►
From YouTube: k8s 1.16 - Week 3 - Release Team Meeting 20190715
Description
Release details: http://bit.ly/k8s116
A
Fantastic
hello
and
welcome
good
day
to
everybody
today
is
Monday
July,
15
2019.
It
is
week
3
of
the
1/16
release,
welcome
all
lovely
to
see
you
all
on
this
Monday
morning,
at
least
for
me,
happy
Monday
to
everybody.
I
hope
you
all
had
a
wonderful
weekend,
I've
popped
the
release,
meeting
link
in
the
chat.
Please
follow
that
through
and
if
you
go
to
week,
3
so
gender,
please
place
your
name
and
your
role
on
there
so
that
we
know
who's
here
and
also
a
housekeeping
wise.
If
there's
a
note
taker,
please
place
your
name.
A
I'd
love
somebody
to
volunteer
for
that.
That
would
be
fantastic
before
we
get
into
the
agenda.
Just
a
friendly
reminder
that
this
meeting
falls
under
the
kubernetes
code
of
contacts,
so
please
be
excellent
to
each
other
in
conversation
and
actions,
and
this
meeting
is
being
recorded
and
will
be
posted
to
YouTube
later,
ok
with
out
for
did
somebody
have
something.
A
A
Ok,
so
we
are
going
to
go
into
regular
schedule
this
week,
we're
following
the
template,
but
I
will
ask
as
we
go
around
if
there
are
shadows
that
have
not
been
introduced.
Can
the
leads
just
give
everybody
each
of
their
shadows,
an
opportunity
to
be
introduced
to
the
team.
That
would
be
fantastic
and
to
kick
us
off
this
morning.
I
will
hand
it
to
Kenny
on
enhancements,
take
away
Kenny
yeah.
B
Thanks
locky,
so
everybody's
inboxes
might
have
gotten
flooded.
Last
week
we
hit
all
hundred
and
50
open
issues
that
are
in
the
K
enhancements
repo,
so
we've
gone
through.
We've
got
to
give
a
big
shout
out
to
all
the
shadows
on
the
team.
Everybody
really
stepped
up.
They
took
about
30
issues,
every
every
shadow
about
30
issues,
they're
responsible
for
and
so
with
that
hitting
them
making
sure
we
get
some
responses
back
from
the
individual
owners
or
the
cig
leaders.
There's
a
few
a
few
few
ones.
That
kind
of
crept
up.
B
I
know
that
lucky
kind
of
tagged
you
on
some
of
those
be
able
to
talk
to
something
individuals
cigs.
When
you
have
some
meetings
and
stuff
like
that,
but
for
the
most
part
we
have
a
really
good
consensus
of
really
where
we
are
after
this
first
week.
So
at
this
current
moment
we
are
tracking
the
51
to
just
be
in
the
tracked
category.
It
just
does
mean
that
you
look
to
do
something
in
116,
whether
you
want
it
to
be
initialized
this
alpha
or
graduated
through
beta
and
stable.
B
However,
with
that
becomes
some
other
risks
that
are
involved,
we'll
talk
about
that
here
in
a
second,
so
51
being
track,
22
is
alpha.
17
beta,
when
I
really
love
to
see,
is
seeing
12
going
to
stable,
considering
last
release
I
think
we
had
like
2
3,
something
like
that.
So
this
gives
us
a
much
better
kind
of
an
idea
of
really
where
the
progression
and
track
of
these
enhancements
are
really
going.
So
we
can
see
that
there's
there's
some
good
movement
that
are
happening
between
some
of
these
stages
for
these
individual
enhancements
as
well.
B
Now,
over
to
the
risk
side,
there
are
only
36
right
now
that
are
slated
with
a
cap.
This
is
for
a
few
things,
one,
it
could
be
a
brand
new
enhancement
and
they
want
to
see
it
come
in
as
alpha.
So
therefore
they
are
going
through
the
process
of
getting
they
kept
merged
into
from
a
provisional
into
an
implementable
state.
