►
Description
Kubernetes 1.18 Release Team Meeting (BurnDown) for 20200306
A
Good
morning
afternoon
evening
and
happy
Friday,
everyone
welcome
to
yet
another
meeting
for
the
kubernetes
released
him.
118
I
will
be
your
host
today.
My
name
is
Jorge
next
and
before
we
get
started,
like
always,
I
have
to
mention
them.
The
kubernetes
community
has
a
code
of
conduct
which
essentially
balls
them
to
please
be
nice
and
kind
to
each
other
at
all
times.
A
B
B
We
can
walk
through
those
right
now.
The
first
one
is
the
sidecar
containers
kept
and
I.
Think
people
are
aware
there
was
a
lot
of
flurry
around
that
towards
the
end
beginning
of
the
week,
as
we
got
closer
to
code
freeze-
and
there
was
an
exception
requested
for
that,
but
they
have
since
deferred
that
to
119
based
off
of
some
discussion
that
was
happening
in
the
PR.
B
Essentially
like
Kubla
code
is
very
complex
and
this
is
gonna
add
more
complexity,
so
it
just
really
requires
more
review,
so
they
wanted
to
wait
for
119,
for
that.
Second,
one
is
exception,
Arsena
issue,
number
enhancement,
number
770
and
that
one
is
related
to
CSI
volumes.
Drivers
they've,
except
they
re
they've,
asked
for
an
exception
for
that
they
had
LG,
TM
and
approved
by
the
deadline
yesterday,
but
it
needed
a
rebase.
B
B
So
those
two
wait
for
that
PR
to
go
through
693
had
the
LG
TM
and
approved
by
the
deadline
last
night,
but
the
test
flakes
kept
it
from
merging
and
then
the
retest
passed.
So
it
looks
good
now
1020
for
the
PR
that
they
needed
in
kitten.
The
KK
repo
was
not
milestone
by
the
deadline
last
night,
so
they've
asked
for
an
exception
today,
and
then
we
have
1559.
B
B
One
comment
saying
that
they
think
that
it
should
defer
to
119
because
it
came
in
at
the
last
minute
and
it's
unfair
to
the
exception
process,
and
then
the
other
one
is
indicating
that
it's
a
long
pole
for
some
other
things
and
should
really
be
considered.
Even
though
it
came
in
at
the
last
minute.
I
have.
C
C
Fields
are
I've
worked
with
the
owners
of
that
particular
body
of
work,
more
quote
well,
over
a
month,
I
would
say,
and
almost
pinging
towards
the
end
on
the
daily
basis.
I
think
personally,
my
fields
are,
it
should
be
deferred
because
I
did
everything
I
could
to
get
them
up
to
speed.
Well
in
advance
of
code
freeze
and
then
code
grease
came
and
we
were
working
till
like
the
very
I
mean
we
went
over
for
this
one
I
mean.
B
It's
that
happens
on
a
lot
of
these
things
that
771
up
above
this
is
actually
something
like
770
and
603
were
things
that
deferred
from
117,
because
it
finish
the
PR
in
time
and
now
we're
back
in
the
same
boat,
where,
like
it,
didn't
get
approved
and
merged
before
coach
Frese,
because
it
was
kind
of
a
last-minute
thing.
So.
B
And
then
there's
two
more
166
is
tank
based
eviction.
They
asked
for
an
exception
before
coach
weeks
because
they
knew
they
weren't
going
to
get
it.
We've
already
approved
that
one
and
then
1393
we
had
removed
from
the
milestone,
but
I'd
add
it
back
today,
everything
that
was
merged
for
that,
except
for
a
PR
that
adds
and
and
tests
and
that
PR
for
the
end
end
test
was
good
to
go
yesterday,
but
it
was
blocked
by
flaky
tests
merging
so
they're
running
three
tests
on
it
right
now.
A
Have
a
request
for
more
a
actually
a
what
you
just
said
about
the
I
forget
the
actual
name
of
the
PR.
But
could
you
actually
a?
