►
From YouTube: k8s 1.16 - Week 10 - Release Team Meeting 20190903
Description
Release details: http://bit.ly/k8s116
A
A
If
you're,
just
tuning
in
here
feel
free
to
go
and
add
your
name
to
the
attendees
list
on
the
agenda
for
today,
as
we're
in
burn
down,
we
have
a
packed
agenda
and
we
have
postcode
freeze.
So
there's
plenty
of
things
for
us
to
work
on
today.
So
I
will
remind
everybody
that
this
is
being
recording,
recorded
and
will
be
posted
publicly
on
YouTube
and
all
conversations
are
subject
to
the
kubernetes
code
of
conduct
which
governs
us
please
be
great
to
each
other
and
treat
everybody
with
respect
in
your
conversations.
B
Everyone
hope
you
all
can
hear
me
so
right
now
enhancements
are
pretty
good.
We're
green
to
go.
I
put
all
the
notes
there
what's
inside
the
doc
earlier
this
morning
there
was
an
exception
that
was
approved
last
week
for
in
in
tests
and
at
this
point,
everything
that
has
been
tracked
or
it
was
being
tracked,
all
the
PRS
for
those
have
been
merged.
So
everything's,
looking
good
from
an
enhancement
standpoint.
A
A
C
Score
first
off
happy:
what
is
it
to
z,
everyone,
I'm
gonna,
give
it
a
complete
the
complete
opposite
story
or
a
of
enhancement,
everything's
broken
and
see
a
signal.
So
let
me
go
down
the
list
of
issues
away
that
we
have
so
actually
before
that
they
see
a
signal
report
it
just
it
just
went
out
a
couple
hours
ago,
I
believe
so
for
a
full
story
of
everything.
That's
been
going
on
a
in
the
in
in
the
last
couple
days
in
last
week.
A
please
look,
please
took
a
take
a
look
at
the
report.
C
A
other
notable
events
that
I
have
to
mention
is
we
recently
just
added
a
new
job
for
what
a
for
the
116
inform
a
for
the
116
in
forming
a
that's
great
dashboard.
The
job
the
job
is
a
is
a
sake
Windows
on
a
so
surgery,
so
gently
just
for
future
for
future
Norris
we're
gonna
be
working
with
sick
windows,
a
to
copy
made
to
make
a
copy
of
that
job
into
into
master,
informing
it
the
job
it
must
be
informing.
C
We
will
make
sure
that
it
has
the
proper
config
for
connotation,
so
they
so
we
can
prepare.
We
can
prevent
all
the
manual
labor
a
for
future
releases
other
than
that
for
animal.
A
for
anyone
interested.
We
are
going
to
have
a
release
team
six
scalability
a
catch-up.
This
is
going
to
be
a
follow-up
on
the
conversation
we
had
a
at
the
beginning
of
a
at
the
beginning
of
the
release
so
we'll
keep.
It
will
keep
y'all
up
to
date
with
a
whatever
comes
from
that
and
I'm
gonna
start
really
I'm.
Gonna.
Stop
really
quick.
C
Sweet
so
now
on
to
all
day,
and
although
all
the
things
are
working,
let's
start
with
master
a
master
walking.
So
we
a
we
currently
have
two
jobs
are
failing
so,
and
you
can
check
you
can
check
out
the
issues
and
in
the
PRS
for
the
so
so
for
one
of
them
in
DC
cause
Buster
cereal.
We
have
a
bunch
of
DNS
related
DNS.
Related
tests
are
failing.
C
We
actually
got
a
got
a
potential
fix
before
we
are
before
we
had
to
open
a
tray
before
we
had
to
open
a
track
in
PR,
so
they
so
the
first
one.
The
first
one
is
a
reference
to
the
PR
daddy
that
is
aiming
to.
That
is
aiming
to
fix
the
test
failures
in
that
job.
Please,
please
check
it
out.
That
did
make
sure
to
include
the
PR
in
the
Co
here
in
the
career,
milestone
to
me
to
make
sure
that
once
it
gets
reviewed
and
approved,
he
it
actually.
C
It
actually
goes
to
the
tide
pool.
Then
next
182
174,
this
one
is
a
failure
on
the
a
GC
Alpha
features
test
or
master
master.
Blocking
we've
been
investigating
this
way.
