►
From YouTube: SIG Cloud Provider 2023-08-30
Description
Meeting agenda: https://docs.google.com/document/d/1OZE-ub-v6B8y-GuaWejL-vU_f9jsjBbrim4LtTfxssw/edit#bookmark=kix.pwxikdozsfan
A
All
right
welcome.
Everyone
today
is
Wednesday
August,
30th
2023,
and
this
is
the
Sig
cloud
provider.
Meeting
Sig
cloud
provider
is
a
sub-project
of
kubernetes
sigs
and
as
such,
we
follow
their
meeting
guidance,
which
basically
says
please
treat
everyone
kindly
and
if
you'd
like
to
speak,
please
raise
your
hand,
although
there's
only
a
couple
of
us
here
so
I,
don't
think
it'll
be
that
big
of
a
deal
so
looking
at
our
agenda
for
today,
I'm
not
seeing
any
sub
project
updates.
A
Although
we
do
have
some
Representatives
here,
does
anyone
want
to
give
a
sub-project
update.
D
A
All
right
cool
thanks,
John
yep,
all
right
so
I
guess
that's
our
provider
updates!
Let's
get
to
the
regular
agenda.
Kirsten
you've
got
the
first.
You've
got
the
first
item.
So
please
take
it
away.
C
C
This
is
just
adding
a
new
flag
to
replace
the
existing
cloud
Excel.
This
would
be
like
mostly
to
be
used
after
migrating.
All
the
cloud
products
out
of
tree
sorry
enter
from
August.
A
C
No,
this
is
this
is
just
like.
Instead
of
the
it
is
going
to
add,
like
a
club
produce
as
external
So.
Currently
we
have
this
cloud
provider,
external
and
some
provider
name
right.
This
is.
C
Yeah,
instead
of
like
using
the
existing
flag
to
say
that
we
are
using
the
external
collaborator,
we
are
adding
a
new
flat.
That
way,
we
can
just
say
it's
external
or
not.
Altribute
will
just
have
like
a
current
I
think
we
say
like
gcp
and
AWS
and
other
stuff
right
in
the
using
the
flag.
A
C
But
yeah
I
think
one
of
the
comment
was
to
just
get
approval
from
both
the
sick,
Lord
provider
and
sign
networking
before
we
can
I
mean
get
approach
from
both
the
six
like.
C
If
it's
going
to
block
any
of
the
I
mean
if
there's
any
unknowns
which
this
might
cause,
they
should
test
me,
but
safety
precautions.
B
C
C
Note
because,
like
it
started
like
okay,
what
is
this
flag
going
to
be
used
for,
like
especially
when
we
remove
the
cloud
providers
from
the
entry
and
that
started
a
discussion
onto
the
one
is
the
tense
another
one
is
the
node
IP
related
things,
and
that
went
into
that
discussion
more
whether
we
can
re
I
mean
remove
the
node
IP
thing
from
clock,
loader
logic
or
not,
are
like
a
and
talking
discussion
to
on
a
different
cap
like
which
is
working
on
some
on
the
Node
IP
particle.
C
A
Yeah
I
mean
to
me
like
at
a
high
level,
I
think
you're
right.
We
still
do
need
to
be
able
to
say
whether
you're
using
a
CCM,
you
know
in
external
mode
or
you're,
not
using
any
CCM,
so
I
think
the
the
binary
nature
makes
sense.
But
yeah,
like
I,
haven't
looked
at
this
issue.
So
I,
probably
you
know,
I
agree
with
Bridget.
Probably
we
need
to
spend
some
time
kind
of
reviewing
and
leave
our
comments
here
and
and
figure
out
what
the
best
step
forward
is
because.
D
C
C
A
Okay,
so
this
came
up
earlier
because
some
of
the
openstack
folks
were
trying
to
update
a
PR
and
they
wanted
to
get
approval
from
us.
But
it
turns
out
that
the
the
aliases
that
we
have
in
the
KH
dot
IO
repo
only
have
I
think
Andrew
and
Walter
as.