Could
you
actually
add
it
in
the
notes?
Marki,
please
just
to
have
a
just
a
cut
the
context.
Yes,
we
are
going
through
all
the
exceptions,
yes,
I
wa.
Thank
you!
So
much
and
other
than
that
does
anyone
have
any
comments.
Questions,
suggestions
for
enhancements.
E
A
E
E
There
are
other
three
terms
that
are
failing
in
Mass:
reforming
the
witness
jobs,
1909
and
2019,
which
were
failing
for
the
same
reason,
but
now
they
seem
to
be
broken.
So
I
want
to
pink
some
people
after
their
meeting
to
check
if
something
change
the
bill
to
package
the
steps
it's
still
failing.
But
this
is
not
a
critical
issue.
I
left
the
comment
there,
it's
failing,
because
the
the
180
release
for
Sierra
to
say
it
doesn't
exist.
E
Yet
so
it's
okay,
not
failing
in
release
misc,
we
only
eight
eight
Souls
flaky
and
only
one
that
it's
reported
as
a
failed
failure,
but
it's
it
still.
Flake,
it's
really
hard
flake
and
in
118
plugging
in
forming
jobs,
there
are
actually
the
same
source
with
master
blocking
and
corresponding
master
reforming
jobs
which
all
the
both
of
us
or
other
investigation
and
the
flagship
that
was
talking
about
the
CIA.
A
One
quick
comment:
I
wanna
put
this
on
the
record
and
I
want
you
to
call
me
out
and
get
some
really
mystery
in
this
situation.
So
at
least
the
what
is
the
one
thing
I
didn't
want
to
mention
in
same
job,
failing
you
master
blocking,
which
is
skew
cluster
latest
a
keep
Quarles
to
a
stable
one
GC,
so
just
to
go
just
over
to
go
through
that
one
really
quickly,
and
please
help
me
out
if
anything
sounds
funky,
so
as
a
name
entails,
this
is
doing
a
skew
test.
A
A
of
coral,
mainly
a
between
lay
between
ladies
and
stable
one,
if
you
actually,
if
you
actually
go
to
the
job
and
a
lot
of
testing
from
and
stuff
it,
it
appears,
then
it
is
doing
that
down
grid.
So
it
starts
with
latest
which
in
this
case,
is
everything
in
the
master
branch
which
is
tagged
with
119
8.0,
and
then
you
down
grids
to
stable
one,
which
I
believe
is
170
a
117
4
and
for
everyone,
that's
wondering
why
I
know
any
of
that
is
because
I
spend
a
lot
of
time.
A
Looking
at
the
logs
a
so
we
know
this
is
just
black
magic
that
a
that
is
appear
that
it's
appearing
now
so
going
through
they're
going
to
the
test.
It
appears
there
was
a
new
resource
that
was
added
on
one
it
in
a
Ingrid's
classes,
which
is
to
manage
ingress
resources,
and
there
is
an
end-to-end
job
that
is
executed.
That
is
executing
the
cube
in
cube.
A
Kernel
inside
of
KK
and
into
an
end-to-end
job
essentially
tries
to
take
data
from
a
from
HDD
and
there's
an
entry
for
group,
a
group
version
and
resource
the
ingress
classes
was
a
a.
The
English
classes
were
certain
in
118.
So
if
he
goes
from
one
a
features
from
that,
the
job
in
118
with
a
bonny
teen
cluster
or
newer,
it
should
work
put
either
a
but
I
believe
the
liability
issue
is
due
to
the
fact
down
a
bit.
Maybe
cuz
we're
running
a
skew
job.
We
are
actually
running
the
end-to-end
test
from
117.
A
Now,
from
enough
from
118
and
117
day,
job
was
not
meant
to
work
with
the
ingress
classes,
and,
given
that
we
are
running
on
a
server,
so
the
actual
cluster,
a
the
actual
monster
is
still
118
something
and
they
end-to-end
tests
are
from
117.
That's
a
that's
where
the
issue
is
happening,
so
I
guess
the
thing
that
I
hear
that
I
want
you
to
double
check
me.
