►
From YouTube: Kubernetes Community Meeting 20190321
Description
The Kubernetes community meeting is intended to provide a holistic overview of community activities, critical release information, and governance updates. It also provides a forum for discussion of project-level concerns that might need a wider audience than a single special interest group (SIG).
See here for more information: https://github.com/kubernetes/community/blob/master/events/community-meeting.md
A
A
This
meeting
is
recorded
and
streamed
so
make
sure
you
don't
say
or
do
anything
that
you
don't
want
permanently
recorded
if
you
are
in
the
Google
Doc
with
the
community
community
meeting
notes
and
are
not
taking
notes,
please
close
it
as
it
might
be
a
little
slow
with
that
many
people
in
there
please
check
back
later
for
the
notes,
though,
today
we
have
three
six
presenting.
We
have
UI
apps
and
windows,
but
first
off
we
have
a
demo
by
Jon
Schnepp
Kay
from
VMware,
and
he
will
give
us
a
tour
and
demo
of
sonobuoy
Jon.
B
First,
we're
gonna
start
with
a
real
high-level
introduction
about
what
is
sana
Bui,
who
runs
sana
buoy
and
why
they
run
sana
buoy.
Then
we're
gonna
talk
about
the
structure
of
sana
buoy.
What
are
all
the
moving
parts
and
then
kind
of
two
demos
really
one
for
a
quick
and
simple
workflow
and
then
we're
gonna
at
least
point
you
in
the
direction
of
how
you
can
go
on
a
more
advanced
workflow
and
then
I'm
going
to
touch
on
our
roadmap.
B
So
what
that
means
is,
if
you're,
a
cluster
operator
you
may
want
to
know,
is
my
cluster
in
a
good
state
before
doing
some
important
work
or
if
you
are
having
problems
to
help
diagnose
that
you
can
run
sonically
now.
The
most
calm
plug
in
we
run
is
probably
the
kubernetes
conformance
tests.
What
those
are
just
a
series
of
tests
maintained
in
the
kubernetes
kubernetes
repo
sauna
buoy
doesn't
maintain
those,
but
we
do
make
an
easy
way
for
you
to
run
those
tests
and
any
cluster
that
you
have
available
so
again.
B
What
that
allows
you
to
do
is
earn
that
cloud
native
computing
foundation
stamp
that
your
product
is
still
certified,
kubernetes,
meaning
that
there's
a
minimal
set
of
promises
about
you
know
if
workloads
are
in
one
cluster
and
they
operate,
and
it's
a
certified
kubernetes
cluster.
If
you
move
those
to
another
certified
kubernetes
cluster,
they
should
still
work
and
sana
Bui
helps
you
do
that
in
a
simple
way.
The
structure
of
sonamoo
is
really
made
up
of
four
parts:
the
CLI,
the
aggregator
workers
and
plugins.
B
Now
the
aggregator
and
the
worker
are
just
images
really
utilizing
the
sana
Bui
binary
as
well,
but
it
all
starts
at
the
CLI
level.
The
users
gonna
run
something
like
sana.
Buoy
run.
Tell
it
what
kind
of
things
you
want
to
have
happen.
What
sort
of
tests
you
want
to
run
the
whole
conformance
suite
so
on
and
so
forth,
and
that
produces
a
whole
bunch
of
resources
inside
your
cluster,
which
includes
an
aggregator
pod.
B
That
aggregator
is
going
to
be
looking
at
the
rest
of
your
configuration
and
your
plugins
and
then
launch
all
those
plugins
with
sana
buoy
sidecars.
Those
are
the
workers,
wait
for
those
results
and
then
aggregate
them
all,
along
with
a
set
of
metadata
about
your
cluster,
to
help
diagnose
any
problems
that
did
exist.
B
The
only
real
contract
between
the
plugin
and
the
worker
is
that
you
set
the
results
in
a
predictable
place
and
then
the
worker
does
the
the
other
part
of
actually
grabbing
them
all,
sending
him
sending
them
to
the
aggregator
so
that
the
client
can
pull
them
down.
That's
why
writing
a
plugin
is
so
easy
in
theory,
it's
just
writing
a
little
bit
of
yeah
Mille,
providing
an
image
to
do
whatever
work.
You
want
it
to
do.
B
So
let
me
show
you
how
this
works.
The
most
simple
command
is
just
sauna.
Buoy
run,
this
would
run
with
all
the
default
options
which
includes
running
both
the
system
D
logs
gatherer
and
the
into
end
conformance
suite,
but
for
the
sake
of
time,
we're
just
gonna
run
in
quick
mode.
