►
From YouTube: Kubernetes SIG Azure Community Meeting 9-6-2017
Description
A
A
So
please
feel
free
to
go
there
and
see
who
attended
and
what
we
spoke
about
and
as
well
as
finding
video
links
to
prior
meetings
and
without
further
ado,
we're
gonna
go
ahead
and
press
through
the
agenda.
This
should
be
a
pretty
quick
meeting
if
there
is
something
that
you
would
like
to
speak
about
in
terms
of
PRS
that
need
attention
or
issues
you're
aware
of
or
any
other
things,
please
put
those
in
the
agenda
and
we
won't
get
to
those
forthwith
so
moving
right
along
work
in
progress.
A
If
anybody
wants
to
participate
in
that,
please
make
a
comment
in
the
issue
and
stay
tuned
for
PRS
or
reference
that
issue.
There
will
be
a
lot
of
work
done
there,
so
this
gives
us
a
chance
to
sort
of
have
a
Greenfield
opportunity
to
to
add
things
that
we
wanted
to
or
reef
actors
to
the
the
cloud
provider
and
because
the
implementation
plan
is
to
have
this
be
in
parallel
to
the
entry
provider.
A
This
gives
us
a
chance
to
a/b
the
functionality
in
a
way
that
would
have
been
impossible
before
so,
as
a
community
I
think
we
have
a
really
beautiful
opportunity
here
to
improve
that
functionality
and
I
highly
suggest
we
take
advantage
of
it
and
make
the
most
of
it
so
anyway.
That
is
the
issue
that
we'll
be
tracking
to
work
and
stay
tuned
on
that
as
we
move
forward,
PR
is
needing
attention
and
review.
There's
an
aging
did
a
new
PR
for
this.
Oh.
B
A
Gonna,
because
we're
in
tree
already
that
work
will
be
probably
duplicated
in
entry
and
out
of
tree,
and
so
there
will
be
a
time
there
where
you
actually
have
to
support
both,
and
that
means
that
if
we
patch
one
we
have
to
patch
the
other
and
in
the
begin
of
a
Greenfield
approach
to
the
out
of
tree
provider,
that
will
make
it
tricky,
because
there
could
be
an
instance
where
a
code
fix
on
one
does
not
necessarily
apply
to
the
other.
So
that'll
be
a
relatively
high
management
burden
for
a
period
of
time.
A
A
A
This
was
actually
in
response
to
having
to
move
this
stuff
out
of
core,
because
the
the
path
that
it
was
in
was
deprecated.
So
this
did
get
some
reviews
before,
but
it
would
be
good
to
have
just
a
few
more
eyes
on
this.
If
possible,
I
did
look
at
it,
but
I'm,
not
the
most
qualified
reviewer,
so
I
would
really
appreciate
it
if
somebody
else
could
just
run
through
these
and
make
sure
they
make
sense.
So
that
would
be
great
there's.
This
oh
go
ahead.
A
A
There's
some
been
some
changes
requested
and
I
want
to
verify
that
those
changes
were
implemented
properly
again,
I'm,
not
the
right
person
to
review
this
code,
but
if
somebody
here
wants
to
look
at
that,
that
would
be
great,
also
paste.
This
link
in
the
chat
in
case
you
don't
have
to
agenda
open
any
questions
about
that.
A
B
B
A
B
Don't
have
a
lg
tv,
LG,
TM,
writes
on
anything
and
kubernetes
I,
just
review
and
comment
and
so
on.
If
you
scroll
down,
if
you
scroll
down
to
lost
meeting
notes,
I
actually
updated
them
was
was
was
all
the
stuff
that
I
said,
I
will
take
care
of
long
time.
Ok,
all
right!
So
like
if
you
long.
If
you
scroll
down
to
the
last
meeting
notes,
you
will
see
like
that
that
there
were
some
reviews
requested
abused,
requested
on
a
lot
of
things.
B
A
So
of
great,
then
we'll
get
this
on
track
for
approval,
then
for
one
night
in
fact,
what
I'll
do
is
I'll?
Add
it
to
the
one
night
milestone
now,
because
it's
not
going
to
go
to
180,
okay,
cool.