This
one
for
a
couple
days:
they
it
the
way
they
want
the
one
thing.
The
one
thing
that's
happening
is
they
actual
copper
Nate
a
person
did
or
something
misconfigure
in
the
in
the
copper
net
is
cluster.
C
That
is
prevented
that
is
actually
preventing
any
and
two-inch
end-to-end
tests
from
actually
a
from
actually
running
where
we
been
working
on
a
on
tracking
this
one
down
and
right
now,
the
possible
culprit
seems
to
be
a
that.
It
seems
to
be
a
PR
them
implemented,
implemented
a
new
feature,
adding
endpoint
slice
to
proxy
IP
vs
and
IP
temple
proxy,
a
proxy
errs.
C
This
is
right
now
we're
really
just
guessing
that
this
might
be
responsible
with
a
we
don't
know
if
this
is
actually
this
actually
the
root
cause,
but
we're
working
with
the
authors
of
the
APR
to
investigate
further,
also
a
not
really
enough
related
to
CI.
You
should
really
do
with
the
cap
for,
for
this
feature,
really
cool
a
other
than
that
we
have
3
flex,
almost
a
must-have
working
in
11
passing
now
moving
on
to
master
conforming
master,
informing
we
also
have
to
just
failing.
C
One
of
them
is
scalability
acid
scalability
on
it
do
180
to
the
leash
deletion
error
on
CI
kubernetes
into
NGC
scale
performance,
a
the
alias
for
it.
I
guess
just
scale
performance
right
now
we
are
awaiting
a
response
from
six
scalability,
although
they
did
mention
that
they
were
aware
of
the
error
and
hopefully
hopefully
they
and
they
will
report,
they
will
report
back
with
the
full
story
on
what's
what's
going
on.
The
other
182
186
is
a
failure
on
DC
scale
correctness,
and
this
one
seems
to
be
actually
really
this
way.
C
This
one
is
also
a
six
scalability
on
job,
but
the
failure
put
the
finger
seems
to
be
in
from
the
two
a
stick
network
on
end-to-end
test:
we're
also
waiting
their
response.
For
that
one
way
we
don't
know
the
whole
story,
we
don't
actually
a.
We
don't
yet
know
exactly
what's
a
what's
wrong
and
other
than
that,
we
once
a
once
tell
job,
which
is
a
which
is
a
conformance.
We
tour
on
OpenStack
all
day
provider.
Openstack
people
are
poor.
They
are
diligently
working
on
it
a
most
recently.
C
C
What
they're
doing
today
during
the
last
couple
days
a
so
for
this
one
will
observe
if
some
a,
if
you
felt
again,
then
we'll
make
sure
to
contact
a
probably
releasing
it
a
release
engineering
into
it,
are
looking
at
this
one
and
that
is
it
from
the
master,
a
from
the
master
release
set
of
things
now,
a
1/16
blocking.
It
actually
has
the
same
failures
as
most
have
walking.
So
hopefully
so,
hopefully,
once
we
get
those
fixes
in
and
116
blocking
is
gonna
be
healthy
and
116
informing
aim.
C
So
this
is
a
this
is
actually
the
job
that
was
Adam
is
sick,
a
basic
Windows,
aka,
IKS
engine,
a
sure,
116
windows.
It
turns
out
the
job
the
job
had
has
been
failing,
since
it
was
adding
do
to
say
what
is
it
an
out
of
scaling
and
went
into
and
test
failing,
and
that
is
it
for
today's
update
on
CI
signal.
Anyone
has
any
questions.
Comments,
concerns.
A
If
nobody
else
I
have
a
few
questions
comments
and
concerns
mm-hmm
I'm
just
walking
that
list,
so
the
two
that
are
failing
in
master
blocking
the
DNS
tests,
I,
there's
a
PR
open
to
get
them
fixed,
there's
just
a
missing
readiness
check,
so
that
should
clean
that
up
and
soon
deep
is
on
that.
So
I
think
we're
okay,
I'm,
just
trying
to
figure
out
which
ones
we
need
action.
The
pods
are
not
ready
in
alpha
features.
A
Rob
from
Google
who
was
working
on
the
endpoint
slice
actually
commented
and
said:
I
think
I
have
a
PR
for
this
late
last
night.
So
we
should
see
something
coming.
