►
From YouTube: Kubernetes SIG Node 20220315
Description
Meeting Agenda:
https://docs.google.com/document/d/1j3vrG6BgE0hUDs2e-1ZUegKN4W4Adb1B6oJ6j-4kyPU
A
B
Welcome
everybody
to
today's
edition
of
sig
node.
It
is
tuesday
march
15th,
2022.,
we've
just
passed
our
soft
freeze,
the
previous
meeting,
and
I
wasn't
there,
but
thanks
folks
for
going
over
the
list
of
things
on
today's
agenda.
I
guess,
let's
start
with
sergey
and
the
status
of
prx.
C
Yeah,
if
you
missed,
what's
going
on
with
pr
this
week,
you
can
check
those
links
in
the
agenda
document.
C
I
found
one
rotten
prs
that
added
aging
some
unit
tests
for
like
emptiness
for
static
ports
from
clayton,
I'm
not
sure
if
it's
still
needed,
because
there
are
so
many
changes
in
the
space,
but
this
is
potentially
one
that
was
rotten
without
us
noticing.
That
may
be
interesting
other
than
that
we
grew
a
little
bit
on
pr
account
not
significantly
so,
let's
double
down
on
reviewing
and
approving
prs.
B
Sounds
good,
what's
our
total
that
we
have
open
right
now?
Do
you
know.
C
Oh
sorry,
I
didn't
basically
it's
a
174.
cool,
so
not
not
too
much
change,
yeah,
it's
plus
three,
comparing
plastic,
yeah
and
yeah.
Just
a
side
note
I
I'm
not
sure
wasn't
anybody
noticed
we
added
kevin
as
approver
for
and
tests
welcome,.
B
B
Hurrah:
okay:
let's
move
on
into
the
agenda!
First
up
we
have
vinay
with
in-place
pod
vertical
scaling
status,
yeah.
D
Hi
sorry,
my
own
release
so
we're
working
on
already
at
work
and
that
got
done
last
friday
that
took
a
lot
of
priority.
There
were
many
last
minute
issues,
so
we've
heard
that
story
before
right.
D
Okay,
so
I'm
not
fully
focused
on
trying
to
get
this
address
directs
comments
and
I
was
reviewing
most
of
them
direct
correct
me
if
I'm
wrong,
but
are
there
any
major
concerns
that
you
have
that
I
missed
it
looks
like
there
are
various
questions
and
I'm
responding
to
them
with
examples
on
why
they
were
done.
D
D
Okay,
let
me
ping
him
on
slack,
so
I'll.
Okay,
what
I'll
do
is
I'll
respond
to
his
comments
that
he
has.
I
think
many
of
them
are
clarifications
and
there
are
some
gaps
in
the
missing
tests
and
stuff,
so
I'm
gonna
be
working
on
that
this
week
and
try
to
get
close
the
loop
on
that.
How
close
I
mean
how
comfortable
is?
Are
we
with
this?
I
know
we
missed
the
soft
freeze
deadline.
D
B
D
Okay,
so
this
can
this
still
has
life
till
29th
of
this
month.
The
code
freeze
date
we
can,
we
can
get
it
through.
Okay,
so.
D
I
need
to
update
the
pr
catch
up
with
the
latest
changes.
I
hope
there
isn't
a
lot
of
churn
from
the
merge.
I'm
gonna
spend
time
on
that
mainly
address
direct's
comments.
See
if
there
is
anything
glaring
that
I
missed
gonna
create
issues.
There
are
a
lot
of
people
who
are
willing
to
help,
so
I'm
gonna
start
assigning
issues
to
folks
and
I'll
take
on
some
myself
over
the
next
month
or
so
try
to
wrap
up
as
many
things
as
we
can.
So
that's
pretty
much
the
nutshell
status.
Any
questions.
B
That's
me,
okay,
and
as
vinay
mentioned,
I
just
also
added
a
note
of
this
in
the
agenda
march.
29Th
is
cooked
freeze.
