►
From YouTube: Kubernetes SIG Node 20220705
Description
SIG Node weekly meeting. Agenda and notes: https://docs.google.com/document/d/1Ne57gvidMEWXR70OxxnRkYquAoMpt56o75oZtg-OeBg/edit#heading=h.adoto8roitwq
GMT20220705-170352_Recording_640x360
A
Is
being
recorded
all
right?
Well
welcome
everyone
to
the
July,
5th
May
22
signal
meeting.
We
have
a
few
items
on
today's
agenda,
but
lighter
attendance
than
usual,
probably
given
the
U.S
holiday
weekend.
A
The
the
first
item
on
the
agenda
I
had
raised,
and
this
is
just
a
call
for
feedback
from
those
who
want
to
give
suggestions
and
I
want
to
be
able
to
close
on
this
this
week,
which
is,
we've
been
trying
to
normalize
our
criteria
for
reviewer
and
approver
Status
within
the
Sig,
as
we
evaluate
past
present
and
potentially
future
requests,
and
so,
if
you
have
a
keen
interest,
which
I
hope
many
of
you
do
on
increasing
your
scope
in
the
Sig.
A
A
So,
if
folks
could
give
their
feedback
by
end
of
this
week,
I'd
like
to
get
updates
and
then
merge
so
we're
in
a
good
spot
for
the
rest
of
the
year,
any
questions
or
thoughts,
not
from
folks
who
may
have
had
a
chance
to
look
at.
Otherwise.
We
can
move
on
to
the
next
topic
and.
B
Sure
so
so
the
next
topic
is
in
place.
Pod
vertical
scaling
by
ubene
is
winning
on
the
call.
No,
he
he
doesn't
seem
to
be,
but
I
I
left
a
comment
on
the
changes
that
may
be
needed
on
the
runtime
side.
So
I
took
a
look
at
transiv
along
with
Kerr,
and
we
left
some
guidance
basically
I.
Think
on
the
system
D
side
we
don't
have
control
over
the
order
in
which
memory
and
CPU
are
written.
While
for
the
other
drivers
we
do
so.
B
A
Okay,
interesting
all
right.
Well,
that
makes
sense,
given
that
we
can't
yeah.
B
B
Yeah
cool
right,
so
the
next
one
is
Adrian,
so
Adrian
I
know
you
had
some
follow
on
response
to
the
security
question.
C
Yes,
yeah
Derek
last
time
brought
up
the
security
discussion
and
after
a
bit
thinking
about
it,
I
think
I.
Remember
we
discussed
this
that
we
want
to
leave
it
for
Alpha
as
it
is,
and
for
beta
we
want
to
and
find
out,
if
there's
a
way
to
provide
additional
authorization
for
for
cubelet,
API,
endpoints
I
think
this
was
the
the
conclusion
we
came
to
I.
C
Don't
know
if
you
remember
Derek
that
we
talked
about
this
a
long
time
ago,
but
I
think
that
that
was
the
point
that
for
Alpha
we
might
leave
it
as
it
is,
and
if
we
move
it
to
better,
we
need
to
look
and
how
to
find
a
way,
if
necessary,
to
add
additional
authorization
to
it.
A
A
For
the
the
history
on
that
I
had
forgotten
where
we
were
so
I
think
it's
fine
to
defer
The
Artsy
stuff
until
data,
so
I
I
will
take
a
look
at
the
pr
as
well,
but
also
trust,
Bruno
and
others
are,
are
fine
to
merge
if
they're
fine
with
it.
So
thanks
Adrian
for
the
follow-up.
C
B
Yeah
I
think
it
makes
sense
to
me
for
now.
I'll
make
another
pass
as
well.
C
B
Great
next
up,
Swati
follow-up
from
last
week's
node
resource
topology,
API.
D
Hi
yeah
I
just
wanted
to
follow
up
on
this,
so
I
had
had
a
chance
to
talk
to
the
kubernetes
release,
engineering
team
and
it
was
actually
earlier
today
and
their
feedback
and
input
was
that
we
can
essentially
follow
the
normal
GitHub
workflow
and
gather
review
from
signal
itself.
D
I
mentioned
to
them
this
that
this
particular
repo
would
be
under
Sig
node,
so
just
want
to
kind
of
let
you
know
that
my
action
item
is
kind
of
dealt
with
and
if,
if
there's
any
input,
please
let
me
know,
and
the
PRS
are
open
for
review.
If,
of
course,
it's
okay
to
proceed
in
this
direction,.