That
means
that
we're
gonna
run
just
the
end-to-end
tests
and
we're
gonna
run
just
a
single
test
and
you
can
see
asynchronously.
B
We
just
create
a
bunch
of
resources
like
a
namespace
service
accounts
and
config
Maps,
but
you
can
check
on
the
status
of
your
run,
using
two
different
commands:
either
sauna
buoy
logs,
which
is
gonna,
gather
all
the
logs
from
the
pods.
Here
you
can
see
these
are
the
logs
from
the
sauna
buoy
pot
itself.
That
aggregator
just
says
that
it's
gathering
your
plugins,
it's
gathering
metadata
starting
the
plugins
and
waiting
for
results
and
then,
above
that
these
are
the
actual
system
logs.
B
So
you
can
see
as
a
run
happens
if
you
want
to
sometimes
those
runs
can
be
pretty
long
now.
You
can
also
check
on
the
status
at
a
high
level
just
by
running
sauna
GUI
status,
which
lists
what
plugins
were
choosing
to
run
and
what
their
current
status
is.
Once
that's
done
and
everything
is
complete,
the
aggregator
again
grabs
and
tars
all
those
up
adds
a
little
extra
metadata
about
the
run
and
makes
those
available
through
sauna
buoy
retrieve.
B
There's
a
couple
folders
and
that
results
tarball,
just
like
I,
said:
there's
information
about
the
hosts,
there's
information
about
the
run,
there's
some
extra
pod
logs
and
resources.
But
if
you're
just
interested
in
the
run
itself,
you
just
look
in
the
plugins
directory.
You
can
add
as
many
plugins
as
you
want
to
run
here,
but
each
one
gets
its
own
results
directory
and
each
plug-in
can
decide
what
constitutes
results
for
itself.
So
the
end
to
end
tests
decide
they're
gonna,
give
you
the
logs,
so
you
can
inspect
those
fully
and
the
j-unit
XML.
B
So
you
can
parse
those.
And
if
you
look
at
ours
we
can
see
that
we
passed
one
test.
So
that's
the
simple
workflow
there's
lots
of
flags
that
you
can
use
to
customize.
You
know
run
this
test,
don't
run
that
test,
use
a
separate
version
or
something
like
that,
but
sometimes
you
need
to
edit
something
that
we
don't
have
a
flag
for,
or
you
want
to
run
your
own
custom
plugin,
in
which
case
you
use
sauna,
buoy
Jen.
B
This
is
gonna
dump
all
of
the
amyl
for
you
that
sauna
buoy
is
using
internally
so
that
you
can
edit
it.
However,
you
see
fit
and
then
you
apply
it
with
Q
control
yourself,
so
you
can
see
here's,
that's
the
namespace,
the
service
count
zone
and
so
forth,
but
the
two
parts
I
really
want
to
point
out
to
you
are
this
list
of
plugins
here
by
default.
B
I
said
we're
running
the
end
to
end
tests
and
the
system,
D
law
and
below
that
we
have
another
config
map
where
each
one
of
those
plugins
is
going
to
be
given
their
own
plug-in
configuration
it's
pretty
much
just
a
standard
pod
spec,
along
with
just
a
little
bit
of
sauna,
Bui
metadata,
saying
whether
it's
you
know
a
run.
One
job
run
once
type
of
plug-in
or
a
daemon
set
and
along
with
a
name
okay.
So
if
you
want
to
run
your
own
plug-in,
you
just
have
to
change
those
two
things
right.
B
B
We
have
version
0.14
coming
out
pretty
soon
in
the
next
couple
weeks
and
the
major
work
there
was
really
improving
the
support
for
the
end-to-end
tests
and
improving
the
script,
ability
things
like
adding
synchronous
actions
and
making
a
little
easier
to
work
around
in
CI
or
scripts,
but
in
v15
we
really
want
to
improve
the
custom
plugin
support
so
that
you
don't
have
to
manually
edit
the
yeah
Mille.
We
want
to
try
and
make
that
as
easy
as
possible,
so
you
don't
have
to
go
through
this.
B
A
few
extra
manual
steps,
if
you
want
to
get
a
hold
of
us
here,
is
our
github,
just
hep
tio
sana
Bui,
we're
active
in
the
sauna
Bui
channel
in
the
kubernetes
workspace
and
our
Twitter
and
a
blog
post
talking
more
about
what
sana
Bui
is
and
some
of
the
roadmap
items
if
you're
interested.
Thank
you
awesome.