I
just
want
to
illustrate
that
those
two
ones
and
master
blocking
actually
have
engagement
from
the
owners
at
this
point
and
we're
hoping
to
get
PR
zine
within
the
next
24
48
hours
here
to
get
them
cleaned
up,
which
I
think
is
good
and
the
scaleability
ones
in
master
informing
the
8
to
180.
It
looks
pretty
flaky
right,
we've
got
three
green
runs.
C
When
it
comes
with
six
with
scalability
test,
especially
especially
the
one
in
80
to
180
and
the
job
takes
an
entire,
the
job
takes
an
entire
day
to
run
so
it's
okay,
not
least
in
the
near
past
flakiness.
He
actually
means
that
something
might
be
that
something
might
be
breaking.
So
we
really
need
a.
We
really
need
someone
from
scalability
to
tell
it
to
tell
us
a
way
where
to
go,
what
to
look
for
or
if
it
really
is
just
a
random
link.
Okay,.
C
A
Okay,
cool
and
we've
had
no
action
on
eight
to
one.
Eight
six,
which
is
the
source
from
it,
looks
like
the
last
comment
was
two
hours
ago
and
they've
called
out
specifically
at
mister
hone
on
the
looks
like
what's
this
PR
I
blowed
answer
finalized
at
a
betta
which
may
may
be
the
thing
that's
happening
here,
it'll
be
interesting
to
see.
If
we
get
in
I
can
I
can
ping
signet
work
to
see
if
we
can
get
somebody
else
to
take
a
look,
but
it
looks
like
at
least
we've
called
out
somebody
there.
A
A
I,
don't
like
a
board
getting
added,
and
it's
been
red
since
it's
been
added
okay,
so
we'll
get
that
cleaned
up
absolutely
alright
and
we
have
a
conversation
with
cig
scalability,
but
probably
the
only
one
that
seems
actionable
right
now.
Is
the
source
ranges
networking
conformance
test
is
in
flight,
the
flaky
one
there
anything
else.
We
need
to
try
and
look
at
I'm,
just
wondering
who
I
could
should
chase
down
today,
but
looking
at
most
of
those
things,
they
seem
to
be
in
hand
or
somebody's
responding.
Mm-Hmm.
C
A
F
Everyone,
so
not
so
good,
rent
this
time
via
read.
So
regarding
issues
we
are
static:
okay,
twenty-eight,
in
other
words
eight
less
than
we
had
to
devise
meeting,
and
we
don't
have
any
critical
urgent
issues
and
regarding
pr's
we
have
24,
and
this
is
70
less
than
we
had
last
week,
which
is
okay,
but
the
problem
is
that
we
have
got
one
critical,
mojit
issue
and
I
think
it
has
been
mentioning
it,
but
by
George
in
this
guy
thing
about
date,
it
had
been
created
today
and
is
still
missing
review.
F
So
I
hope
it
will
get
soon.
But
it
is
the
reason
why
we
are
at
time,
because
these
people
that
appeared
and
we
got
appears
we
have
two
peers
that
have
HTM
and
approve
and
are
the
milestone,
so
they
are
probably
going
to
get
in
and
beside
that
the
queue
is
drained,
so
we
don't
have
any
other
peers.
So
the
end
regarding
all
helping
issues
we
have
42.
So
this
is
regarding
numbers
and
yeah.
So
the
only
thing
I
can
be
is:
should
we
start
removing
other
issues
and
PRS
from
we?
B
A
Okay,
that's
fine!
Thank
you
for
calling
that
out.
I
think
we're
OK
to
start
removing
issues
and
peels
from
the
milestone.
If
they
haven't
been,
if
there's
been
no
activity
and
you've
pinged
them
and
said
so,
if
there's
no
activity
put
them
in
milestone
next
I'm,
not
sure
of
II
actually
do.
We
would
just
want
to
milestone
them
rather
than
keep
them
to
the
next
milestone,
and
have
people
read
milestone
them
when
they
pick
them
up
once.
G
You
know
yes,
we
were
issues,
especially.
We
need
to
make
sure
that
we're
not
losing
signal
of
issues
that
were
considered
important
to
resolve
for
this
milestone.
Prs
are
more
reasonable
to
take
out
of
the
milestone,
especially
if
they
don't
have
LG,
TM
or
approved
for
activity,
but
I
think
we
need
positive
confirmation
that
an
issue
is
not
blocking
before
we
lose
track
of
it
in
the
sea
of
non
milestones,
issues.
