►
From YouTube: Kubernetes Community Meeting 20170413
Description
We have PUBLIC and RECORDED weekly video meetings every Thursday at 10am US Pacific Time.
https://docs.google.com/document/d/1VQDIAB0OqiSjIHI8AWMvSdceWhnz56jNpZrLs6o7NJY
Demo: Gitlab; Releases 1.7, 1.6, 1.5; SIG Networking; SIG Windows; SIG Docs.
A
B
C
C
Oryx,
so
basically,
you
know
just
to
kind
of
cut
down
some
of
the
time
here.
I've
already
got
this
pre-installed
I've
got
a
git
lab
Enterprise,
Edition
installation
running
on
communities
on
Google
cloud
platform
and
already
logged
in
so
I'm.
Just
gonna
jump
right
in
and
kind
of
go
through
our
what
we
call
our
idea
to
production
flow
and
really
it's
kind
of
you
know,
trying
to
encompass
the
entire
software
development
lifecycle.
C
All
from
within
you
lab
so
here,
I'll
just
go
and
create
a
new
group
I'll
make
it
public
and
that's
automatically
going
to
create
a
new
matter,
most
team
for
us
as
well
I'm
matter
most,
is
our
slack
OpenSocial
slack
equivalent
that
we
ship
with
it
and
here
I'm,
just
going
to
copy
and
paste
a
URL
for
a
repo
I'm
going
to
clone,
save
myself
some
typing,
it's
a
really
trivial
thing:
I
probably
could
have
typed.
It
live
just
as
fast.
C
C
And
then
the
last
thing
I
need
to
do
is
set
up.
Auto
deploy.
I
said
this
is
a
really
simple
repo
has
no
CI
CD
configured
or
anything
so
I'm.
Just
gonna
set
up
this
template.
It's
a
pretty
straightforward
template,
but
basically
gives
me
a
full
build
test
clean
up
in
a
production
staging
pipeline.
I
just
need
to
configure
one
thing
on
it.
C
And
I'll
go
ahead
and
commit
that
in
the
master
great.
So
that's
all
the
setup
stuff.
So
now,
I
kind
of
wanna
go
through
our
flow
here.
I'm
gonna
start
actually
from
matter
most,
as
we
here
get
lab
believe
that
you
know
today.
More
of
your
team
conversations
are
happening
in
chat
and
not
starting
out
on
issues,
and
so
we
really
wanted
to
integrate
with
your
chat,
client
and
so
here
there's
this
new
command
I've
just
added.
C
First
time,
I
wrote
it
I
just
need
to
connect
to
my
github
account,
but
now
from
here,
I
can
do
a
few
things.
But
now
it's
really
simple,
just
the
issues
and
deployments.
But
the
idea
is
that
you
know
we're
talking
about
some
idea
and
when
I
call
it
that'd
be
a
really
good
thing.
Let's
make
an
issue
out
of
it.
I
can
just
go
ahead
and
do
that
right
from
my
chat,
client
and
then
everybody
can
see
and
then
get
links
and
all
that
kind
of
stuff.
So
you're
gonna
create
a
new
issue.
C
C
If
you're
familiar
with
that
on
the
first
time,
I
need
to
set
up
I'll
just
create
a
couple
columns
to
do
and
doing
and
done
and
I'll
just
add
that
new
issue,
and
then
let's
kind
of
a
developer,
I'm
gonna
come
in
and
say
alright,
let's
mark
that
to
doing
and
I'll
start
work
and
that
lets
everybody
know
that
it's
now
being
worked
on
now,
of
course,
to
start
working
on
it.
I'd
obviously
have
to
start
writing
some
code.
C
So
I'm
gonna
show
something
that
it's
actually
pretty
cool
I've
got
these
environments,
since
I
did
set
up
Auto
deploy
this
actually
automatically
deployed
to
staging
for
me
and
I
can
click
this
little
button
here
and
I
will
SSH
into
that
staging
server
and
have
full
access
to
all
the
content
there
and
wouldn't
generally
recommend
doing
this
for
real.
This
is
more
for
like
debugging
purposes,
but
it's
also
great
just
for
hacking
around
and
testing
out
some
things,
and
so
here
I'm
just
going
to
make
a
little
change.