Please
tell
me
if
this
is
how
wrong
assumption
this
is
not
X
strictly
released
blocking
right.
A
E
A
Yeah
I
also
a
also
a
to
add
more
context.
The
reason
my
wave
1
masking
this
right
now
is,
you
know,
giving
the
we're
on
a
what
is
it
go?
Freeze
you
want
to
make
you
want
to
make
sure
that
we
it
you
know
just
a
double
check
to
make
sure
that
we
don't
have
something
completely
broken
in
coordinating
but
yeah
other
than
that.
Does
everything
sound
reasonable
to
you
yeah
or
does
anyone
have
any
other
comments?
Questions
suggestions
about
this
discussion.
A
C
Good
morning,
good
afternoon
and
good
evening,
happy
Friday,
aw
bug
triage
is
currently
tracking
red
for
open
non-test
flakes.
We
are
at
7:00,
that's
down
from
that's
down
five
four
priority:
critical
urgent.
We
are
at
one
that
is
down
one
previous
week.
Priority
important
soon
is
at
two
that
is
down
to
four
open
KK.
We
are
at
26,
which
is
down
to
four
priority
critical
urgent.
We
are
at
two
which
is
down
to
priority.
Important
soon
is
at
16,
which
is
down
two
and
finally
open.
Kk
issues
is
at
23,
which
is
down.
C
Six
I
do
have
a
few
notes.
We
have
a
link
to
a
doc
that
tracks
items
over
milestone
travel.
We
are
also
waiting
to
see
if
the
new
exception
requests
are
going
to
be
approved,
and
then
we
will
remove
their
remaining
ones
into
feature
milestones
on
Monday.
We
will
also
work
with
you
George
to
ensure
that
have
approval
before
removing
any
milestone
items.
Finally,
I
did
add
a
note
for
exception.
Request
eight
one
three
well,
this
particular
item
is
not
tracked
across
multiple
milestones.
It
was
opened
two
days
prior
to
code.
C
A
A
F
A
G
Happy
Friday
Oh,
so
we
are
finishing
up
loading
on
the
sega
domes
for
the
blog,
and
that
involves
each
of
the
folks
on
comms
team
stack,
ranking
their
choices
for
features
that
they
think
are
worth
highlighting
and
then
we'll
be
reaching
out
to
the
relevant
teams
working
on
those
features
to
collaborate
on
that
blog
post.
That
will
eventually
go
out
so
we're
on
the
path.
A
Per
month
twice
and
sold,
thank
you
very
much
and
also
for
the
audience.
If
anyone
is
interested,
if
anyone
is
interested
in
the
pay
in
what
comes
he's
doing
right
now,
Paul
are
there
a
couple
more
notes
on
a
how
the
whole
boring
a
on
take
items?
It
works
so
take
a
look
if
you're
interested
a
quick
PSA
and
moving
on
to
release
branch
manager,
Carlos.
H
A
A
It's
okay
at
some
point
during
the
day
and
the
PAPR
tiny
essentially
turns
hunk
of
a
country's
was
merged
and
some
PRS
were
removed
from
if
I
look
here,
I
yeah,
I,
guess
independent
of
that.
So
a
so.
The
work
to
things
on
a
got
a
code
freeze
happen
a
little
bit
earlier
than
most
people
are
used
to
and
which
it,
which
is
not
an
a
which
is
not
an
issue
and
the
other
one
is
an
some
peers
were
removed
from
them
from
the
milestone
a
in
the
nature
of
strictly
following
they.
A
The
deadline
of
a
5pf,
a
APR,
has
to
be
appropriate,
GTM
and
I'm,
paying
the
milestone
before
before
end
of
day
and
I.
Guess
just
wanted
to
go
well
for
I
wanted
to
talk
for
a
quick
discussion
right
now.
They
just
to
get
a
general
feel
from
from
us
Crowley's,
French
management
and
mostly
from
Bach
triage.
