►
From YouTube: Kubernetes 1.10 Release Burndown Meeting 20180316
Description
A
Welcome
everybody:
it
is
Friday
March,
16th
2018.
This
is
the
kubernetes
1.10
release,
burndown
meeting
and
I.
Am
your
host
Jay,
singer,
Dumars
and
also
release
lead
and,
if
you'd
like
to
follow
along
after
the
fact,
the
meeting
minutes
are
available
at
bit.
Dudley
slash,
kids,
110
burn
down
and
you
can
see
what
we're
talking
about.
They're
gonna
go
quickly
through
the
timeline
this
week,
just
clean
up
clean
up
clean
up
trying
to
get
some
of
these
major
long
learning
issues
resolved
getting
some
traction
there.
A
A
Final
fast
forward,
all
that
good
stuff
associated
with
code
freeze
lifting
will
do
the
PR
that
basically
does
the
lift
at
some
point
there.
It's
ready
to
go
timeline
for
week.
13
is
actually
the
release
and
it's
on
Monday
the
29th
and
then
we'll
do
the
retro
on
Thursday
metrics
review
I
put
a
link
in
here,
so
that
will
be
forever
more
for
the
issue
tracking
spreadsheet.
That
Josh
does
will
cover
through
that
in
just
a
minute
and
open
issues
in
the
milestone.
A
A
It's
down
I'm,
not
sure,
there's
a
nine
there,
but
maybe
it's
changed.
Let's
take
a
look
six!
Okay,
so
there
are
six
in
down
by
five
and
none
of
the
other
metrics
are
trending
any
particular
way
open.
Prs.
The
milestone
is
pretty
low
right
now
at
four
we're
up
by
nine
since
last
check,
so
velocity
is
slowing
as
it
should
getting
into
when
code
freeze
lists
and
hopefully
we'll
have
minimal
changes
going
in,
because
those
are
the
ones
that
need
to
be
cherry
picked.
A
Open
PR
is
the
milestone.
Let's
go
over
those
real
quick,
we
have
the
the
one.
Obviously
that
will
remain
until
the
release,
which
is
the
cube,
medium
merge,
and
then
we
have
this
one.
That's
the
added
unscheduled
taint
and
the
autoscaler
a
and
B
VARs
to
Kuby
and
veto
hotfix.
So
that's
the
first
one
is
six
one,
one
six
one
which
is
a
recent
ad
and
six
one,
one,
one
nine
which
has
been
there
for
a
few
days
and
six
one
one.
One
nine
is
just
waiting
for
an
LG
team
and
approve
and.
A
There's
one
milestone
and
complete,
which
is
needs
sake,
assigned
to
it
and
proof
for
milestone
labels.
Obviously
that's
tied
to
the
sake.
I
didn't
treeoche
what
six
should
on
it,
but
maybe
we
can
figure
that
out
in
the
course
of
this
meeting
and
long
term,
there's
nothing
out
there
right
now,
so
that's
kind
of
the
rundown
of
where
we're
at
and
let's
go
ahead
and
go
into
issues
so
Josh.
You
want
to
take
the
floor
from
it.
B
Yeah
well
so
we're
continuing
to
make
progress.
What
I'm
happiest
about
is
one
of
our
major
blocker
issues
just
got
closed,
which
was
the
cubelets
not
I,
don't
know
something
about
cubelets
and
pod
creation:
hey
that
one
that
one
has
been
officially
fixed
which
takes
that
out.
So
really
what
we're
looking
at
still
pending
major
issues.
So
there's
basically,
three
things
on
resolve:
one
is
the
performance
issues
which
are
making
progress
but
not
closed,
and
hopefully
somebody
else
knows
more
about
those
that
I
do
on
this
call.
B
B
Basically,
everything
on
that
is
done,
except
for
the
taint
which
there's
a
PR
open
for
the
annoying
thing
about
the
taint
PR,
which
is.
B
Six
one
one,
seven
one
I
think
or
six
one
there
we
go
the
annoying
thing
about
the
tape
PR
is
we
need
sign-off
from
one
of
the
owners
and
controller,
and
it
seems
that
everybody
in
that
owners
file
is
out
today,
so
we're
not
going
to
get
that
result
until
Monday
the
because.
B
B
B
B
The
so
that's
it,
we've
got
a
bunch
of
other
things
that
we've
got
a
few
other
sort
of
actually
really
nothing
else.
We've
got
a
couple
of
failing
tests
on
them
to
check
on
the
other
thing.
I
wanted
to
actually
check
is
with
a
signal
here.
He
is
weather.
We've
had
several
cycles
of
good
test
results
for
the
upgrade
and
downgrade
tests,
so
that
Tim
and
I
can
stop
tracking
those.
A
Assuming
that
that
is
the
case,
then
yes,
we
should
we
should
be
able
to
get
a
better
run
after
that,
hey,
let
me
see,
is
Caleb
on
the
call
just
out
of
curiosity,
yeah,
hey
so
in
terms
of
timing
for
the
reading
edit
cut.