So
if
you
have
anything,
you
want
to
get
done
for
the
124
release.
You
have
a
little
under
two
weeks:
okay,
moving
on
danielle
what
you
got
annual
report,
things.
E
Oh
yeah,
so
as
part
of
doing
the
sig
annual
report
there's
a
couple
of
things
that
I
can't
do
or
would
like
some
broader
input
on.
E
E
If
there's
anything
that
is
different
to
you
know
the
like
kubernetes
default,
contributor,
documentation
that
exists,
and
we
had
the
outstanding
thing
of
don
and
derek
having
a
document
somewhere
a
few
months
ago,
and
so
it
would
be
nice
to
get
that
cleaned
up
and
moved
into
the
community
repo.
As
part
of
this.
A
I
derek
is
terrific
here.
I
think
the
director
needs
the
last
review
about
that
one
and
I
a
priority
approval
couple.
People
already
agree,
but
I
think
the
darker
maybe
hold
us
up
like
the
last
opinions,
and
what's
that
I
think
I
can
direct
derek
is
not
here.
I
can
direct
and
and
let's
move
forward,
yeah.
A
E
Cool
and
then
there's
a
second
part
to
that
of
updating
our
contributing.md
to
be
sort
of
friendlier
to
people
who
want
to
start
contributing
to
node.
E
If
you
take
a
look
at
that
page
today,
it
is
a
pile
of
links
followed
by
like
a
little
bit
of
documentation
from
dims
and
before
I
take
a
pass
at
that,
I
wanted
to
know.
If
anyone
had
opinions
or
things,
they
would
like
to
make
sure
included.
E
C
Yeah,
I
think
what
may
be
really
helpful
is
some
notes
on
cod
organization.
I
think
many
slack
questions
of
new
contributors
is
how
code
is
recognized
and
why
it's
happening.
C
We
may
have
some
links
to
seek
architecture,
but
I
think
for
kubernetes
itself,
like
which
folders
we
use
and
why
this
might
be
useful
cool.
E
That
is
everything
from
me
cool,
oh
one,
last
quick
thing:
actually
is
there
anything
that
happened
last
year
that
we
should
that
we
want
to
share
with
the
community
to
highlight
from
the
sig
I
added
like
getting
test
passing
and
finally
deprecating
docker
shim,
but
like
there's.
B
Probably
something
else
that
I
mentioned,
we
ran
a
bug,
squashing
party
and
it
was
pretty
successful.
A
Can
we
have
a
lot
of
either
promote
the
effort
better
better
to
ga
and
also
remove
depicted
many
features?
I
think
the
maybe
circuit
can
help
because
they
actually
contribute
a
lot
remove.
The
data
code
include
dynamic,
the
kubernetes
config
right,
so
we
can.
We
can
loop
of
those
things.
This
is
basically
make
our
code
bits
maintainable
right,
it's
similar
as
the
docker
shield
depletion.
We
also
actually
remove
a
lot
of
the
deplete
a
lot
of
the
code
and
remove
some
other
code
too.
So,
and
those
could
be
also
highlighted
there.
B
Okay,
anything
else,
danielle.
B
F
Hello,
hey,
so
this
is
a
quick
update
on
the
cap.
We
have
been
working
on
around
pod
sandbox
conditions,
so
we
had
a
few
review
and
comments
from
lana
and
derek
earlier.
So
this
is
a
quick
update
on
the
fact
that
I
simplified
the
overall
cap.
Quite
a
bit,
there
were,
like
you
know,
say
two
conditions
that
you
pointed
at
lana
simplified
that
to
one
it
also
removes
some
of
the
at
least
like
one
of
the
things
that
direct
pointed
out
and
also
handles
termination,
pretty
pretty
nicely
so
yeah.
F
It's
it's
been
simplified
quite
a
bit,
and
I
think
it
aligns
a
little
more
with
some
of
the
existing
ready
conditions.
So
would
be
great
if
I
can
get
some
more
feedback
on
it.