A
If,
if
we
follow
it,
we
follow
the
comment,
the
cut
that
Katherine
made
on
slack
a
whether
we
can
a
so
so
essentially,
let's
say
that
we
did
a
code
freeze
the
same
way:
either
say
the
same
dates
way
for
the
next
release,
so
on
Thursday
a
we
have
the
deadline,
a
we
have
code
freeze
on
a
on
a
Thursday
and
then
a
turn
on
a
actually
turn
on
code,
freeze
that
they
add
they
through
that.
Okay,
what
what
would
you
all
think
a
bubble?
You
all
think
about
that
I.
H
A
H
F
H
Like
a
comment
yesterday
in
the
PR
yeah
but
I
like
I,
was
not
completely
like
clear
on
my
method,
because
I
put
on
the
end
of
the
day,
yo
D,
but
I
didn't
mention,
which
time
zone.
What
maybe
was
my
fault,
okay
and
I-
did
that
when
I
thought
that
the
PR
was
like
I
read,
approve
it
and
merge
it,
I
thought
was
like
getting
more
late
in
the
US
time
zone,
but
they
only
spoke
me
first.
Even
he
said
okay,
it's
that's!
Not
not.
H
A
Actually,
I'm
done
actually
on
that
note,
I
want
to
follow
the
line
of
the
drain
of
data,
a
code
freeze
since
you're
doing
the
PR.
It's
up
to
you
this
case.
It's
up
to
your
discretion.
You
was
young,
every
aquarium
chart
a
whoever
is
you
know
from
the
release
branch
management
side,
it's
up
to
your
discretion,
I'm
gonna!
Whenever
you
a
whenever
you
merge
it,
okay.
D
A
Really,
don't
really
don't
think
the
vena
strict
about
times
a
time
zone
should
matter.
So
what
would
you
a
moment
to
think
about
this
process?
So
we
make
the
deadline
for
code
freeze
a
Thursday,
but
we
would
be
if
we
actually
turned
on
a
code
phrase
that
they
hate
the
day
after
work.
That
way
we
can
we
get
that
way.
We
can
it
that
way.
We
can
actually
be
strict
about
about
PRS,
PRS
and
issues
I
make
it
I
make
it
into
the
milestone.
D
A
Having
to
do
a
lot
of
checking
up
then
on
this
last
point,
this
is
something
that
I.
This
is
something
that
I
would
love
to
hear
from,
but
a
bug
triage.
If
anyone
has
thoughts
because
a
because
the
other
part
is
a
I
old,
a
wish
a
we
should
do
say
we
should
do
something:
the
Dawson
a
that
wasn't
just
cut
off
in
turn
a
I
think
we
should
do
something
to
loosen
just
cut
off
a
the
entire
community.
A
You
know
like
okay,
it's
four
feet,
a
it's
five
o'clock
and
sorry
you
didn't
make
it,
but
the
other
thing.
We
should
definitely
make
a
hard
deadline
to
allow
for
1:18
to
soak
in
CI
and
also
a
figure
out
a
process
that
is
not
going
to
make
it
sort
troubled.
A
so
troublesome
for
Bactria
stood
a
for
bug,
triage
and
other
people
to
do
their
thing.
So
from
the
bug,
triage
thing,
a
marquee
or
anyone
from
the
DIMM
hey
and
do
you
all
have
any
thoughts
opinions
at
this.
C
C
Thank
You
Jeremy
appreciate
that
so
yeah
5:00
p.m.
but
then
what
we're
doing
is
we're
saying
5:00
p.m.
but
we're
not
actually
doing
anything
until
the
following
day:
okay,
that
we're
giving
we're
giving
we're
saying
5:00
p.m.
but
we're
not
actually
taking
any
action
on
any
of
the
open
items
until
the
following
day.
A
C
Be
very
honest
with
you,
my
my
personal
opinion
is
no
matter
what
we
do.
We
will
still
get
complaints.
So
I
say
you
said
code
freeze.
Is
it
5:00
p.m.