A
C
A
rake,
sorry,
no
slides
again
this
week
and
I
will
actually
try
to
keep
it
really
short.
If
I
can
so
hi
everybody
I
am
Aaron
of
sleek
beard.
I
am
your
kubernetes
114
release
lead
today's
cat
t-shirt
didn't
really
print
as
lightly
as
I
wanted
it
to,
but
for
you
francophones
out
there,
it's
the
talk
act
so
because
countdown
is
happening,
we
are
almost
there.
C
C
There
was
an
example,
I
believe
of
this
happening
for
some
freaking
behavior
earlier,
so
our
plan
right
now
is
to
make
our
go
no-go
decision
on
whether
to
actually
cut
the
release
as
scheduled
on
Monday
March
25th
at
8
a.m.
Pacific
time
to
give
us
just
a
couple
extra
hours
in
case
anything
happens
to
come
up
last-minute,
so
as
you
are
finding
bugs,
and
things
at
this
last
minute,
which
you
can
really
do
to
help
us
out,
is
linked
to
the
V
114
known
issues
issue
here
on
github.
C
This
will
really
help
our
release
notes
team
make
sure
that
they
have
identified
all
of
the
possible
known
issues
you
can
see.
We
have
somebody
here
talking
about
a
core
DNS
issue
and
also
an
issue
about
google
of
failing
to
restart.
This
is
super
helpful
for
us.
It's
also
super
helpful
for
end
users,
and
this
is
also
really
helpful
for
us
to
understand
the
need
and
urgency
of
cutting
up
actually
straight
out
the
door.
I
know
we.
C
We
have
this
reputation
that
the
dotto
release
of
kubernetes
is
seen
as
not
terribly
stable,
but
you
know
I
actually
haven't
seen
too
many
cherry-picks
zoom
in
at
the
last
minute
here
and
right
now.
The
length
of
this
issue
also
makes
it
appear
as
though
we
don't
actually
have
too
many
known
problems.
That's
recognized
that
could
all
change,
but
you
can
really
help
us
out
by.
If
you
see
something
say
something.
C
Our
CI
signal
report
for
the
week
was
posted
to
kubernetes
that
on
Monday
things
in
general
are
looking
a
little
bit
better,
although
we
still
definitely
have
some
long-standing
flakes.
So
as
a
reminder
to
everybody,
if
you
care
about
excuse
me,
if
you
care
about
kubernetes
1:15,
you
should
be
caring
about.
What's
going
on
in
the
release
master
blocking
dashboard,
this
one
could
block
the
release
of
1:15.
C
The
informing
dashboard
gives
us
some
information
about
what's
going
on
and
then
the
upgrade
dashboard
gives
us
some
information
about
if
I
tried
to
upgrade
from
the
latest
released
version
of
kubernetes
to
what
is
in
last
year
right
now,
what
would
that
look
like
what
the
release
team
cares
about?
Is
the
1:14
blocking
dashboard
and
oh,
my
gosh?
This
is
actually
happening
on
camera.
There
are
no
failures
on
this
dashboard.
Can
you
believe
that
it's
amazing?
C
C
We
definitely
have
some
flakes
that
are
just
they're
just
happening
all
over,
but
one
of
the
things
that's
really
important
for
us
as
we
look
to
decide
whether
or
not
to
cut
a
release
is
to
see
if
a
given
commit
can
pass
a
test
three
times
in
a
row
and
we're
seeing
a
lot
more
of
these
in
the
present
than
we
did
in
the
past.
So
that's
a
good
sign.
C
Because
things
look
like
they
are
headed
on
track
and
on
schedule.
We
are
going
to
be
taking
over
the
community
meeting
next
week
for
a
retrospective
of
how
the
lease
went.
You
have
probably
been
through
this
before
if
you've
been
a
longtime
attendee
of
the
community
meeting,
we
encourage
anybody
and
everybody
to
participate
here.
C
I'm
sure
those
of
us
on
the
release,
team
who've,
been
involved
in
this
day
and
day
out
have
some
opinions,
but
we
would
also
love
to
hear
from
people
who
are
just
contributors
to
kubernetes
or
even
people
who
want
to
try
and
get
a
sense
of
what
has
been
happening
with
the
release
and
what
could
have
been
easier
for
the
communications
perspective.
So
we
have
this
document
that
literally
anybody
can
edit.
C
You
can
talk
about
what
we
did
that
we
said
we
do
well
what
could
have
gone
better
and
we'll
walk
all
through
all
of
that,
thanks
to
the
wonderful
moderation
help
of
Jase.