B
This
is,
of
course,
we're
putting
a
lot
of
these
15
or
sorry
I
think
there
was
like
15
that
either
just
did
not
have
a
cap,
and
then
there
was
four
extras
that
were
and
sort
or
some
kind
of
a
limbo
state.
If
you
will
with
this,
is
we're
gonna
be
looking
at
keeping
this
at
risk
because
it
just
it
if
it's
not,
there
emerged
already.
We
know
that
we're
gonna
be
hitting
enhancer
freeze
here
in
like
three
weeks,
something
like
that
into
July.
B
Well,
not
even
that
15
days
about
then
so
that
kind
of
gives
you
an
idea
that
if
nothing
is
happening
by
the
end
of
enhancer
freeze,
it's
probably
not
going
to
happen
by
any
time
after
that.
So
we
really
want
to
see
those
caps
get
in
before
enhancement
freezes
over,
so
those
the
ones
were
keeping
at
risk
for
now,
if
by
some
chance
they
do
get
merged,
then
we'll
take
off
and
we'll
put
them
back
into
the
track
category.
A
Thanks
Kenny
yeah
and
just
a
note
on
the
enhancement
phrase,
as
you
mentioned,
it
is
July
30th,
Tuesday,
July
30th,
which
is
15
days
out
from
today,
which,
if
you
look
at
scheduling
it's
an
extra
week
than
we
usually
have
so
we
started
enhancements
tracking
a
week
earlier
so
I'm
hoping
that
we
can
avoid
any
last-minute
kerfuffles
or
extensions.
But
you
know
how
that
goes.
One
of
the
things
I'm
trying
to
do
is
communicate
with
the
sig
leads
as
well
just
to
make
that
date
really
well
known
so
that
they
don't
come
in.
A
B
C
A
D
Actually
I'll
be
taking
over
a
Veronica
was
not
a
was
not
able
to
do
to
some
last-minute
issues
and
before
we
actually
get
started
with
the
thousand
things
that
are
going
wrong
in
CI.
Let
me
introduce
a
couple
people
that
I've
been
doing
a
lot
of
work
with
CI
signal
these
past
couple
of
weeks
and
I
see
them
on
the
call
right
now.
A
Craig,
Craig
Peters,
carlos
pinero
and
Nicolas
Murat
is
populated
in
butcher
that,
if
you're
listening
and
you're
not
three
please
take
the
floor.
Introduce
yourselves
hey
guys.
E
G
D
D
They
are
labeled
us
and
she'll,
be
able
to
add
or
delete
notes
it
technically
be
technically
what
the
tests
actually
do
is
that
they
create
a
replication
controller
and
they
they
wait
for
enough
replicas
hotpot
they
to
be
able
to
be
available
seems
to
be
that
the
test
failure
is
actually
relatively
straightforward.
Same
pots
are
stuck
in
a
pending
state
for
a
couple
minutes.
D
Maybe
they
may
timeout
is
the
heated
time
out
and
then
after
that,
there's
another
message
for
okay,
both
Sandler
a
I'm,
not
running,
and
everything
may
put
you
know
by
a
by
that
time.
The
test
is
already
marked
FAA,
as
failure
so
trying
to
get
trying
to
get
a
hold
of
someone
from
a
plus
or
lifecycle
to
make
sure
it
to
make
sure
that
the
test
failure
is
really
shows
that
there's
nothing
there's
nothing
a
particularly
wrong
in
that
area.
D
Now,
let
me
go.
Let
me
go
with
the
AC
bumpers,
a
release
miscellaneous
a
it's
looking
healthy,
a
nine.
So,
lastly,
last
week,
I
last
week
when
I,
when
I
talked
it,
when
I
last
talked
to
you
all
the
work,
a
polish
there
were
a
couple
jobs
and
were
failing
seems
to
be
that
the
branch
management
team
a
they
are
so
awesome.
They
fix
all
of
them
in
a
single
week.