B
Cool,
I
probably
won't
have
time
to
go
through
it.
I
don't
know
if
derek
will,
but
I
know
that
he
also
has
a
really
long
list,
so
I
will
do
my
best
if
other
folks
from
the
community
want
to
look
through
this
and
maybe
give
feedback
on
like
how
this
may
or
may
not
help
you.
That
would
also
be
super
useful.
B
Great
thanks
anything
else
on
that.
B
Cool
any
other
business,
that's
the
end
of
our
agenda
for
today,
short
meeting.
A
May
I
ask
a
little
bit
status
update
on
that
auto
cpu,
because
I
think
that's
the
regression
in
the
1.22,
so
all
our
currently
needs
we're
seeing
all
the
release
will
have
that.
I
know
the
clayton
is
not
here,
but
I
saw
the
david
porter
is
here.
Things
keep
a
lot
of
time
yeah,
so
I
can.
G
Provide
a
status
update
for
that
yeah,
so
the
pr
still
in
progress.
I
think
it's
a
lot
closer
to
getting
done.
Basically,
we
identified
an
additional
issue
a
little
bit
earlier
last
week,
related
to
that.
So
basically,
there
was
an
issue
where
evicted
pods.
G
G
So
it
was
possible
that
a
pod
could
be
in
this
kind
of
weird
state
where
ready
would
be
true,
but
the
phase
would
be
terminal-
and
this
was
also
an
issue
because
from
the
service
standpoint
like
traffic
would
still
be
sent
to
this
pod
in
in
this
terminal
state.
So
we
identified
that
issue
as
well
and
it's
kind
of
related
to
this
pr.
G
So
I
added
another
regression
test
to
capture
that
issue,
so
we
kind
of
have
two
regression
tests
added
now,
one
for
the
out
of
pods
issue
and
second
for
this
issue
as
well,
and
so
clayton
actually
updated
the
pr
and
has
included
both
of
those
tests
and
both
are
now
passing
in
in
the
pr.
So
I
think
we're
actually
very
very
close
to
getting
that
in
and
it
should
solve
both
of
those
issues.
G
So
yeah
clinton's
been
doing
a
great
great
work
on
fixing
it
actually
and
I've
been
testing
it,
and
I
think
it
addresses
both
those
issues.
So
I
think
we
should
have
it
very
soon
and-
and
I
think
the
ferris
state
yeah
yeah-
the
next
step
will
be
it's
kind
of
a
larger
change,
so
we
should
get
it
in
and
we
should
have
ci
passing,
oh
yeah,
the
other
kind
of
thing
to
mention.
I
added
like
as
part
of
the
testing
for
this
pr.
G
It
requires
kind
of
creating
a
lot
of
pods
in
parallel,
and
I
noticed
in
kubernetes
that
in
general
we
don't
have
a
pre-submit
serial
job
so
like
we
were
unable
to
actually
test
this.
In
pre-submit,
because
we
don't
have
a
pre-submit
serial
job,
so
I
went
ahead
and
we.
G
Yeah,
so
I
wasn't
able
to
find
such
a
job,
so
I
I
I
created
a
job
in
test
infrared
that
does
that
absolutely
gonna
to
test
this.
So
that's
the
other
thing
that
I
had
to
add,
and
I
think
that
will
help
test
and
test
this
as
well
and
pre-submit
so
anyways,
that's
the
other
thing
and
yeah
once
we
once
we
get
the
change
in
we'll
have
to
cherry
pick
it.
G
I
guess
all
the
way
back
to
122
it's
a
little
bigger
change,
so
we'll
have
to
go
through
that
and
make
sure
there's
no
conflicts
or
anything
like
that.
But
yeah,
let's
go
to
the
status
update
for
it.
B
Going
once
going
twice:
okay,
let's
call
it
nice
short
meeting.
You
get
40
minutes
back!
Here's
everybody!
Nice
to
see!
You
see
you
next
week,
thanks
bye,.