The
only
thing
that
I
ask
is,
if
you
put
something
in
that
dock,
you
put
your
name
next
to
it,
and
you
be
prepared
to
show
up
next
week
and
make
sure
that
you
know
you
actually
talk
to
what
you
have
put
in
the
dock.
So
come
one
come
all
to
the
114
Festivus
retro.
C
You
have
all
sorts
of
room
to
get
you
code
landed
for
115,
I.
Believe
I'll
have
to
go
back
and
confirm
this
for
the
retro,
but
it
sure
seems
like
we
actually
merged
our
backlog
of
PRS
after
a
code
thaw
faster
than
we
ever
have
before
super
cool.
Ok,
these
graphs
are
taking
a
little
bit
of
time
to
render
you
can
click
on
that
link
yourself.
When
you
look
at
the
meeting
that's
later,
if
you
want
to
find
out
or
just
go
to
Bella
Driven
Kate
SIA,
we
also
have
a
115
release
team.
C
So
I'd
like
to
I,
already
told
you
all
that
we
we
sort
of
Claire
is
you're
going
to
be
your
115
release
lead
and
we
now
also
have
staffed
all
of
the
other
roles
with
me
positions.
The
shadow
position
selection
is
TBD.
So
if
you
are
interested
in
that,
please
hang
out
in
signally
slack
Channel,
please
come
to
seek
release
meetings,
cig
release,
it's
great
and
I'm,
also
very
appreciative
of
those
of
you
who
have
started
thinking
about
what
you
intend
to
land
in
115
and
putting
together
some
plans
for
that.
C
So
we
already
have
of
115
milestone
and
we've
had
it
out
there
for
a
little
while.
So
if
there
are
some
enhancements,
you'd
like
to
lay
it
in
115,
now
is
a
great
time
to
be
going
through
the
work
of
authoring
caps,
finding
the
right
people
to
review
those
caps
and
getting
those
tracking
issues
in
the
115
milestone.
So
we
can
start
to
understand
what
is
coming
ahead
with
that
I.
C
Think
I'll
just
also
go
ahead
and
talk
about
patch
releases
a
little
bit
so
the
114
one
patch
release.
There
is
kind
of
like
no
hard
and
fast
rule
for
when
the
dot
one
patch
releases
have
historically
come
out.
I
took
a
look
at
some
of
the
data
with
Tim
pepper
who's
running
a
patch
release
management
team
and
it
seems
like
anywhere
from
a
week
to
three
weeks,
has
historically
been
when
we've
cut
those
out.
C
So
I
would
say
for
now
assume
the
two
weeks
from
the
planned
release
date
of
114
zero
will
will
send
out
114
one,
so
that
would
imply
a
cherry-pick
deadline
of
Friday,
the
fifth
and
the
patch
release
going
out
on
Friday.
The
8th,
if
you
are
subscribed
to
the
kubernetes
dev
mailing
list
and
read
it
in
any
way,
shape
or
form
first
off.
Thank
you
very
much.
That's
the
best
way
for
us
to
communicate
to
our
contributor
community.
C
C
A
Silky
smooth
sounds
of
sick
beard.
Thank
you,
so
much
Aaron,
a
great
great
update,
we're
very
much
looking
forward
to
release
and,
of
course,
the
retro
next
week.
So
thank
you
so
much
next
up,
we
have
a
quick
contributor
tip
of
the
week.
This
one
comes
from
Bob
Dylan.
If
you
are
a
contributor
to
kubernetes
and
if
you
have
an
issue
with
the
CN
CF
CLA,
if
you
change
jobs,
if
you
lose
a
loss,
account
loss
t
account
access
to
your
email.
Things
like
that.
A
You
can
contact
the
Linux
Foundation
helpdesk
at
health
desk
at
our
T
dot,
Linux
Foundation
org
that
that
one
is
way
too
hard
again.
Thank
you
so
much
Bob
for
for
that,
one
we're
gonna
dive
into
these
sig
updates
the
first
sig
we
have
a
ciggy
y
and
then
we'll
have
sig
apps
and
sig
windows.
So
for
sig
UI
we
have
Jeffrey
CK.
D
Ok,
so
apparently
I
should
unmute
myself
before
sharing
sorry
about
that.
There
we
go
all
right.
Are
we
good
thumbs
up
Jonas,
yeah,
perfect,
all
right
hello?
My
name
is
Jeffrey
sigue
I'm,
one
of
the
chairs
of
sig
UI,
and
let's
get
right
into
this.
