interested
in
that,
we
can
also
set
up
a
meeting
on
that
to
discuss
it
and
then,
obviously,
when
the
proposal
comes
out,
review
it
so
the
next
one
was
wasn't.
A
The
dear
yeah,
thanks
to
def
about
reviewing
the
open
ignition
pr
and
zephy
just
said,
you'll
take
that
over
yeah.
I
can
also
have
a
look
as
well
if,
if
needed
as
well,
if
you
want
another
pair
of
eyes,
cool
so
says
ft,
do
you
want
to
do
the
next
one
on
the
conformance
tests.
C
Sure
yeah
richard
brought
up
that
maybe
we
can
consider
about
getting
confirmed
with.
We
are
already
running
the
conformance
test,
but
we
do
not
submit
our
test
results
to
communities
conformance
repository
to
get
officially
compartment.
We
were
discussing
if
it
would
make
any
sense
for
our
case.
I
checked
it
with
nadir
earlier.
C
He
he
made
a
good
point
regarding
so
we
are
actually
blocking
kubernetes
releases.
Our
test
signal
is
blocking
kubernetes
releases,
so
it
is
a
good
indication
and
may
be
good
enough
for
not
being
officially
conformant,
because
it
is
too
much
work
if
we
think
about
it.
We
have
three
different
branches,
viva
tree
v1,
alpha
4
and
v1
beta
1,
and
we
need
to
submit
test
results
for
each
kubernetes
versions.
C
To
that
repository
I
mean,
if
there
is
any
significant
benefit
of
being
confirmed,
we
can
consider
it,
but
in
my
point
of
view
it
is,
it
may
be
too
much
work.
A
Yeah,
I
I
guess
one
comment
I
have
on
that
is:
is
for
people
that
build
their
their
solutions
based
on
kappa
and
when
things
like
rfps
come
in,
they
always
ask
the
question:
are
you
conformant
and
if
you
use
cappy
and
kappa,
you
can't
really
necessarily
answer
that
question
as
true
so
yeah?
I
don't
know
whether
that's
good
enough
reason
to
do
it
or
not.
A
Yeah,
well,
we
can
always.
We
can
always
discuss
it
in
chat
as
well
one.
We
can
always
write
up.
Go
on
to
left.
C
Also
one
thing:
I
noticed
that
the
conformant
projects
are
mostly
does
not
belong
to
like
kubernetes
six,
so
I
mean
it
makes
more
sense
the
outside,
like
outside
projects,
to
get
compartments,
and
also,
as
I
said
since,
we
are
actually
blocking
current
releases
on
our
conformance
test
results.
C
We
could
actually
say
we
are
confirmed
but
did
not
get
the
package.
I
mean
I'm
definitely
open
to
discussions.
If
this
is
important.
A
A
Was
it
cap
v
and
capd
under
the
under
the
covers
so
interesting
to
see
what
what
they
used
to
be
conformant,
but
we
could
follow
on
that
in
an
issue
cool,
so
the
next
one
is
we,
we've
got
an
older
tickets
here
about
moving
away
from
make
to
to
use
something
like
mage
a
bit
like
well,
we'll
see,
what's
the
ruby
on
rails
rake,
but
in
go
so
there
was
some
discussion
about
this
for
a
while,
but
I
was
just
wondering
if
there's
any
benefit
to
that.
E
Yeah
I
mean
what
right
in
the
description
right,
the
make
file
itself
has
become
unwieldy.
Maybe
our,
I
guess,
are
there
concrete
issues
that
that
we
need
to
fix
with
a
make
file?
I
have
maybe
one
sort
of
on
on
my
mind,
but
maybe
that's
maybe
that's
what
we
could
do
in
instead
of
of
this
pr
is
just
identify.
E
A
F
Yeah
kind
of
along
the
same
lines,
it
sounds
like
there's
a
few
kind
of
smallish
annoyances,
but
you
know
if
our
build
is
complex,
it's
likely
to
be
complex
wherever
we
move
it.
I'm
guessing
moving.
Complexity,
doesn't
necessarily
eliminate
complexity,
and
there
is
definitely
a
probably
a
large
switching
cost,
not
just
to
the
person
that
does
the
pr,
but
to
everyone
to
now
learn
the
new
build
system.
F
A
E
Yeah,
I
guess
yeah.
I
just
request
for
comments
to
this
pr.
If
you
have
like-
and
you
know
some
anecdotes
about.
Oh
this,
the
you
know
this
part
of
the
make
file
is
confusing.
Whatever
went
you
know,
maybe
as
a
new
contributor
etc.
When
I
was
trying
to
accomplish
something
specifically
and
then
maybe
you
know,
we
can
take
each
of
those
comments
and
maybe
make
it
an
issue
or
just
or
just
maybe
just
make
it
an
issue
and
just
tag
tag
decision.
E
Then
we
can
close
this
and
but
not
forget
the
fact
that
that
we
wanted
to
so
yeah.
I
guess
either
comments
or
or
create
an
issue.