D
D
We
still
need
to
look
for
an
owner
for
a
for
this
hazard
for
these
failures,
but
it
seems
that
there
was
some
work
from
six
scalability
on
trying
to
get
a
they
trying
to
bootstrap
kubernetes
clusters
that
had
no
access
to
the
Internet,
and
so
the
firewall
rule.
It's
some
complication
that
arises
from
that
still
so
need
mitigate
need
to
get
a
hold
of
someone
from
sick
network.
That
wants
done
really
knows
how
a
the
tests
are
supposed
to
be
running
and
what
a
whatever
requirements
actually
need,
a
whatever.
D
Whatever
things
actually
need
to
tend
to
test
in
the
test,
the
next
one
there's
a
it
worried
that
were
a
couple
issues
that
we
discovered
that
we
discover
a
couple
weeks
ago
with
a
cubicle
test
and
another
poll
in
unemployment
test.
The
underlying
issue
seems
to
be
the
we
were
trying
to
do,
something
that
has
been
has
been
deprecated
in
kubernetes
1:15,
for
example,
trying
to
create
a
deployment
with
an
extensions.
D
We
want
better
one
API
group,
so
we
fix
the
test,
but
now
it
seems
that
the
actual
jobs
and
sells
our
misconfigure
it
not
in
a
couple
of
them.
We
we
actually
been
trying
to
go
from
kubernetes
114
to
116,
where
we
were
supposed
to
a
where
what
we
actually
intended
to
do
was
go
from
115
to
what
16
everyone
16
on
16-bit
day,
the
master
branch
Raina.
D
So
the
next
step
that
we're
going
to
take
is
just
open
up
a
couple
peers
to
try
to
fix
a
try
to
fix
a
jobs,
get
a
conversation
going
to
see
if
that
is
actually
the
the
right
thing
to
do.
If
you
look
on
the
issues,
a
I
will
link
a
Google
Doc,
where
we
actually
call
a
where
we
actually
call
in
detail
about
the
failure
and
all
the
things
that
we
all
these
are
wrong
with
it.
So
you
can,
you
can
go
there
for
full
details.
D
The
next
one,
a
CSI
mock
volume,
test,
failing
need
to
create
a
need
to
create
an
issue
for,
though,
a
need
to
create
an
issue
for
doubt.
So
that
is
so
that
we
can
track
it,
but
seems
relatively
straightforward.
There
is
a
kinder
test.
The
doc
has
it
has
been
failing
so
also
need
to
create
an
issue
for
is
a
fraud
on
one
and
a
really
cool
thing.
D
I
still
have
I
still
don't
know
why
it
actually
happened.
Instead,
during
the
weekend,
there
were
a
bunch
of
issues
with
the
CIE
for
some.
For
some
reason,
many
jobs
were
not
able
to
actually
create
a
kubernetes
cluster,
so
they
they
failed
multiple
times.
Some
other
jobs
just
seemed
to
have
a
random
prior
issues,
as
as
it
seems
them
the
latest.
Then
there
are
multiple
ones:
I
had
not
I
have
not
finished
so
so
they
so
the
plan.
The
plan
of
attack
for
those
ones,
are
close
to
way.
D
Give
more
time
a
give
more
give
a
more
time
for
CIT
run,
see
if
we
actually
just
get
more
signal
on
things
are
actual
failures,
catch
up
with
the
people
as
sick
testing,
testing
knobs
it
in
case
they
actually
know
a
David
April.
They
probably
don't
know
what
happe
what
happen
during
the
weekend.
Just
catch
up,
see
if
there's
anything
there,
the
release,
team
or
just
a
broader
community,
a
missed,
needs
to
do
or
fixed
and
yeah.
That's
everything
that's
coming
running,
see
a
signal
right
now.
Anyone
has
any
questions.
Comments,
concerns
I,
have.
H
I
H
Remove
them
from
the
master
informing
board
and
that's
going
to
be
a
PR
and
a
strongly
worded
email
soon,
not
strongly
worded.
It's
just
going
to
be
an
email
to
note
and
cluster
life
cycle.