B
B
E
Yeah,
so
this
is
a
quick
follow-up
on
one
of
the
final
comments
on
the
cap
that
quite
a
few
of
you
all
reviewed.
So
this
is
around.
The
Pod
has
Network
condition,
and
that
cap
was
approved
in
the
final
week
by
by
Don,
but
she
had
one
comment
about
like
potentially
looking
into
like
changing
the
meaning
or
not
the
meaning,
but
where
the
initialized
condition
is
applied
for
pods.
That
I
have
no
init
containers
today.
E
So
that's
done
like
super
early
like
before
anything
in
the
pot,
sandbox
and
stuff
gets
created
right.
So
that
seems
a
little
like
weird,
and
we
we
did
describe
this
in
the
in
the
original
camp
for
pot
has
network
so
to
address.
Dawn's
com
and
I
was
wondering
like
I
was
thinking
of
basically
starting
up
a
new
cap,
because
that
seems
like
a
different
problem
than
the
original
one,
that's
being
solved
by
bottas
Network.
E
Would
there
be
any
major
concerns
around
this
change
in
behavior
of
this
existing
condition,
and
you
know,
should
I
start
off
a
new
cap
to
kind
of
discuss
that
separately
in
one
or
25.
While
we
make
progress
on
the
model
Network
or
would
there
be
any
other
ways
of
approaching
it
of
that
comment.
A
Hey
since
I
looked
at
this
earlier
and
I
was
just
catching
up
on
Don's
comment
here,
I
think
the
way
it's
written
like
we
can't
necessarily
assume
any
specific.
Like
the
way
the
connection
is
defined
right
now.
It
just
means
I
have
an
IP
right.
A
So
if
we
change
the
ordering
at
which
the
IP
appears
or
The
Ordering
of
the
conditions,
toggle
I,
don't
think
it
actually
changes
the
meaning
of
the
the
cap
right
so
in
in
that
Spirit,
like
I,
wouldn't
feel
like
you
should
be
pressured
to
do
anything
additional
than
what
you've
already
been
graciously
signing
up
to
do
with
this
cap,
and
so,
if
others
think
there's
value
on
applying
ordering
to
the
conditions
that
would
be
cool
but
I,
don't
I
don't
feel
like
we
lose
or
gain
anything
without
the
ordering.
B
F
I
am
hi
thanks,
hi
everyone,
my
name
is
Natalie
vodko
I'm,
one
of
the
co-chairs
for
Sig,
docs
and
I
thought
I'd
come
over
and
add
something
to
the
agenda
here.
Regarding
reviewers
for
docs
reviews,
We've
recently
had
a
few
reviews
needed
for
signode
that
have
kind
of
been
open
for
quite
some
time
and
I
wanted
to
come
and
discuss
with
folks.
F
What
we
thought
were
reasonable
kind
of
timelines
to
hold
reviews
too,
before
we
can
kind
of
push
contributions
to
maybe
becoming
closed
or
reopened,
or
so
on.
The
psycho
that
I've,
just
linked
to
I,
can
see.
Danielle
specifically
responded
with
the
certain
Alias
that
we've
been
using
for
reviewers
is
actually
likely
not
active.
So.
F
Of
come
here
and
bring
some
kind
of,
let's
say,
shot
a
light
on
us
needing
to
get
some
kind
of
reviewer
help
from
Steve
node
with
a
few
PRS
and
how
we
can
kind
of
move
forward
with
the
you
know
possible,
realistic
timelines
for
reviews
in
future.
A
Hey
Natalie,
so
I
think
you
saw
that
this
is
a
good
week
to
raise
this
topic,
given
that
we're
trying
to
close
on
growth
in
the
other
areas,
I
think
probably
your
topic
calls
out
a
gap
in
my
existing
doc
that
I
posted,
which
was
the
importance
of
the
project
website
review
as
a
part
of
review
or
approval.
A
So
I'll
look
to
make
updates
on
that
this
week,
but
it
looks
like
from
the
pr
there's
eight
things
so
I'm
assuming
a
set
of
us
can
probably
get
through
this
in
the
next
week
or
so
and
then
probably
when
we
have
a
bigger
audience
next
week,
we
can
try
to
get
an
updated.
Alias
I
would
think
anyone
who's
a
an
existing
TL
and
the
Sig
or
in
the
Ed
project
should
be
in
an
alias
that
can
review
any
of
these
topics.
F
Definitely
just
we're
definitely
just
looking
for
having
if
the
Alias
is
something
that
we
want
to
kind
of
keep
going
with
great,
just
some
kind
of
alignment
on
how
do
we?