If
we
do
that
on
Monday,
do
you
do
you
do
that
before
or
after
code
phrase
gets
lifted,
it
doesn't
matter.
A
C
One
thing
needs
to
note
is
that
the
GC
presubmit
cause
briefly
out.
Last
yesterday
afternoon,
we
due
to
the
logging
syncs
that
driver
logging
Cinco
de,
is
reached.
That
means
the
logging
sink
was
leaking
and
we
manually
cleaned
up
the
logging
sinks
and
unplugged
the
test,
but
I'm
sure
why
it
was
leaking
in
the
first
place,
and
it's
probably
just
like
slowly
piling
up
was
fine.
C
C
It's
probably
got
not
going
to
affect
this
release
because
pining
up
was
like
kind
of
slow,
but
I
mean
not
sure
like
when
next
time
they
were
like
bite
us
again
for
the
release.
Ii
I
looks
like
we
get
more
fixes
ain't.
We
is
still
pending
on
6-1
6-1.
It's
pending
a
Kubo,
wonder
if,
like
top
that
approver
can
just
prove
that,
so
we
can
get
some
like
guys
signals
over
the
weekend.
A
F
C
A
A
F
C
E
C
F
F
C
A
A
F
A
F
A
F
G
I
I
looked
at
it,
it's
fine,
I'm
I
have
a
X
open
to
correctly
skip
I
forgot
how
ginkgo
worked
and
when
things
get
said,
okay,
so
those
three,
the
three
that
we
intended
to
skip
still
got
run
and
so
they're
red,
but
we
know
they're
red
and
we
will
correctly
skip
them
shortly.
So
that's
not
concern
okay,.
A
Okay,
so,
okay,
so
in
terms,
let's
go
back
to
your
issue
tracking,
your
singing,
so
the
Stila
lady
sweets,
this
Joffe
right
here
in
theory,
it
could
already
be
fixed,
we'll
find
out
when
we
give
the
RC
and
the
yellow
ones
are
serve
all
in
progress
for
the
most
part.
Is
that
feel
right
to
you,
Josh
the
Alpha
features
one
is
that
is
that
still.
B
That
is
actually
hung
up
on.
The
demon
sets
okay.
So
yes,
our
remaining
failing
tests
as
far
as
I
know,
and
please
correct
me
if
I'm
wrong,
I
haven't
actually
I'm
in
the
process
of
you
know.
I
already
went
through
everything
that
github
tells
me
has
changed
since
last
night
and
now
I'm
going
through
every
every
one
of
these
in
detail,
because
github
doesn't
tell
me
about
all
types
of
changes.
B
E
B
A
All
right
that
actually
is
relatively
heartening,
so
so
the
plan
is,
it
stands
now
as
we'll
get
the
stuff
merged
for
that
other
stuff.
The
RC
will
cut
some
time
during
the
day.
Monday
get
some
signal.
If
signal
looks
decent,
we
proceed
with
code
lifting
code
freeze
if
for
some
reason
that
we
need
to
continue
the
bisect
work.
If
we
live
code,
freezes
that
kinda
negatively
impact
the
ability
to
troubleshoot
that.
D
H
A
Consider
Janice
my
snarky
female
alter
ego,
so,
okay,
cool,
fantastic!
Thank
you
Tim,
all
right,
so
things
are
looking
good.
C
You
wanna
hit,
hurt
or
yeah
you
see,
oh
so,
for
a
great
test.
I
take
the
glanced.
Look
there's
like
what
one
last
issue
is
like
Tony
mentioned.
We
need
to
risk
it.
The
one
of
the
proxy
be
one
test
Danny
to
be
skipped.
The
other
one
is
the
I
think
it's
sharing.
The
same
issue
is
the
flu
entities
dinner
was
that
since
that
car
is
merged,
hopefully
that
will
bring
the
Syriac
or
it
has
to
be
green
as
well.
C
A
Yeah:
okay:
let's.
A
C
A
C
C
We
still
you
to
document
this
essay
now
issue
a
sweet
end,
but
if
this
one
9pr
can
make
the
dungaree
test
to
pass
I
think
it's
a
good
thing,
so
we
can
see
if
there's
other
issues
with
the
Dongri
past
right.
That's
currently
the
dungaree
they
self
do
not
happen
yet
so
like
now,
the
test
has
been
ran.
C
A
A
All
right,
cool.
A
This
is
the
documentation
of
the
known
issue:
there,
yes
cool,
okay,
all
right
anything
else
sent
for
in
the
bun
nope
great.
Thank
you,
user-facing
documentation,
Jen
is
out,
but
I
think
these
are
mostly
correct.
So
pretty
much
still
we're
at
marketing
I
still
need
to
review.