It
sounds
like
they
have
consist
gently,
not
had
an
owner
for
multiple
cycles
and
we
often
get
into
the
situation.
So
if
no
one
wants
to
own
them
and
no
one
is
actively
working
on
fixing
them,
then
they
don't
need
to
exist.
D
So
actually
I'm,
gonna
I'm
gonna
take
off
my
CI
signal
head
and
put
my
testing
cluster
like
a
psycho
hat.
There
is
currently
so
there
is
the
reason
why
those
jobs
are
not
maintained
is
because
the
mechanism
for
bootstrapping
the
clusters,
it's
the
cluster
directory
in
the
Gaiety
repo,
which
has
been
deprecated
for
more
than
two
years.
There
is
ongoing
work
by
a
cluster
lifecycle
and
sig
architecture
to
actually
refactor
that
into
something
that
into
something
the
community
wants
to
mind.
D
It's
something
that
the
community
wants
to
maintain
so
and
in
a
min
that
case,
if
everyone
is
okay
with
if
every
day
so
I
guess
my
proposal
will
be,
can
we
actually
either
keep
them
on
mastering
forming
or
very
preferably
moving
somewhere
else
somewhere
else
release
miscellaneous?
We
just
keep
them
in
there
so
that
people
can
can
keep
I
can
keep
an
eye
on
them.
D
H
So
I
think
that
that
is
part
of
that
is
a
good
plan.
I,
don't
want
it
on
our
board,
saying
that
we
I
mean
it's
on
our
board,
so
it
basically
means
that
we
own
it
right.
So
the
so
my
thought
was
either
a
cig
release,
orphaned,
board.
That
says
like
this
is
clearly
stuff
that
people
have
decided
not
to
own
right
or
remove
them
altogether.
All
right
so
I'm
fine
to
do
either,
but
there
needs
to
be
some
sort
of
hey.
We
don't
own
this
hey
someone
do
something
about
this.
Okay,.
J
D
H
H
D
It's
a
player,
so
the
plan
is
that
we
at
least
at
least
get
started
with
some
tangible
work.
They've
in
have
something
set
up
for
a
for
1/16
I'll.
Look
for
the
issue.
I
know
I,
have
it
somewhere
in
one
of
my
thousand
tabs
and
I'll
link
it
in
the
in
the
notes
and
in
the
secret
list
channel
for
anyone.
That's
interested.
D
So
that
right
up
its
I
added
it
to
the
comments
in
the
in
the
actually
in
the
actual
issues.
So
if
you
go
s
or
if
you
go
to,
they
see
a
signal
master
informing
a
on
the
failing
jobs,
there's
a
misconfigure
upgrade
down
with
and
skill
jobs
both
of
those
issues.
A
the
last
comment
contains
a
link
to
the
document.
Ok,
perfect
thanks.
D
D
The
misconfigured
amiss,
so
there
are
a
couple.
There
are
a
couple
areas:
the
firewall
and
the
sig
testing
sig
testing
in
candor
jobs.
Those
are
normal
failures.
You
know
they
they
happen,
they
always
happen.
We
just
interfere
just
need
to
fix
them
the
misconfigured
jobs.
That
is
probably
something
that
has
been
broken
for
a
while.
D
We
just
never
notice
it,
because
this
is
the
first
time
that
we
actually
deprecated
an
API
group,
and
you
know
now
we
actually
need
to
really
be
looking
for
those
a
for
looking
up
for
specific
Persians,
again,
probably
and
the
other
end.
Other
stuff
is
parishes
and
they
happen
a
every
every
every
now
and
then,
but
they
they
also
happen.
A
H
A
But
this
is
the
release
where
it's
actually
like
gone
right,
so
this
is
yeah.
Anybody
move
115
to
116
and
using
extensions
v1
deployments
or
break
yep,
okay
yeah.
We
just
make
sure
we
at
least
then
we
can
say:
we've
done
our
part
on
making
it
very
well
known
kind
of
want
that
at
the
top
of
the
release.