C
We
just
took
the
server
now:
I
can
click
on
this
little
button
there
that
actually
pulls
me
to
the
staging
server
and
I
can
see
that
my
change
is
live
now.
It's
live,
but,
of
course
it's
not
in
source
control.
So
what
we
really
want
to
do
here
is
go
back
in
and
put
this
into
source
control,
so
go
back
to
that
same
file
and
click
Edit
from
here
and
make
the
same
change.
But
now
I'm,
just
gonna
go
totally
radical
and
turn
it
into
a
header.
C
Now,
remembering
that
we
have
we're
all
doing
this
to
work
on
an
issue,
I'm
gonna
actually
assign
this
to
a
topic
branch
named
with
the
vert.
With
the
issue
number,
it
was
a
shoe
number
one
and
then
that
helps
us
that
when
we
go
to
create
a
merge
request,
it
knows
that
it
automatically
it'll
close
issue
number
one.
So
I'll
just
go
ahead
and
finish
creating
that
merge
request,
and
we
see
that
now
that
we
got
this
merge
request.
It
actually
automatically
kicked
off
a
pipeline.
C
I
see
I,
see
the
pipeline
dive
in
a
little
bit
further
here,
and
we
see
it's
a
pretty
simple
pipeline
right
now.
It's
a
build
review
and
clean
up
the
build
step
is
creating
a
docker
image,
pretty
straightforward
that
finished
and
review
it
takes
that
docker
image
and
sends
it
up
to
a
this
awesome
o
thing:
a
review
app.
Basically,
it's
a
temporary
ephemeral
application,
that's
running
in
a
production
like
environment,
so
you
can
actually
see
your
changes
and-
and
we
see
if
it
go
back
to
the
merge
request,
we
get
this
nice
little
notice.
C
C
This
is
really
really
great
because
you
know,
as
a
viewer
I
don't
want
to
just
trust,
reading
the
code
and
having
to
interpret
it
specially
if
it's
a
CSS
change
or
something
that's
really
hard
to
visualize
and
obviously
I
could
pull
the
code
down
locally
and
test
it
that
way,
but
especially
for
like
product
managers
or
designers,
other
folks
that
just
don't
to
bother
even
developers,
they
just
want
to
bother
pulling
the
code
down.
You
can
go
and
take
a
look
at
it,
so
it's
really
really
valuable
that
way.
But,
of
course
we
do
sport.
C
C
And
you
know
that
will
show
up
in
this
discussion
thread
and
anybody
who's
on
there
would
get.
You
know
an
email
notification
and
all
that
kind
of
great
stuff.
So
so
it
looks
all
good.
The
code
looks
good.
The
review
up
looks
good,
we're
gonna,
go
ahead
and
merge
this,
and
now,
if
I
go
back
to
the
the
pipelines
page
here,
I
see
that
it's
kicked
off.
C
Another
CSU
D
pipeline
not
just
for
the
topic
branch
but
for
master
and
because
of
course,
has
been
merged
in
master,
and
so
it's
rechecking,
the
CI
make
sure
that
everything's
still
good,
but
also
it's
going
to
go
and
publish
this
directly
to
our
staging
server,
which
we
already
had
before.
But
now
it's
going
to
update
that
staging
server.
C
Actually
one
of
the
really
cool
things
if
I
go
back
to
this
merger
quest,
even
though
it's
already
been
merged,
we
now
know
that
it's
no
longer
deployed
to
review
because
that
app
has
been
destroyed
because
the
branch
has
been
destroyed.
But
now
it
has
been
deployed
to
staging
and
I
can
click
directly
from
there
and
see
that
staging
has
now
been
updated,
which
really
brings
everything
full
circle
to
the
developers
working
on
the
merge
quest.
C
They
can
see
everything
and
have
visibility
into
when
this
goes
into
staging
when
it
goes
into
production
and
all
that
kind
stuff.
So,
speaking
in
production,
let's
go
ahead
and
ship
this
to
production
I
could
do
that
in
a
couple
of
different
ways.
You
know
you
can
go
into
the
environments
tab
here
and
I
can
click
on
this
and
deploy
from
whatever
is
in
staging
into
production
I'm
going
to
pipelines.