Thank
you
come
I
and
there's
something
else.
So
cal
are
you
actually
are
you
remember
the
kubernetes
or
one
in
a
zone?
No.
B
A
B
A
B
A
B
C
B
Remember,
I
read
the
description
of
the
scheduler
issues
that
we
have
around
disks
and
types
and
the
end.
Yes,
I
want
to
open
this
discussion.
This
is
not
the
right
forum
for
it,
but
we
at
least
need
to
start
having
these
discussions.
I,
don't
Kelly
Kelly
I'll
just
go
ahead,
and
do
it
all
right.
So
I
need
to
put
lackeys
like
into
this,
because
he
is
that
the
best
guy
in
terms
of
user
experience
will
come
to
see
things.
Yeah
agree.
B
No
other
than
just
please,
please,
please,
please
take
a
look
at
the
azure
channel
on
slack.
Yes,
there
are
a
bunch
of
questions
that
can
easily
be
answered
and
get
people
unlocked
some
questions
around
how
things
work
question
around.
How
can
I
do
this
and
that
can
be
answered
as
simply
as
two
liner
or
a
reference
to
sun
block
and
it
greatly
so
I'm,
trying
like
at
least
I,
spend
them
a
total
of
two
hours
a
week
or
an
hour
a
week
contest,
so
just
I've
treated
people
go
and
understood.
A
Yes,
that
isn't
the
announcements
and
I
100%
agree.
I
was
actually
helping
that
guy
mark
last
night,
try
and
diagnose
why
he
was
having
issues
with
no
deaf
grades
looks
like
he
might
have
uncomfort
uncovered
some
sort
of
incompatibility
with
darker
versions
and
175,
so
we're
gonna
have
to
figure
that
out.
Well,
just
what
we
need
so
yeah
I
echo
what
you
said
in
the
more
eyes
on
that
the
better.
Also
Stack
Overflow
has
a
ton
of
questions
all
the
time
for
us
that
we
can
answer.
A
C
A
A
Okay,
so
just
a
quick
update
about
where
we're
at
in
the
various
release
cycles
we're
gonna,
try
and
cut
one
dot
beta.
One
today
today
is
also
the
first
burndown
meeting,
and
it's
gonna
happen
in
about
an
hour
an
hour
from
now
or
45
minutes
from
now.
If
you
want
to
be
involved
in
that
as
a
spectator,
I
can't
imagine
why
somebody
would,
but
you
may
want
to
there-
is
a
crew
Nettie's,
burned
down
meetings.
Google
group
that
you
can
join
and
you
will
get
an
invite
to
the
meeting.
A
I
can
also
add
you
directly
since
I
on
the
calendar,
and
that
would
be
great
to
have
people
just
interested
in
the
release
process.
You
might
want
to
be
involved
in
1-9.
As
that
happens,
and
just
see
what
happens
behind
the
scenes.
We
are
basically
on
track.
There's
a
pretty
high
level
of
churn
around
testing
right
now.
For
example,
the
cops
AWS
test
is
completely
broken.
There's
a
packin
janitor
job
that
handles
all
the
cleanup
for
those
things
is
also
broken.
A
Kubernetes
cross
was
broken
yesterday
because
that
somebody
changed
an
nth
type
and
and
exceeded
the
value
of
the
of
the
allocated
memory,
and
so
there's
lots
of
like
constant
things
happening
with
a
semi,
so
make
you
pretty
much
on
a
daily
basis,
but
we're
trying
to
work
through
those
we're
also
working
on
getting
the
1.8
upgrade
tests
and
the
the
current
version
ski
of
policy.
It
really
applies
for
two
versions,
so
this
will
be
one
6,
+,
1
7.
So
basically
the
1
5
path
to
1.8
will
be
deprecated
just
for
everybody's
knowledge
and
notification.
A
D
A
It's
I'd
see
it
mainly
because
I'm
in
all
the
sig
groups,
so
sig
testing,
like
channel,
has
usually
something
in
it.
People
are
also
raising
awareness
and
sig
release
and
also
when
I
do
I'm
doing
all
these
managing
of
PRS
Aminata
I
see
them
there.