Pacific
time
and
then
turn
on
everything
the
next
day,
and
that
gives
everybody
a
chance,
but
you
tell
them
it's.
5:00
p.m.
and
I
met
I,
give
them
sort
of
a
six
to
eight
hour
grace
period,
if
you,
but
but
even
with
that
I
can
I
can
assure
you
that
there
will
still
be
complaints.
Yeah.
A
Absolutely
but
at
least
on
the
boat
rejection
from
anyone
in
the
in
the
team
that
is
also
affected.
Will
it
change
the
dynamic
a
lot
if
we,
if
we
actually
turn
on
Godfrey's
either
they
afterward
do
us
in?
You
was
someone
who
does
both
triage
it
would
that
would
I
have
a
huge
effect
and
you're
like
what
it.
A
C
D
Something
so
I
totally
agree
with
everything
that
marki
has
just
said,
and
what
is
more
important
to
me
is
that
all
of
this
is
communicated
very
clearly
to
everyone
involved,
and
that
is
that
is
properly
documented
and
that
there
are
emails,
reminding
people
that
these
rules
are
in
effect,
so
that
we
don't
have
the
same
situation
that
we
have
this
time.
Where
there's
some
in
certainty
and
we're
improvising
a
little
it
and
going
with
with
the
flow.
So
I
just
wanted
to
add
this
strong.
A
Okay,
it's
it
also.
Thank
you
very
much
for
all
the
a
for
all
your
input.
Everyone
I'll
try
to
surface
the
conversation
in
some
way
in
some
way
to
aid
in
slack
or
mainly
a
Australian,
the
mailing
list,
just
to
just
it
just
to
double
check
with
everyone
and
try
to
get
more
a
try
to
get
more
complaints.
A
So
the
only
things
I
think
really
matter
is:
if
there
is
anything
that
we
can
make
to
any
anything
that
we
should
that
we
should
do
any
any
change
that
we
should
that
we
should
perform
to
make
a
buck
triage,
a
work
easier,
smoother
and
if
we
can
also
a
help
and
give
a
little
bit
more
be
a
more
bit
more
breathing
room
to
people.
A
also,
but
at
least
at
least
those
are
my
general
thoughts
they
meet
at.
Does
anyone
have
any
more
comments,
questions
suggestions,
anyone
anyone
wants
to
add.
A
Going
once
twice,
okay,
so
again
a
thank
you
very
much
Carlos
in
release
branch
management.
In
a
sorry
for
dragging
on
that
one,
a
little
bit
a
married
to
split
I,
don't
see
a
message
in
anything
like
it's
on
the
line
right
yeah,
so
moving
on
to
my
part,
so
might
released
lead
update
a
couple,
a
couple,
more
deadlines
to
keep
in
mind,
because
that
is
a
whole,
a
that
is
a
that
is
the
most
fun
part
about
the
released
him
after
you
get
what
a
there
you
get
done
with.
A
One
thing
have
to
start
playing
for
the
next
one
and
the
next
one
is
dog's
deadline.
Prs
I
have
to
be
ready
for
review
on
next
Monday
and
that
is
Monday
March
9
and
they
wanted
him.
Beta
2
is
planned
to
be
released
on
Tuesday
March
10,
which
is
next
Tuesday,
and
please
correct
me
if
anyone
thinks
so
in
any
of
those
states
are
wrong
other
than
that
I'm
just
going
to
go
on
and
hey
ahead
with
my
public
service
announcement
down.
A
If
there
is
anything
whatsoever
that
you
think
it's
funny
anything
that
you
think
is
interesting,
anything
that
you
learn
anything
that
you
think
you
might
be
done
better
in
the
future,
please
feel
free
to
add
in
them
in
the
release
retrospective.
So
we
can
discuss.
So
we
can
have
a
larger
discussion
about
your
own
ideas,
issues,
comments
and
with
that
I'm
gonna
stop
and
ask.
Does
anyone
have
any
final
comments,
questions
suggestions
for
anything
whatsoever
that
we
have
discussed
today.