C
It
can
do
a
bunch
of
different
things,
but
let's
go
back
to
chat
and
kind
of
complete
the
circle
here
and
one
of
the
great
things
about
chat
upstairs
to
do
everything
in
the
clear
so
people
know
what's
going
on,
and
here
I
can
say
I'm
going
to
exploit
whatever
is
in
staging
deploy
it
to
production,
and
now
everybody
will
get
that
notification.
I
click
on
the
link
and
I
can
see.
C
I
know
the
deploy
happening,
live
it's
really
fast
cuz,
it's
just
pushing
up
the
docker
image,
but
it
is
done
and
and
again,
if
I
go
back
to
that
merge
request.
I
can
see
that
now
it's
running
in
staging
and
production
and
if
I
click
on
that
link,
actually
sometimes
I
don't
have
my
readiness
probes
set
up
properly
yet
so
the
so
say
the
deploy
is
done
before
it
actually
finishes.
So
there
goes
the
good
finish
level:
it's
nice
yeah.
C
So
now
it's
live
in
production
all
the
way
and
and
which
is
pretty
darn
cool
couple
of
all
the
great
things
now
that
I'm
on
the
environments
tab
here.
This
is
a
new
feature
we
just
launched.
In
nine
point,
oh
I'm,
pretty
proud
of
I
can
now
see
the
deploy
board,
which
is
really
just
showing
me
some
the
status
of
all
the
incident.
Sorry,
the
pods
in
this
case
on
communities
right
now,
there's
just
one
pod,
but
if
I
go
over
to
I'm,
gonna
cloud
share
a
show
here
up
and
I'm
gonna
scale.
C
B
D
I
did
not
know
that
this
is
super
awesome
and
I
will
have
to
we'll
have
to
figure
this
out
very
cool.
Is
there
anyone
else,
41.7
Ivar?
No,
you
had
one
more
thing
before
we
get
into
the
new
features
work.
Is
there
anyone
else
who
had
any
other
1.7
updates
other
than
a
call
for
a
release,
see
I
think.
D
D
D
D
G
I
will
show
the
actual
document
that
have
created
it.
We
have
generated
the
CPM
medium
that
happened
last
Tuesday,
so
this
document
describes
the
a
global
vision
of
how
should
we
enhance
our
feature,
submission
process
and
feature
feature
workflow
for
just
a
releases
startin
from
1.7.
This
work
has
been.
G
This
document
has
been
included
in
like
lots
of
work.
It
has
been
done
during
the
last
time
and
it
collected
our
experience
for
again
from
there
one
of
six
releases
and
prioritizes,
so
please
feel
free
to
review
this
dog.
We
have
a
great
on
this
version
at
the
6:00
p.m.
meeting.
At
the
same
time,
we'd
like
received
a
global
community
feedback.
Regarding
this
also
one
major
note
that
baba's
has
highlighted
your
insignia
meeting
that
this
proposal,
this
procedure
should
not
be
under
for
the
features,
but
for
any
major
in
significant
change
to
Panetta's.
G
So,
if
we're
speaking
about
some
major
bug
fixes-
or
something
like
that,
so
which
it
also
starts
to
consider
this
procedure
for
that,
but
that's
not
a
question
for
our
upcoming
governance
and
government
committees.
We've
also
discussed
there's
also
I
have
created
the
PR
for
with
a
new
template.
This
parrot
has
actually
been
created
with
proximate
one
months
ago,
but
before
because
of
huge
lot
of
people
with
1.6
and.
G
Non-Existence
of
the
newer
features
features
procedure.
We
haven't
discussed
that
with
more
details,
so
I
have
to
updated
that
with
the
latest
situations
from
the
community,
so
please
review
it
and
Olivia
Colin
stirrer
also
would
like
to
receive
the
final
suggestions
and
final
comments
on
those
items
on
PR
and
on
the
document
for
Monday
and
we
expecting
to
Google
I/os
these
two
items
on
Monday.
A
H
Hi,
can
you
hear
me?
Yes,
we
can
hear
me
good
yeah,
so
I
said
I'd
give
an
update.
We
had
a
face-to-face
meeting
in
Berlin,
which
was
very
well
attended.
I
thought
about
40
people
showed
up,
and
so
it
was.