This
is
how
many
times
have
we
deprecated
these
long-standing
api's
and
kubernetes,
not
much
okay,
cool
thanks
for
the
update
judge.
H
I
L
L
Thank
you,
so
I
don't
really
have
news
right
now.
It
is
a
little
bit
during
the
cycle.
So
there
are
some
queries
in
the
document.
So
far
we
are
standing
about
I,
think
66
be
ours,
often
yes,
and
about
issues
it
is
72
and
I
have
a
question
about
Suffolk.
When
should
we
set
up
the
issue?
Npr's
tracking
documented?
These
are
someone
who
can
help
me
that
maybe.
N
O
Hi
everyone,
my
name,
is
Juanita
I
am
from
Hyderabad
India,
so
I
am.
This
were
my
very
first
Amazo
shadow
for
dogs
and
the
another
best
part
is.
My
brother
is
also
a
shadow
for
bug.
Triage
Tarun,
so
we
both
are
shared,
is
working
for
two
different
groups
for
this
release.
That's
something
very
big
for
us.
Thank
you.
N
Thank
you
very
much
for
Donna
for
the
dog.
We
don't
really
have
any
major
updates
right
now:
I'm
working
with
the
shadows
to
understand
the
process,
so
that
there's
going
to
be
very
good
understanding
on
how
you
know
the
whole
thing
works
with
dog
in
case
any
one
of
them
wants
to
be
the
Future
lead
and
also
we
are
working
on
the
process
and
ants
meant,
especially
with
Jane
Jane
angel
is
been
very
helpful
on
working
weeks
how
to
improve
dog.
N
You
know
dog
release
process
and
finally,
I've
been
going
through
the
the
announcement
tracking
sheets
so
to
finalize
how
we
are
going
to
distribute
talks
among
ourselves
and
shadows
and
also
I'm
planning
to
have
a
meeting
with,
but
the
shadows,
this
Friday
to
walk
them
through
the
process.
So
that's
all
the
other
time
right
now,.
P
H
So
one
note
on
the
release,
notes
really
quickly.
Sorry
I
mean
the
so
sasha
has
a
PR
up
for
KK
right,
a
change
to
the
change
to
the
github
template.
The
issue
template
our
key,
our
template.
Rather
that
includes
a
section
for
Docs
all
right,
so
the
idea
is
that
the
idea
is
that
we
should,
at
some
point,
via
magical
automation,
be
able
to
pull
in
the
same
way.
We
do
with
the
release.
H
Notes
section
within
the
PRS
be
able
to
do
the
same
with
Doc's
right
so
include
a
description
of
the
doc
and
a
link
to
the
actual
doc
right,
so
that
should
be
coming
soon.
I'm
going
to
release
the
hold
on
it
today
after
I
send
an
email
to
kubernetes
stuff
so
moving
forward.
If
we
have,
if
you
have
people
who
have
PR
soap
and
are
you
tracking
pr's,
so
this
is
like
bug
triage
and
whoever
may
be
looking
at
stuff.
Make
sure
try
to
encourage
people
to
include
a
Doc's
link.
A
Q
A
K
Yeah,
so
we
will
be
caught
up
at
you
this
week,
I
all
let
Steven
update
since
he's
gonna.
Do
it
I
again.
H
H
If
I
send
out
a
doodle
I'm
like
not
sure,
if
lik
we
can
close,
it
open
and
close
a
doodle
within
less
than
a
day,
but
it
will
be
recorded
I'm
going
to
find
a
somewhat
convenient
time
for
people
to
pop
into
that
and
and
then
after
we'll
post
it
on
the
cig
release
playlist.
So
the
idea
of
so
we
did
this.
We
started
doing
this
last
cycle.
Just
so
people
can.
H
H
So
yeah
we're
going
to
be
so
we
want
to
plan
a
shadows
meeting.
We
did
this
last
cycle
and
we
got
really
good
feedback
going
through
some
of
like
the
the
exit
interviews
and
the
applications.