B
A
All
right
so
I
wanted
to
bring
up
I
guess,
like
probably
this
would
come
up
in
the
triage,
but
since
we
have
been
pinged
about
it,
I
just
wanted
to
bring
it
up
specifically
so
I
believe
her
name
is
Lakshmi.
She
pinged
and
Chad
about
this
docs
related
issue.
That
came
up
so
she
had
been
asking
me
about,
like
in
in
our
running
Cloud
controller
manager,
docs
in
the
official
kubernetes
documentation.
A
You
know
it
says:
Cube,
API,
server
and
Cube
controller
manager
must
not
specify
the
cloud
provider
flag,
but
currently,
if
you're
trying
to
run
an
external
cloud
provider,
you
need
to
specify
the
flag
for
external.
So
she
was
looking
to
to
put
forth
a
PR
for
that
and
created
an
issue
and
was
just
looking
to
see
if
we
would
approve
it
on
triage.
A
B
A
Yeah
she
and
I
went
back
and
forth
on
it
in
terms
of
like
getting
the
the
guidance
updated,
so
I
think
that
part
yeah,
totally
agreed
that'll
need
some
more
reviews,
but
I
think
in
terms
of
just
getting
the
the
issue
accepted
like
it
probably
makes
sense.
Okay,
so,
hopefully
is
it.
Does
anyone
knows
the
command
like
that?
Is
it
triage
Dash
accepted.
B
A
Okay,
so
next
topic
looking
to
poke
an
Azure
maintainer
about
this
issue
here,
so
glad
that
Bridget
you're
here.
B
So
I
took
a
look
at
this
one
and
I
scrolled.
All
the
way
down
and
I
was
like.
It
looks
like
the
last
thing.
El
Nico
actually
said
to
this
guy
was
you
should
rebase
and
I
did
pay
him
over
on
kubernetes
slack
just
to
kind
of
see
if
he
wanted
to
try
to
do
the
rebase
but
which
it
looks
like
maybe
has
not
happened
yet,
but
you're
saying
that
maybe
rebasing
isn't
going
to
solve
it
or
well.
A
It's
going
to
need
to
be
fixed
in
a
rebase,
but
here
was
the
I
think
the
big
issue
was
that
you
know,
like
Damiano,
opened
this,
this
PR
on
May
16th
things
have
changed
and
what
has
happened.
Yeah
is
that
there
have
been
successive
changes
that
have
come
in
and
he's
rebased
it
a
couple
times
to
follow.
Those
so
like
I,
think
we're
trying
his
patience
a
little
bit
to
like.
E
A
B
Yeah
I
think
I'm
gonna
have
to
check
in
with
penpei
and
see
what
is
happening
because
I
think
the
most
recent,
if
I
scroll
back
up
I,
remember
there
being
a
discussion
of
hey.
We
want
to
put
this
on
hold
or
something
but
I'm
trying
to
remember
if
this
is
that
one
yeah,
so
at
least
at
one
point,
pancake
was
like
we're
working
on
this.
You
know,
let's
hold
so
I'm,
going
to
try
to
see
if
I
can
get
an
answer
from
him
as
to
are
we
still
holding?
B
Find
out
if
there's
something
that
isn't
reflected
in
the
issue
that
we
can
communicate.
B
A
A
It's
fine,
if,
if
it's
fine,
if
there's
like
still
discussions
happening
on
your
end,
we
just
it's
like
it
has
been
some
time
yeah.
Well,
it's
been
it's
been
a
couple
months,
but
also
like
and
I
think.
This
is
probably
like,
maybe
a
unique
problem
that
redhead
has,
which
is
like
as
a
downstream
consumer
as
we
go
to
rebase
some
of
these
things.
We
we're
hitting
this
now
that
we
can't
like
we're
having
trouble
injecting
changes
on
top
of
the
rebases,
and
you
know
that's
our
problem.
You
know
for
sure,
but.
A
A
So
we
were
recently
our
security
team
was
doing
an
audit
and
review
of
Cloud
controller
managers
in
openshift,
and
one
of
the
things
they
noticed
was
that
the
CCMS-
or
at
least
the
ones
they
were
looking
at,
did
not
have
comprehensive
audit
logs
for
the
activity
that
the
cloud
controller
manager
is
taking
with
the
cloud.