So
last
cycle
we
actually
only
presented
two
months
ago,
so
there
hasn't
been
too
much
progress,
but
we've
submitted
two
G
sock
proposals
that
have
been
accepted
and
we're
looking
to
fill
them.
The
first
one
is
to
add
support
for
C
RDS
on
the
dashboard.
D
E
D
D
The
code
is
more
or
less
done,
but
right
now
we're
just
waiting
for
an
official
GCR
repo
to
push
the
container
to,
and
that
will
be
ready
for
the
to
fuel
release.
We
are
now
distributing
multi,
Arch
manifests
and
by
now
I
mean
when
we
eventually
cut
our
two
oh
release,
and
we
have
switched
our
CI
over
to
use
kind
for
Oliver
testing
and
that
has
been
a
wonderful
transition.
D
So
what
are
we
doing
now?
Well,
we
are
still
trying
to
finish
the
angular
migration,
keep
in
mind.
We
were
on
an
older
version
of
angular
that
was
a
couple
years
old
and
to
move
to
the
new
version
required
more
or
less
a
rewrite,
also
I'm.
Sorry,
if
my
dog
is
coming
through
over
my
mic,
she
really
wants
to
play
so,
as
we
are
doing
this
full
rewrite
we're
just
trying
to
reach
feature
parity
with
the
older
version.
D
So
we
have
this
umbrella
issue
that
is
tracking
all
of
the
efforts
there,
I
can't
say
when
we
are
going
to
get
it
done,
but
I'm
hoping
to
try
and
have
a
to
a
release
out
by
keep
county.
You
so
finish
the
migration
we're
hopefully
going
to
on
boards
and
ji-suk
students,
and
once
all
that
is
undone,
we
are
going
to
work
on
better
OS
support
and
just
a
better
authentication
mechanism
general.
D
A
C
F
Hey
everyone,
so
my
name
is
Adnan
a
bill
saying
I'm
one
of
the
co-chairs
of
sick
cats.
If
I
can
take
next
perfect,
so
what
we
why
we
been
doing
against
this
last
quarter,
I
think
we
gave
an
update
in
early
January.
Since
then,
one
of
the
things
that
we
have
is
quite
recently
is
a
POC
implementation
of
of
a
sidecar
kept
of
the
sidecar
kept
rubber,
which
is
all
about
essentially
adding
kind
of
support
for
side
cars
in
the
life
cycle
of
starting
up
starting
off
and
tear
it
down.
F
Containers
in
a
pod
and
then
there's
two
caps
that
are
being
worked
on
that
are
almost
I
was
not
yet
merged,
but
in
a
state
to
be
merged,
but
not
quite
implemented
and
and
still
some
stuff
to
be
worked
out
and
discussed
there.
Those
are
the
stateful
set
or
your
resize
gap
and
the
adding
max
unavailable
support
for
stateful
set
faster
rollouts
and
then
with
the
application
CID
stuff
we
worked
on.
There
is
implementing
status
in
the
controller
and
an
adoption
of
resources.
I'll
go
over
that
in
a
bit
so
plans
for
upcoming
cycles.
F
F
So
we
kind
of
I
think
we
went
over
this
in
the
last
update
as
well.
The
main
issue
with
cron
jobs
before
we
take
it
to
GA
is
improving
the
scalability
of
the
controller
and
switching
it
to
using
shared
informers.
The
API
itself
is
stable
and
we
don't
expect
it
to
change,
but
this
is
where
the
majority
of
that
work
needs
to
happen.
F
The
application
controllers.
So,
if
you
aren't
familiar
with
the
applications,
the
ID,
essentially
it's
a
way
to
to
group
a
set
of
resources
that
are
related
to
an
application
and
provide
a
kind
of
unified
up
status
for
an
application
and
also
manage
that
application.
So
things
that
we've
worked
on
again
status,
so
you
can
kind
of
get
in
an
aggregated
status
of
of
a
whole
application,
and
also
resources
that
are
related
to
the
application
are
now
adopted
by
the
controller.
So
dione
references
are
set
to
to
be
owned
by
the
application
cid.
F
F
We
also
kind
of
mentioned
this
last
time
as
well,
but
beta
api's
are
going
to
be
removed
and
for
workloads
fairly
soon,
so
we'll
continue
to
serve
the
beta
api's
in
114
and
115,
but
going
forward
in
116
they'll
no
longer
be
served
by
default
with
an
option
to
re-enable
them,
but
they
will
be
removed
entirely
in
118.
So
just
a
heads
up,
but
that's
happening,
a
quick
update
on
compose
K
compose,
which
is
a
sagat's
project.