A
A
couple
of
weeks
ago,
Jordan-
this
is
just
one
case
in
point-
is
there's
an
issue
from
early
2018
from
Clayton
Coleman
that
something
about
stale
reads
which
is
critical
urgent,
but
it
hasn't
been
worked
on
for
a
year
and
a
half
yeah,
and
it's
so
basically,
it
just
keeps
getting
kicked
from
release
to
release
and
nobody's
actually
working
on
it.
Every
release
team
has
to
go
through
and
decide
whether
it's
important
in
which
point
Clayton
says
yes,
it's
important
and
it's
critical,
but
nobody's
actually
working
on
it.
G
That
issue
in
particular
and
I
would
agree
that
it
is
not
prioritized
correctly
given
reality,
I
would
start
by
asking
on
the
issues
in
the
milestone
for
an
analysis
of
whether
this
is
release
blocking
and
give
that
a
day
or
two,
and
if
there's
no
response,
then
we're
more
free
to
start
kicking
things
out.
If.
G
F
F
G
I,
if
it's
been
moved
from
milestones
before
and
there's
no
new
information,
that
indicates
why
it's
important
for
this
release,
then
that's
reasonable
to
move
things
that
have
recent
activity
where
they're,
actually
getting
drilled
down
and
and
looked
into.
We
want
to
just
double-check
with
the
people
who
have
more
information
on
those
issues
before
we
move
them
out.
I.
H
G
I
could
make
one
request
new
issues
and
PRS
that
we
move
out
if
we
could
just
keep
a
list
of
those
may
be
here
in
minutes
or
as
you
go
through
the
market
somewhere,
so
that,
because
once
we
move
them
out
of
a
milestone,
it
is
almost
impossible
to
find
them
again.
If
someone
said
oh,
no
wait,
there
was
one
that
was
really
important
where,
where
did
it
go?
It
would
be
really
helpful
just
for
visibility
to
have
that
list.
G
I
I
will
break
the
streak
of
bad
news.
Doc's
is
green.
We
have
one
outstanding
PR
that
I
have
reached
out
to
the
team
on
an
update
on
that
today.
Doesn't
look
like
there's
been
too
much
activity
on
it,
but
from
what
they
mentioned.
It
doesn't
look
like
it's
that
big
of
a
deal
so
should
be
quick
and
easy
for
them
to
nail
that
down
and
have
everything
ready
for
review,
and/or
merged.
A
J
Pleasure
so
we're
green.
We
have
to
pee
ours
in
for
usability
improvements
to
the
release,
notes
website,
we're
also
doing
our
weekly
update
of
the
release.
Notes
talked
as
well
as
editing
for
style
and
grammar,
and
the
only
thing
we're
waiting
for
is
just
the
update
to
the
coms
draft
for
major
themes,
and
we
spoke
with
taylor,
and
we
should
hopefully
have
that
by
end
of
day
today
and
that
any
questions
or
concerns.
K
Hello,
everybody,
okay
had
a
wonderful
weekend.
Today
we
are
tracking
at
yellow.
Do
you
have
to
get
those
themes
over
we'll
be
doing
that
by
end
of
day
today,
we've
had
a
little
bit
of
an
issue
in
getting
some
responses,
but
you
know
fully
expected
with
the
holiday
weekend
and
everything
go
freeze
and
everything
else,
but
so
we've
done
a
few
SIG's
to
get
first
draft
of
their
blog
posts.
We
haven't
had
any
submissions
yet,
but
we
did
have
one
get
pulled
out
and
that's
from
cui.
K
That's
dashboard
rewrite
won't
be
making
it
in
that's,
not
complete
and
so
we're
gonna
I
think
we
had
six
total.
So
now,
we've
got
five
to
those
things.
We're
looking
at
responses
from
going
to
be
engaging
with
the
CNC
F
about
the
overall
themes
as
well.
Once
we
have
those
kind
of
quantified
and
then
get
that
out
and
get
an
idea
of
just
the
scheduling
for
all
the
PR
and
everything
that's
the
CNC
F
is
going
to
be
running
other
than
that
I.