Please,
if
you
have
some
some
experiences
with
a
make
file,
something
something
that
you'd
like
to
get
fixed.
A
Right
so
sid
f!
Yours
is
the
next
item.
C
Yes,
last
week
I
updated
the
bank
cappy
version
to
v1.1.0
beta.
I
updated
the
unmanaged
clusters
e2e
test
but
did
not
get
updating
eks
test.
Yet
so
probably
it's
fading
because
of
that,
but
but
eksters
are
constantly
failing
and
I
don't
see
it
when
I
open
the
console.
There
is
a
summary
area
right.
I
don't
see
that
failure
in
the
council-
I
don't
know
why.
C
No,
I
see
it,
it
is
just
a
bit
down
below
sorry
yeah
yeah,
please
ignore
yeah
yeah,
I
see
it,
it
is
the
periodic
eks
e2e
main
test.
A
A
Yeah
that'd
be
good
and
that's
probably
explaining
there
was.
I
was
running
the
tests
on
a
change
for
auto
scale,
group
tags
it
kept
failing
on
the
cluster,
cut
or
win
it,
and
that's
just
it's
probably
that
isn't
it
so
I'll
get
that
fixed
I'll.
Add
that
too,
to
my
list
to
do
cool
thanks
to
def,
so
dimitri
you,
you
have
the
next
house.
G
Room
thanks
yeah,
so
I
I
started
a
conversation
in
slack
and
wanted
to
introduce
this
topic
here,
but
basically
the
gist
is
we
have.
We
have
some
kubernetes
package
builds
and
kubernetes
versions
that
have
some
metadata
at
the
end,
like
fips,
for
example,
but
the
ami
name
doesn't
support
plus
signs
in
it,
and
so
we
substitute
and
build
the
amise
with
a
hyphen.
G
But
then,
when
trying
to
use
this
with
the
like,
the
image,
lookup
format
feature
is
the
the
format.
Basically,
the
go.
Template
is
just
taking
the
kubernetes
version
as
it
is
and
not
doing
any
manipulation
to
it,
obviously,
and
then
so
yeah.
I
just
wanted
to
see
if
you
know
first,
if
anyone
ran
into
this
or
if
people
would
be
open
to
having.
You
know
some
sort
of
logic
that
converts
a
like
a
a
valid
several
version
to
something
that
amis
can
actually
be.
F
Every
time
this
is
incredibly
timely
because
I'm
internally,
I
just
picked
up
a
ticket,
an
internal
ticket
to
look
at
ami
lookups.
F
A
C
How
do
you
do
the
lookups?
So
are
you
not
manually
adding
the
image
names
because,
right
now
we
do.
We
do
not
do
lookups
on
separate
repositories
for
for
the
mis,
how
the
lookups
work
in
your
cases.
G
We're
using
lookups
in
our
own
aws
account
on
a
different
account,
so
it's
yeah.
These
are.
These
are
amis
that
we
are
building
yeah
and
absolutely
am
I
as
we're
building,
and
also
doing
that
substitution
of
changing
the
plus
to
a
hyphen.
Actually,
packer
is
doing
that
for
us.
F
Well,
for
eks
we're
using
ssm
parameter
store
because
amazon
publishes
amis
themselves
and
and
makes
those
available
in
the
look
up
in
each
aws
account
for
each
region
for
different
architectures,
so
that
that
only
really
solves
that,
although
you
can
put
custom
things
in
parameter
store.
I
wonder
if
there
is
an
extension
of
that
mechanism
that
could
be
leveraged
by
kappa
to
publish
our
custom
ones.
C
Yeah
there
has
been
discussions
a
few
months
back
supporting
different
kinds
of
lookups,
but
we
didn't
make
much
progress
on
that,
but
yeah.
It
would
be
great
to
think
about
it
to
cover
many
scenarios
and
not
just
look
up
the
name
matching.
Also.
C
We
should
make
sure
that
the
changes
are
not
specific
to
what
we
are
seeing
here
and
if
you
have
different
use
cases,
let's
keep
track
of
them
and
do
the
changes
at
once.
My
suggestions.
A
C
A
A
No
cool,
I
guess
the
next
question
is:
do
we
start
with
the
issue
triage
again
or
do
we
wanna
start
that
from
the
next
meeting,
if
anyone's
interested.
A
C
We
don't
have
lots
of
issues
came
to
the
holiday
season.
If,
if
you
want,
we
can
resume
it
in
the
next
meeting.
Up
to
you.
A
Yeah-
it's
probably
it's
probably
better,
isn't
it
maybe
too
much
for
the
first
meeting
back
cool?
We
can
do
that.
So
we'll
we'll
add
a
add
some
time
we
can
take
the
time
off
on
the
next
meeting,
can't
be
the
remaining
time
to
do
the
issues
brilliant
well,
if
no
one
else
has
any
anything
else
to
discuss,
we
can
call
it
a
day
there
thanks
for
joining
and
happy
new
year.
Everyone.