It
was
quite
an
eye-opener
to
me
that
the
there's
a
lot
of
people,
you
know
building
things
around
subin's
networking
at
CNI
that
that
don't
kind
of
speak
up
so
much
in
the
slack
or
wherever
they
just
kind
of
get
on
with
it.
H
In
the
nature
of
these
discussions,
when
you
get
a
bunch
of
people
who
are
kind
of
very
own
degrees,
a
lot
of
them
are
coming
up
to
speed.
So
we,
the
network
policy,
discussion
I,
was
in
and
we
basically
spent
most
of
time
kind
of
agreeing
what
the
terms
were
pretty
much.
Everyone
wants
an
egress
rule
that
targets
a
cider
that
was
a
that
was
a
fairly
consistent
point
and
then
there's
a
lot
of
detail
around
you
know:
do
we
actually
want
different
roles?
H
You
know
like
a
developer,
can
make
some
kind
of
a
network
policy,
but
an
administrator
can
override
that
and
that
gets
into
a
whole
bunch
of
other
things
that
are
kind
of
beyond
the
basic
subject
of
networking.
That's
those
things
are
going
to
have
to
be
solved
before
people
are
really
happy
with
the
what
we
do
there.
H
Multiple
interfaces-
I
was
not
actually
personally
in
that
discussion,
but
the
I
believe
the
interest
is
in
things
like
if
I
have
one
network
that
I
want
to
do
my
application
traffic
on
and
another
network
I
want
to
talk
to
my
storage
array
on
or
something
like
that.
Then
right
now,
kubernetes
only
supports
one.
H
You
run
multiple
kind
of
CNI
plugins,
underneath
that
and
and
it
kind
of
segues
into
the
other
topic,
which
is
ipv6
because
ipv6,
you
kind
of
end
up
with
not
exactly
two
interfaces
but
often
two
addresses
and
ipv4
in
an
ipv6
address,
and
we
kind
of
have
to
but
most
of
the
plumbing
into
kubernetes,
to
accommodate
those
things
for
ipv6
and
and
then
it
sort
of
follows
through
that
you'll
have
more
of
the
plumbing
available
from
multiple
interfaces.
So
what
did
people
say
about
ipv6?
H
Yeah
it's
it's
kind
of
coming
along
and
there's
a
few
bugs
a
few
things
that
have
worked
properly.
If
you
try
it,
but
it's
nearly
usable
and
of
course,
in
I
P
V
six
world,
most
people
don't
want
an
overlay
network
or
anything
like
that.
They
just
want
to
use
the
kind
of
global
irritable
nature
of
ipv6.
H
I
H
I
H
I
J
Okay,
any
more
questions:
Brian
I'm,
not
joking!
Is
there
any
publicly
facing
dashboard
that
folks
are
interested
in
maintaining
or
creating
with
regards
to
testing
against
different
versions
of
kubernetes
and
different
potential
deployment
scenarios,
because,
as
some
folks
might
not
may
or
may
not
be
aware,
some
of
the
issues
around
our
back
and
C&I
plugins
was
an
issue
for
one
six
release.
I
Yes,
so
what
we've
been
doing
is
we're
starting
to
spin
up
a
lot
more
variations
with
the
repo
that
we
have
called
the
Taconic
installer
and
we'll
be
putting
those
into
the
test
grid,
trying
to
try
out
essentially
a
bunch
of
the
different
daemon
set
to
play
well
networking
options.
So
that's
something
that
we're
thinking
about
and
I
do.
As
somebody
pointed
out,
we
do
intend
to
put
them
at
the
test
grid
to
at
least
make
them
visible.
K
In
addition
to
that,
we'll
be
in
working
with
Microsoft
and
there's
going
to
be
a
lot
of
announcements
at
dr
cohn
next
week,
coming
from
Redmond
around
native
overlay
drivers
for
for
Windows,
Server
2016
and
our
goal,
there
is,
to
you
know,
creating
a
partnership
between
tigrerra
cloud
based
on
the
ouvea
nano
vs
space,
as
well
as
Microsoft,
to
produce
CNI
plugins
on
by
utilizing
the
networking
overlay
drivers
for
Windows
2
to
produce
a
comprehensive
networking
strategy
for
us.