It
was
grandfathered
into
the
kubernetes
walk.
F
F
So
we're
looking
people
to
we're.
Looking
for
people
to
contribute,
there's
lots
of
stuff
in
the
workloads
API
is
there's
the
control
controller
changes
that
we
need
to
make.
So
if
you
interested
in
getting
involved
in
Windows,
come
along
the
max
unavailable
for
stable,
set,
the
volume
resize
taking
pod
disruption,
budget
to
GA
still
need
to
be
worked
out,
and
then
there's
also
the
portable
service
definitions,
care
which
hasn't
really
moved,
and
that
might
be
something
that's
interesting
to
get
involved
in
as
well.
A
E
Thank
You
Jonas,
hi
everybody.
This
is
Michael
I'm
one
of
the
co-chairs
for
sick
windows
and
give
you
guys
a
quick
update
today.
The
the
biggest
thing
that
we
want
announced
is
with
kubernetes
1.14
windows,
support
that
includes
windows,
node
supporting
kubernetes,
as
well
as
the
ability
to
schedule
windows.
Server
containers
in
kubernetes
is
Koenji
a
so
this
was
a
huge
and
long
journey
for
us.
We
started
the
sake
about
three
years
ago.
E
We
started
from
kind
of
showing
the
art
of
the
possible
by
hiking
a
lot
of
the
kubernetes
components,
including
the
cube
LED
and
the
you
proxy
to
to
get
windows
containers
to
be
scheduled,
then
Microsoft
and
the
community
at
that
CNI
support
which
was
a
tremendous
update
in
functionality,
and
then
we
go
to
being
stable
now,
where
you
can
use
a
lot
of
the
features
and
capabilities
of
kubernetes
on
windows,
with
some
limitations,
of
course,
which
are
well-documented
for
use
for
users.
So
what
do
you
do
in
the
last
cycle?
E
The
biggest
thing
obviously
is
a
stable
release
of
windows
node.
We
have
a
cap
that
has
a
very
detailed
listing
of
what
works
and
what
doesn't
work
on
Windows
and
that's
something
that
we
worked
very
closely
with
seed
architecture
and
other
teams
like
signal
to
to
document
everything
and
make
sure
that
no
end
user
ends
up
in
a
situation
where
they're
not
aware
of
what's
going
to
happen.
The
second
thing
we
worked
on
is
an
alpha
release
of.
What's
called
group
managed
service
accounts.
E
This
is
a
feature
that
Microsoft
has
that
allows
you
to
run
a
container
under
an
Active
Directory
identity.
So
if
you
have
external
resources
in
your
network
like
a
database
or
a
file
share
or
anything
else,
that
requires
an
Active
Directory
identity.
Now
you
can
authorize
your
container
to
be
able
to
access
those
those
resources
and
carry
that
identity
with
it
throughout
network
calls.
E
So
that
was
a
big
amount
of
work
that
we
did
in
collaboration
with
ckp
I,
see
Karka
tech,
chure
and
C
goth,
and
we
look
forward
to
progressing
that
throughout
the
future
releases
and
the
other.
Two
big
things
is
test
automation
we
added
added
and
added
more
tests.
We
got
them
to
green.
We
got
them
to
run
on
Google,
Azure
and
vSphere,
so
that
our
end
users
can
be
assured
that
you
have
a
stable
release
of
kubernetes
that
they
can
run
in
production
and
we
have
extensive
end-user
documentation.
E
So
on
the
testing
update,
specifically
beyond
the
fact
that
you
know
we
have
our
test
grid,
that's
linked
there
and
running
on
either
GCP
and
vSphere
were
able
to
kind
of
mold
the
conformance
test,
two
things
that
should
pass
everywhere
versus
things.
That
should
only
pass
on
Windows
and
things
that
should
only
pass
on
Linux,
so
our
team
did
a
terrific
job,
they're
kind
of
segmenting,
the
test
fixing
whatever
was
necessary
and
even
modifying
a
lot
of
the
Linux
tests
so
that
they
work
very
well
on
Windows
in
the
future.
E
We
look
into
a
conformance
profile
for
Windows
and
that's
something
that
we'll
explore.
But
the
big
ask
here
is:
if
you're
a
cig
that
wants
to
introduce
new
conformance
tests,
please
come
and
talk
with
cig
windows
who
can
address
those
for
compatibility
and
see
if
there's
any
changes
that
we
want
to
make
in
there.