So
I,
don't
know
what
the
best
way
is.
I
mean
you
can
look
at
test
grant
and
whatnot,
but
I
just
find
that
they're
just
looking
at
the
the
regular
channels
to
be
the
way
to
do
it.
Ok,.
A
So
ok,
once
7
&
6,
there's
no
plan
out
yet,
but
I
miss
anticipating.
That
will
probably
happen
next
week.
1
6
10
doesn't
have
a
plan
yet
same,
probably
next
week
or
the
week
following
a
CNS
engine,
zero
six
zero
is
out.
I
haven't
heard
anything
about
zero.
Whatever
would
be
the
next
probably
zero,
seven
zero,
but
anybody
from
the
ACS
so
I'd
want
to
speak
to
that.
A
Alright,
so,
lastly,
announcement
inside
we
already
hit
these
thanks
Cal
for
hitting
this
announcement
early,
but
we
need
more
eyes
on
stackoverflow
and
specifically
the
azure
signature,
channel
and
slack,
and
if
you
haven't
looked
at
this
pull
request
for
the
RFC
proposal.
Please
do
it,
so
it
will
change
the
way
things
are
done.
This
isn't
an
official
announcement,
but
it's
something
you
all
should
probably
know
about
is
it
there
is
a
movement
afoot
that
will
probably
happen
in
sega
architecture
to
basically
gate
all
of
the
contributions
into
master.
A
So
essentially
we're
gonna
probably
move
to
a
non
dirty
master
model,
which
means
we
will
have
a
branch
that
gets
merged
in
at
release
time
as
opposed
to
committing
to
master.
So
that
is
a
relatively
humongo
change
in
procedure
and
process.
It'll
probably
have
very
wide
far-reaching
impacts,
but
Brian
grant
is
probably
going
to
be
the
person
driving
that
so
it
you
have
a
stake
in
that
race
and
want
to
be
involved
in
the
discussion.
Please
attend
Cir
sake,
architecture
meetings,
which
are
a
Thursday
at
9:08,
8:30
a.m.
PST.
A
There
is
not
one
this
week,
but
the
following
week.
You
can
also
join
kubernetes
sick
architecture
and
Google
Groups
and
get
the
information
there.
These
are
things
you
know
it's
hard,
I
I
go
to
all
I
could
probably
go
to
more
sick
meetings
than
anybody.
I,
don't
know
anybody
who
goes
to
more
than
I
do
for
better
for
worse
I
and
there.
There
are
a
lot
of
things
that
seem
peripheral,
but
when
they
emerge
in
the
in
the
sort
of
mainstream
discussions
it
can
be
really
impactful.
A
So
I
will
try,
in
the
sake
to
to
bring
those
up
for
people
who
who
are
interested
in
learning
more
and
getting
involved
in
other
areas
and
other
sakes.
So
I
would
say
say:
your
architecture
is
definitely
something
we
should
look
at.
Another
interesting
thing
for
this
group,
specifically,
is:
there
is
a
lot
more
happening
in
sig
Windows
now
than
there
has
been
I
know.
The
past
sick
windows
was
kind
of
flagging
a
bit
but
Jason
Messer
Microsoft
has
now
got
three
FTEs
dedicated
to
upstream
work.
A
There
is
a
new
version
of
Windows
coming
out
that
is
actually
going
to
support
multiple
containers
for
per
pod
and
they
will
share
all
the
things
that
you
should
expect.
They're
also
trying
to
find
parity
in
terms
of
feature
richness
in
in
the
the
container
ecosystem
for
Windows
a
lot
of
good
stuff
there,
and
it's
going
to
mean
that
eventually
we're
looking
at
true
parity
between
Windows
based
nodes
and
Linux
based
nodes.
A
So
if
you
haven't
attended
cig
windows
in
the
past
because
it
didn't
seem
like
it
was
happening,
it
is
starting
to
happen
and
you
can
see
the
agenda
and
notes
that
bitly
slash
cig
windows,
if
you're
just
curious
and
catching
up
I,
don't
really
have
any
other
announcements
per
se.
Anybody
have
anything
they
want
to
add.