So,
for
example,
updating
load,
balancer
type,
Services
or
you
know
interacting
with
the
node
instance
objects,
and
so
they
wanted
us
to
to
look
into
the
possibility
of
kind
of
improving
the
audit
logging.
A
Now
this,
obviously
every
cloud
provider
is
going
to
be
doing
things
a
little
differently
and
they
have
their
own.
You
know
kind
of
mechanisms
for
the
controller,
but
maybe
it's
something
we
could
add
to
the
cloud
provider
kind
of
root.
You
know
Library.
Maybe
we
could
put
a
framework
there.
That
would
give
everyone
kind
of
some.
A
You
know
very
similar
functions
that
you
know
so
so
all
the
cloud
providers
could
use
the
same
set
of
functions
for
doing
auditing
type
stuff
or
or
creating
the
audit
logs
anyway,
like
I
I,
don't
know
what
the
ultimate
solution
would
be,
but
I
wanted
to
bring
it
here
just
to
kind
of
open
the
discussion
up.
So
does
anyone
have
thoughts,
I
guess.
B
A
Base
right,
the
behavior
yeah
so,
for
example,
like
if
a
user
creates
a
service,
and
that
requires
the
CCM
to
update
a
load
balancer
in
the
infrastructure,
having
like
an
audit
log
of
that
or
having
a
log
of
that
activity.
So
those.
B
A
Right
right,
then,
having
an
audit
log
of
that
or
having
some
sort
of
comprehensive
log
of
that
you
know
kind
of
like,
and
this
can
get
like
I
guess,
we'll
have
to
decide
how
verbose
we
want
it.
But
you
know
something:
that's
like
user
X
has
tried
to
do
y
in
situation.
Z
and
the
result
was
you
know,
ABC
or
whatever.
B
B
Devil
is
going
to
be
in
the
exact
details
that
they're
hoping
to
see.
A
No
I
think
absolutely
we
can
and
in
fact
that's
kind
of
why
I
wanted
to
bring
it
just
lightly
like
this.
Is
that
yeah?
If,
if
this
sounds
like
an
interesting
topic
for
the
Sig,
then
yeah,
what
I
would
do
as
the
next
step
is
probably
put
together
a
document
that
I
could
share
with
everyone.
A
Just
saying
like
this
is
what
I
would
want
to
see
us
do
and
if
that
turns
into
a
cap
for
us
or
something
or
whatever
you
know,
whatever
the
next
step,
is
there
I'm
happy
to
go
that
direction
so
yeah
like
if
you
think
it
sounds
interesting
and
nobody
else
on
the
call
has
has
kind
of
any
objections
or
anything
yeah
like
I'm,
happy
to
take
the
next
step
of
kind
of
documenting
this
and
and
proposing
it
as
like,
an
enhancement
to
the
Sig
or
something.
D
A
Okay,
cool
well
I
will
take
as
an
action
then
to
write
up
a
description
of
what
I
think
we
should
do
here,
and
then
we
could
discuss
it
further
as
a
group
or
maybe
propose
it
as
a
cap
or
something.
A
A
A
Yeah
totally,
okay,
so
I
guess
that's
the
end
of
the
regular
agenda.
Let's
take
a
look
at
the
open
issues
to
see
if
we
have
anything
that
still
needs
triaging.
D
A
A
A
E
E
So
the
person
who
wrote
most
of
this
code
is
Nick,
so
I
would
I
mean
it
might
make
sense.
I
I,
don't
know
if
Nick's
on
the
call,
but
it
might
make
sense
to
to
at
least
CC
Nick.
B
E
I
would
do
that
I
mean
if
you
could
CC
me
as
well.
I'm
hoping
Nick
will
take
it,
but
otherwise
I
I
did
the
reviews
for
this.
So
I'm,
probably
the
the
I
make
sense
for
me
to
be
the
number
two
on
it.
E
E
And
thank
you
Bridget
sorry,
I'm
late
hi.
Everyone.