Overall,
we
got
really
good
feedback
on
that.
So
we
want
to
do
that
again,
so
we're
thinking
that
we
hold
two
of
these
meetings,
one
that
will
be
convenient
for
APAC
region
and
one
that
will
be
good
for
us
base
time
zones
so
stay
tuned
for
that
it
will
probably
be
Monday
Tuesday
next
week.
A
If
you
have
any
questions,
if
you're
raising
any
of
those
enhancements
yourself
feel
free
to
ask
the
enhancements
team
or
me
on
what
state
that
actually
is
that
wanted
to
set
out
into
open
discussion.
I
sent
up
a
survey
last
week
for
more
friendly
EU
APAC
time
zone,
release
team
meetings,
I've
taken
a
quick
scare
of
that
of
the
responses.
So
far,
if
you're
really
interested
now's
your
time
to
vote,
we
have
not
too
many
responses,
so
I'm
really
looking
to
get
a
quorum,
and
that's
really
for
the
folks.
A
You
know
that
are
in
Indy
and
AIPAC
and
making
this
meeting
extremely
light
for
them.
We
appreciate
you
being
here
if
you're,
very
interested
in
having
another
one
I
have
asked
for
that.
Those
survey
results
to
come
back
by
Wednesday
and
then
we'll
make
a
decision.
The
release
leads
we'll
make
a
decision
that
brings
us
to
any
other
business.
That's
not
on
the
agenda.
I
see
one
being
typed
in.
Is
there
anything
else
or
any
questions
for
me
specifically.
A
G
H
Hello
again
so,
as
it
creeps
up
on
us
every
every
quarter,
it
is
time
for
the
cig
release,
update
for
the
community
meeting,
so
I
guess
either
myself
or
Tim
oral
tag
team
or
something
for
that
this
Thursday.
So
if
you
have
ideas
of
what
you
would
like
to
talk
about-
or
you
did
something
really
cool
within
the
last
cycle-
that
you
want
us
to
shout
out,
I'm
going
to
drop
a
doc
and
I
will
do
that
very
soon
and
you
can
just
dump
ideas
in
there
will
be
putting
slides
together
this
week.
So
yep.
A
Yeah
I
think
just
off
the
top
of
my
head
tests.
Infra
does
the
whole
thing:
roll
release,
note
site
there
to
two
large
pieces
and
maybe
the
establishment
of
the
release,
branch
management
and
your
release.
Engineering
team
I
think
they're
the
biggest
functional
changes
off
the
top
of
my
head,
but
anything
else
that
people
have
so
linked.
We're
talking
that.
H
So
sorry,
fine,
so
Jeff
you
had
ideas
about
that.
Moving
forward,
I
don't
have
Jeff.
So
the
markdown
yeah.
R
Hi,
yes,
yes,
good
all
right,
the
markdown
is
still
going
to
persist
between
releases,
so
we
do
still
need
to
generate
it
for
the
new
release,
but
over
the
next,
probably
three
to
four
weeks,
I'm
going
to
be
working
with
a
nago
and
then
the
release
notes
Bosch
utility
to
have
that
call.
The
golang
utility
yeah
exactly
but
yeah
so
once
like
Onaga
was
actually
run,
it'll
actually
generate
the
JSON
and
then
PR
it
against
the
release
notes
website,
but
that
doesn't
preclude
us
from
generating
the
release.
Notes
markdown
that
Glenn's
talking
about
sorry.
R
Part
of
the
problem
right
now
is
whenever
there
is
a
dot
release
so
like
1:15,
one
gets
generated,
we're
not
actually
generating
a
1:15
1
release,
notes
JSON
for
the
website
that
we
should
automate
so
when
they
just
do
release
that
gets
P
yard
against
the
release
notes
website.
So
that's
what
I'm
going
to
be
working
on
in
the
background
and
the
shadow
is
to
try
and
get
that
to
actually
use
the
goaline
utility
rather
than
a
bunch
of
Voodoo
bash.