How
do
we
contact
folks
in
the
sea
to
get
a
tech
review
on
certain
docs
that
are
kind
of
being
updated
and
changed
and
added
to
and
and
what
does
what
to
to
your
think
in
particular,
would
be
a
reasonable
time
frame
to
expect
reviews
just
so
that
we
as
context
for
everyone
in
the
room.
F
Here
we
and
C
docs
get
a
lot
of
PRS
coming
in
and
we
try
and
do
what
we
call
PR
wrangling
each
week
where
we
have
someone
kind
of
on
shift
that
is
pushing
contributions
kind
of
through
through
the
pipeline,
not
only
in
the
English
language,
but
also
in
different
localizations
and
and
we
try
and
keep
our
PR
numbers
that
around
100
to
150
and
that
can
kind
of
be
tough.
F
So
so
this
is
part
of
the
reason
of
of
us
trying
to
kind
of
always
push
contributions
through
and
and
and
we
just
want
to
make
sure
we
can
we're
working
with
sticks
to
what
their
expectations
are
and
what
they're
capable
of
too
yeah.
A
So
one
thing
Sergey
who
isn't
here
today,
but
he
maybe
a
few
others
have
done
this
is
traditionally
when
Sergey
is
here.
A
We
have
a
report
on
like
where
we
are
with
respect
to
incoming
PRS
and
outgoing
PRS,
but
we
haven't
been
covering
the
website
project
repo,
so
I
feel
like
Sergey,
could
probably
update
us
report
out
with
that,
and
so
we
could
get
some
visibility
on
the
same
topic
that
you're
talking
about
wrangling,
so
yeah
I'm
sure
we'll
just
catch
up
with
Sergey
and
the
broader
group
when
they're
around
next
week
for
timeliness,
though
what
is
the,
what
are
you
seeing
with
other
cigs?
F
Not
the
bad
thing,
it's
Ebbs
and
flows
right.
It
also
depends
on
the
kinds
of
just
availability
and
also
number
of
reviewers.
We
ourselves
are
also
trying
to
actually
up
our
review
account
much
the
same
as
you
are,
which
is
also
really
great
like
to
see
the
doc
as
well
and
they've
gotten
some
inspiration
there
I
would
say
like
up
to
a
month
or
two
like,
let's
say
a
couple
of
months:
it's
not
usually
it's
not
that
unusual.
F
Unfortunately,
in
your
case
with
the
list
that
we've
given
we've
got
some
stuff,
that's
like
six
months
and
Beyond
in
terms
of
in
terms
of
waiting
for
review,
which
is
kind
of
which
is
kind
of
why
we're
trying
to
mutually
help
each
other
out.
Basically,.
F
But
it
really
is
Ebbs
and
flows.
We
don't
mind
waiting
if,
if
things
need
the
time
as
well,
but
I
also
thought
coming
to
the
meeting
and
chatting
to
folks
rather
than
us
just
constantly,
maybe
posting
in
the
channel
might
be
a
nicer
way
to
approach
stuff
and
it's
I'm
also
really
willing
to
work
on.
F
If
you
wanted
to
add
to
your
existing
document,
as
you
mentioned,
like
this
idea
of
also
reviewing
you
know,
project
docs
changes
to
the
case,
documentation,
I'm
happy
to
help
out
on
what
Sig,
docs
kind
of
does
in
terms
of
our
time
frames
or
our
general
kind
of
expectations,
so
that
you've
got
an
idea
feel
free
to
tag
me
in
the
in
the
dock.
If
you,
if
you
want
me
to
review
anything
yeah.
A
So
if
you
scroll
down
this
document,
look
on
June
14th,
that's
probably
the
last
day
we
run
a
recent
report.
Yes,
that,
typically
we
advertise
is
how
are
we
doing
with
incoming
and
outgoing
and
that
type
of
thing
and
I?
Just
don't
think
the
query
covers
the
repo
that
your
identity,
fine
and
so
right,
we'll
try
to
break
it
out,
but
I
think
that's
good
feedback
and
we'll
try
to
do
better
here.
So
thanks
for
calling
attention
to
us,
hey.
F
No
problem
we're
always
just
trying
to
do
better,
so
I
really
appreciate
it,
and,
and
thanks
for
letting
me
come
and
speak
to
you
today.
Official.
B
E
Are
we
planning
for
the
internal
freeze
next
week
the
soft
freeze.
B
B
I
guess
we
are
done
so
thank
thanks
for
joining.
Everyone
see
you
all
next
week,
thanks.