Your
also
needs
to
review
those
if
you,
if
you
want
to,
if
anybody
wants
to
click
on
that
link
there,
this
sudden
need
to
answer
these
questions
that
definitely
needs
something
because
right
now
it's
just
blank
and
that's
that's,
probably
not
good.
So
anything
anybody
wants
to
out
there.
A
I
J
I
I
I
Fine,
so
we
should
be
fine
anyway,
if
you
go
and
look
at
the
file,
we
are
so
we,
as
I
said
we
are
now
in
the
editing
phase,
as
what
I've
done
is
basically
given
the
whole
file.
A
little
slight
great
highlight,
as
we
finish,
sections
and
and
PRS
were
taking
off
that
highlight.
So
if
you
see
something
that's
kind
of
grayed
and
it
looks
wonky,
we
haven't
gotten
to
it.
Yet
so
don't
panic!
We
are
down
to
86
pages.
So
that's
good,
we'll
keep
hacking
at
it
until
it's
reasonable.
I
I
Yeah
well,
we.
What
we
were
saying
yesterday
is
we're
going
to
just
take,
take
a
combination
of
the
preview
blog
and
just
sort
of
adapt.
It
essentially
there's
other
quote-unquote,
exciting
things,
so
that's
it
cool.
No,
it
did
you
have
anything
else.
You
wanted
to
add.
I
know,
you've
been
tracking
people,
who've
responded
and
who
haven't
no.
A
E
B
So
they
had
a
discussion
about
things,
but
their
conclusion
was
basically
they
needed
to
wait
for
class,
which
is
why
the
real
conclusion
came
from
my
meeting
with
Klaus
at
9:00
p.m.
my
time,
so
that
I
could
catch
him
in
the
middle
of
this
day.
Okay,
so
yeah
and,
like
I,
said
we
just
went
through
what
was
required
for
the
feature
and
the
one
there's
one
there's
one
outstanding
PR
associated
with
that
general
feature
that
is
not
going
to
close
for
110
and
Klaus
says,
is
not
required.
Okay
work
so
as
an.
A
Alpha
feature,
what
we
want
to
probably
do
is
just
give
heads
up
to
the
you
know
the
patch
manager
Meister
about
that,
because
typically
the
the
patch
releases
we
don't
put
feature
work
into.
But
if
it's
sort
of
bleeding
over
from
the
release
and
and
it's
an
alpha
thing-
I
think
that's
probably
an
okay
exception.
But
I
want
to
give.
B
A
A
A
C
C
It
should
be
nothing
to
do
with
110
itself.
It's
this
resources,
you
know
yeah.
We
can
either
take
a
different
zone
or
bump
Dakota
someone,
okay,.
A
J
Jase
mossad
isodd,
sorry
I,
just
jumped
on
last
second
I
had
another
meeting.
A
could
could
I
talk
to
you
after
this
meeting
or
stay
on
this
line.
I
want
to
discuss
that
open
issue.
Yes,
why
don't
you
just
go
ahead
and
discuss
it
now,
yeah.
J
J
A
J
I
could
summarize
so
I
think
we
all
agree
that
next
step
for
the
downgrade
issue
is
making
sure
that
we
have
some
sort
of
advisory
and
on
documentation.
That's
already
being
worked
on.
The
open
question
right
now
is
about
two
ways
to
actually
make
this
better.
One
is
to
make
the
antenna
the
downgrade
test
green
by
modifying
the
test
to
remove
the
finalizar
automatically.
J
The
challenge
with
this
is
route.
Fs
Wyman
has
been
looking
into
it
and
he
says
it's
non-trivial
and
it
might
be
easier
to
go.
The
alternate
alternate
route
alternate
route
is
actually
modifying
1.9
to
remove
the
finalizer.
If
it
is
present,
the
ball
is
already
working
on
that
I.
Don't
think
either
of
these
two
things
should
be
1.10.
Release
buckers
having
the
Advisory
I
think
should
be
sufficient
week.
I
guess
this
is
a
question
to
the
release
team.
J
A
Have
cut
releases
in
the
past
with
a
green
signal,
knowing
that
there
was
a
test
that
was
known
to
fail,
super
not
preferred
so,
but
on
the
other
hand,
if
we're
just
gaming
the
system
to
get
a
green
signal,
it's
kind
of
a
wash
I
mean
send
Senor
sort
of,
though
like
this
is
sort
of
in
your
ballpark.
So
what
do
you
think.
J
The
question
is
around:
we
understand
why
the
downgrade
test
is
red.
At
the
moment.
We
know
what
is
required
to
fix
it.
The
fix
is
going
to
happen
independent
of
the
1.10
release,
it'll
be
in
the
1.9
branch.
In
the
meantime,
we
have
an
advisory
that
we're
going
to
be
putting
out
with
the
1.10
release
independent
of
the
fix.
The
only
remaining
work
is
to
do
extra
work
to
make
this
test
green
by
manually,
removing
the
finalizar
it
doesn't
add
any
end-user
benefit.
It's
just
going
to
make
the
test
green.