K
L
L
L
A
E
A
Thanks
emeritus
Lee
Josh,
moving
on
to
the
release,
lead,
update
current
status,
I'm,
going
with
red
on
the
release
for
two
things:
the
the
outstanding
jobs
in
release,
blocking
that
are
red
and
the
scaleability
jobs
in
the
release,
informing
master,
informing
and
master
blocking
I
want
to
get
those
sorted
out
before
we
go
to
yellow
and
subsequently
green.
Hopefully,
next
upcoming
milestones
for
this
week,
we're
cutting
B
to
two
tomorrow.
The
docks
deadline
was
today.
So
please
have
all
your
PRS
ready
to
review.
A
It
sounds
like
we're
only
waiting
on
one,
which
is
great
Thank,
You,
Seth
Young
is
cutting
B
to
two
tomorrow.
Hopefully,
we
can
clean
up
the
two
jobs
in
master
blocking
today
and
at
least
get
a
clean
signal
from
master
blocking
that
beta
2
has
some
sense
of
stability,
at
least
in
the
things
that
are
failing.
It
is
daily
burned
down
this
week
so
in
next
week
for
the
next
two
weeks
leading
up
to
release,
so
this
meeting
will
be
every
day
at
this
time,
as
I
said
for
the
folks
in
Europe.
A
If
this
time
is
inconvenient
on
Tuesdays
or
Thursdays,
let
me
know-
and
we
can
reschedule
it
99,
which
is
next
Monday-
is
the
dark,
sould
Docs,
p
oz
need
to
be
reviewed
and
merged
by
next
week.
So,
if
you
do
have
Docs
p
oz
out
there,
please
get
work
on
them
this
week
with
the
docs
team
to
get
them
in
shape
ready
to
merge
next
week.
Nine
ten
is
next
Tuesday
I
believe
just
popping
over
and
have
a
look.
Yes,
nine
ten
is
our
c1
and
it's
also
code
Thor.
A
So
we
have
another
another
week
of
code.
Freeze,
there's
another
thing
of
point:
Josh.
Sorry
Stephen
is
using
this
time
tomorrow
for
a
cig
release
meeting.
So
please
show
up
to
that
meeting
because
a
lot
of
what
he
was
be
discussing
regards
the
work.
That's
been
done
by
this
team
and
it's
really
longer-lived
items
that
need
to
get
spun
out
and
have
a
separate
conversation.
I
know
we
parked
a
few
things
from
different
teams
over
time,
so
feel
free
to
show
up
and
add
things
to
the
agenda
tomorrow.
A
So
we
will
reconvene
on
Thursday
at
9:00
a.m.
US
Pacific
for
the
next
bone
down
meeting
a
couple
of
questions.
For
me
as
I
know,
this
has
been
discussed
in
previous
releases
is
one
is:
do
we
want
to
continue
our
seas,
which
I
think
is
an
interesting
exercise,
maybe
for
the
sig
release
team,
but
last
people
question
whether
our
seas
were
useful
or
whether
it
should
be
another
beta.
A
A
A
A
M
I'm
here
so
after
reviewing
the
PR,
it
looks
like
all
I
need
to
do
is
just
submit
the
initial
draft
of
some
verbiage
on
what
the
backporting
fixes
look
like
I
was
planning
on
putting
it
here,
I'll
send
a
link
in
the
chat
just
making
sure
we
all
agree
on
this.
I
was
gonna,
submit
the
PR
over
the
weekend.
M
And
I
think
this
is
gonna,
be
an
iterative
work.
I'll
submit
a
PR
based
on
the
discussion
we
had
in
this
issue
here
and
I.
Think
folks
are
gonna
chime
in
we'll
make
some.
You
know,
revisions
to
it
and
I
submitted
I.
It's
not
so
much
about
the
document
itself
is
just
having
a
place
to
reference
it.
Besides
this
issue
of
why
we
back
poor
things,
you
know
outside
the
supported
releases
and
what
dictates
that.
A
A
I
have
this
one.
It's
okay,
I
will
still
walk
through
this
still
committee
and
get
this
done
before
the
end
of
the
release.
I
have
this
one
as
well
same
state
yep,
getting
the
what
verbage
around
the
different
terms,
shadow
improvement,
July
29th,
it
looks
like
it
is
assigned
to
Josh,
Josh
and
Stephen,
which
I
assume
is
okay.
Does
this
need
to
stay
and
to
do
for
now?
Are
you
actively
going
to
be
working
on
this?
Yet?
No.