E
In
addition
to
that,
of
the
in
kubernetes
is
out
there
with
pretty
handy,
unstable
script
for
deploying
and
getting
your
cluster
up
and
running
with
all
the
infinities.
Our
documentation
has
links
and
guides
for
each
one
of
these
see
nice.
So
what
do
you
want
to
do?
Next?
Obviously,
we're
not
done.
We
have
a
tremendous
work
ahead
of
us
even
after
GA.
E
We
want
to
have
support
for
kube
ADM
and
as
well
as
cluster
api
and
we're
gonna
collaborate
this
across
the
lifecycle
on
that
we
are
already
started
working
on
CRI,
container
D
support
and
that
work
also
bring
the
runtime
class
for
hyper-v
isolation.
So
now
you'll
be
able
to
run
hyper-v
isolated
containers
on
kubernetes
as
well
we'll
move
forward.
E
The
group
managed
service
account
work
into
beta
and
then
obviously
we're
gonna
keep
adding
more
seein
eyes,
more
storage,
plugins
and
definitely
more
tests
with
the
GA
coming
out
or
some
customers
are
going
to
file
some
bugs
we're
gonna
address
those
and
in
subsequent
release
and
make
sure
that
we
have
a
very
stable
ecosystem
for
kubernetes
to
grow
on
windows.
I
want
to
a
quick
shout
out
here
to
some
of
the
folks
that
wouldn't
be
here.
We
add
them
between
microsoft,
docker,
vmware
and
people.
E
E
We
wouldn't
be
here
without
you
and
we
really
appreciate
all
your
effort
and
we
look
forward
to
continuing
working
with
all
of
you
guys
too,
as
you
move
forward
after
1.14
you're
going
to
contribute
to
have
weekly
meetings,
we
have
recorded
from
many
meetings
if
you
wanna
view
them
offline
and
if
you
wanna
come
in,
come
and
find
some
bugs
in
our
project
board
and
help
contribute,
and
if
you
want
to
find
us
we're
on
slack
as
well
as
a
mailing
list,
feel
free
to
come
in.
Ask
questions,
engage
the
community
and
also
contribute.
A
C
C
That's
called
conform,
its
definition,
that's
all
about
describing
like
what
conformance
is
and
what
behaviors
are
and
are
not
conforming,
and
I
would
encourage
people
who
are
interested
in
this
from
sig
windows
to
make
sure
that
they
show
up
and
participate
in
a
slack
channel
called
Keats
conformance,
and
we
have
a
meeting
that
is
scheduled
for
I
want
to
say
Tuesday
at
noon.
Pacific.
The
information
for
all
of
this
is
available
in
city.
Architectures.
Read
me.
C
If
you
are,
anybody
are
interested
I
know
there
was
a
doodle
sent
out
to
both
that
can
see
and
see
African
formats,
mailing
list
and
Sagarika
texture,
mailing
list
and
I
neglected
to
check.
I
didn't
see
your
names
or
or
any
of
the
other
names
that
she
called
out
in
your
shoutouts,
which,
just
to
underscore
I
know
I
I
personally
saw
Patrick
links
face,
show
up
repeatedly,
multiple
times
at
many
meetings
throughout
the
community,
and
so
this
was
definitely
an
amazing
broad
community
effort
and
lots
of
support.
C
E
So
Patrick
is
the
right
guy
that
attends
those
meetings
have
viewed
some
of
them
offline,
but
you
know
they
asked
earlier
was
a
little
bit
different
if
you're
sick
is
adding
more
conformance
test,
make
sure
that
you
also
are
aware
of
it.
But
yes,
we
are
were
also
engaged
in
the
conformance
working
group.
A
All
right
awesome.
Thank
you
again,
thanks
to
all
sig
chairs
here,
for
your
updates
really
really
useful,
we're
gonna
move
over
to
announcements
and
then
do
some
shoutouts.
The
first
announcement
is
that
the
contributors
summit
website
for
Barcelona
is
live.
So
if
you
are
a
new
contributor
to
the
kubernetes
project,
we
will
have
two
different
workshops
for
you,
one
which
is
called
all
the
101
workshop,
where
newcomers
to
kubernetes
and
two
open
source
project
in
general
can
learn
how
to
be
a
contributors
to
the
kubernetes
project.
A
Then
we
will
have
a
201
workshop
for
newcomers
to
kubernetes,
but
not
newcomers
to
open
source.
In
general,
you
might
have
been
a
contributor
to
other
open
source
projects.
So
now,
we'll
give
you
a
little
more
advanced
advanced
workshop
around
that
for
current
and
existing
contributors.