So
that's
basically
what
she
has
been
mainly
focusing
lately
and
that's
been
consuming.
F
F
Okay,
can
everybody
see
this?
Yes,
yes,
okay,
so
this
is
a.
This
is
a
tracking
matrix
for
the
site's
reorganization
and
a
deprecation
of
the
kubernetes
guides.
Section
guides
previously
held
I
would
say,
probably
70
to
80
percent
of
the
hand,
authored
content
on
kubernetes,
I/o
and
by
hand
authored
I
mean
actual
markdown,
that
people
wrote
not
reference
documentation
and
because
of
that
it
had
a
lot
of
nested.
It
had
some
deeply
nested
tables
of
contents.
F
We've
gotten
a
lot
of
this
stuff
through
to
its
new
homes.
Only
a
couple
of
stragglers
remain
and
we'll
start
editing
it
in
place,
and
we
hope
to
have
this
effort
completed
before
the
1.7
launch.
This
means
that
there'll
be
some
dok
dok
implications
for
the
1.7
launch.
For
those
of
you
who
edit
docs
for
new
features,
you
may
find
that
what
you're
used
to
editing
is
not
in
the
place
where
you
thought
it
was
or
not
in
the
place
where
it
used
to
be,
if
you're
looking
for
a
place
for
new
documentation.
F
If
you
don't
normally
write
Docs
but
you're,
looking
for
the
place
where
your
documentation
should
go
for
1.7,
it's
not
gonna
be
in
guides
anymore.
Guides
is
deprecated
our
release,
our
Docs
release,
person,
andrew
chen
and
also
myself
and
also
steve
perry
will
be
available.
We
want
to
get
involved
in
the
1.7
release
early
to
help
people
find
out
where
their
content
should
fit
in
the
new
structure.
What
kind
of
tasks
documentation
tutorials
or
concepts
you'll
need?
F
So
as
we
pick
the
release
team
41.7,
those
people
that
volunteer
should
be,
hopefully
they
can
work
with
us
and
the
actual
docs
content
providers
to
figure
out
where
their
new
content
should
go.
I
would
prefer
to
do
that
earlier,
waiting
until
after
code,
freeze
or
the
last
week
or
so
of
crunch
before
the
release
goes
out,
would
not
be
ideal.
In
this
case,.
D
F
More
difficult
to
do,
given
our
current
hosting,
because
while
we
maintain
separate
branches
for
each
doc
for
each
for
each
release,
github
agent
is
our
hosting
mechanism.
Only
lets
us
serve
one
at
a
time
unless
we
do
some
crazy
stuff
with
cloning
release
branches,
two
different
repos
and
that
has
seriously
mucked
up
our
search
engine
optimization
and
our
search
engine.
F
The
short
answer
is
we're
currently
not
versioning
documentation,
even
though
we're
maintaining,
where
we're
containing
the
branches
for
the
previous
releases,
and
if
we
want
to
look
at
hosting
multiple
versions
for
the
actual
hand,
authored
documentation,
we're
going
to
have
to
look
at
changing
the
hosting,
we
still
have
version
documentation
for
all
the
automated
auto-generated
reference
documentation.
So
for
API
Doc's
for
cute
control/command
references.
F
D
B
B
D
B
G
It's
one
more
note
regarding
Cuban:
it
is
a
deterministic
servant,
as
you
mentioned
it
has
been
announced
before,
but
during
this
week
we
figure
finally
polish
the
schedule.
So
if
you
follow
the
link
that
is
attached
to
to
the
community
notes,
you'll
find
a
schedule
of
oxygenators
de
austin
so
feel
free
to
attend
them
if
you'll
be
in
Boston.
Today's.
D
D
Only
updates
on
governance,
which
I'm
happy
to
share
our
that
we
have
three
dates
scheduled
and
blocked,
and
we
are
getting
our
plane
flights
together
for
the
seven
people
that
are
going
to
answer
our
four
questions
and
then
implode
themselves.
No
just
explode
the
group,
the
the
self,
the
self
limiting
governance
group
of
seven
is
leading
on
April
28
and
May,
8
and
9,
and
with
those
three
days
we
hope
to
have
answers
for
four
questions
and
bringing
something
back
to
the
community.
We
may.