We
will
also
have
face
to
face
meetings
for
all
six
that
register.
A
So
if
you're
you're
in
e-cig
and
you're
participating
in
a
seg
make
sure
that
your
sig
is
actually
registered
to
have
a
spot
for
during
the
contributor
summit,
and
you
can
check
out
the
event
site
and
the
notes
here
and
register
there.
We
also
have
a
blog
post
yesterday.
It
was
posted
yesterday
is
blog
post
by
Parris,
Pittman
and
me
regarding
previous
and
future
contributor
summits.
So
please
check
that
out
as
well
on
the
content,
I'm
gonna
call
on
George.
Here
you
had
some
fantastic
updates
on
YouTube
yeah.
G
So
it
took
us
a
while,
but
we
have
now
crossed
over
10,000
subscribers
to
the
kubernetes
YouTube
channel,
so
give
yourselves
a
hand,
that's
four
thousand
one
hundred
and
eighty
seven
videos
that
you
have
published,
and
it's
also
your
annual
reminder
for
SIG's.
Please
continue
to
upload
your
meeting
videos
if
you're
behind
toss
them
in
a
bucket
or
something
and
ping
me
and
I
can
help
you
catch
up.
So
just
a
reminder
there
that
people
are
consuming
these
videos
and
they
are
really
useful
when
you
go
on
trips
to
have
that
backlog
there.
A
Awesome,
thank
you.
Now
we're
gonna
move
out
move
over
to
shoutouts.
We
have
a
bunch
of
shoutouts
this
week,
so
thank
you.
Everyone
who
are
shouting
out
to
all
the
wonderful
people
within
the
community
here
the
first
one
is
from
Andrew
cuts,
he's
giving
a
shout
out
to
Steve
Kuznetsov
and
send
Lew
for
helping
him
get
things
straightened
out
for
some
upcoming
fun
foundational
changes
to
the
way
VMware
is
running
end
to
end
tests
and
also
a
shout
out
to
Benjamin
elder
since
the
docker
file
that
was
being
used
was
stolen
directly
from
him.
A
Next
up
is
steven
augustus
is
giving
a
shout-out
to
benjamin
elder
for
his
new
profile,
pic,
which
is
hilarious
and
confided
in
the
shout
outs
channel.
So
highly
recommend
checking
that
out
next
up
is
sack
core
listen.
He
is
giving
a
shout
out
to
curse
Pittman
for
her
need-to-know
emails
for
the
kubernetes
chairs.
He
learned
so
much.
It
is
a
great
email
put
together.
What
Perris,
every
week,
sending
out
sending
that
out
to
all
the
sig
chairs
and
the
elder
Benjamin
elder,
is
also
giving
a
shout
out,
together
with
Tim
st.
A
Clair,
to
kneel
lift
1
2
3
on
reviews,
tests,
infra
and
reviewing
kind.
So
big,
big
shout
out
there,
Aaron
Craigan
Berger
is
giving
a
shout
out
to
Valerie
Lansky
for
a
bot
that
tries
to
all
label
issues
with
the
right
sig
label.
Well,
we
all
love
automation
and
we
know
that
Aaron
is
a
big
fan
of
automating
everything.
So
this
is
a.
This
is
a
great
thing
to
see.
A
Jim
angel
is
giving
a
shout
out
to
Michael
Michael
and
Craig
Peters
for
writing
a
massive
amount
amount
of
Windows
documentation
needed
for
the
1.14
release,
and
also
a
big
thanks
for
involving
the
sigdoc
screw
very
early
in
the
process.
This
is
super
important
for
for
big
new
releases.
Make
sure
that
sig
docs
knows
about
it.
Tim
pepper
is
shouting
out
to
Erin
friggin,
burger
and
Maria,
and
Tata
and
Tala
and
Josh
burgers
for
work
across
recent
months
towards
a
simplified
and
cleaner
sig
released
master
blocking
config
and
to
all
the
cigs.
A
Organizational
stuff
can
be
boring.
But
it's
super
important,
so
really
great
stuff
to
offload
to
machines
again,
more
automation,
awesome!
Those
are
the
shoutouts
here,
really
really
nice
to
to
a
large
portion
here
of
the
community.
Thank
you
all
for
for
sending
those
out
and
with
that
we're
going
to
wrap
up
this
community
meeting,
and
we
want
to
say
thank
you
to
all
the
presenters
today
for
the
demos,
the
sig
updates,
the
the
release
updates
and
everything.
Thank
you
so
much
and
thank
you.
Everyone
who
joined
us
and
we'll
help
too.