►
From YouTube: Kubernetes SIG Windows 20180717
Description
Kubernetes SIG Windows 20180717
A
A
B
B
C
Okay,
I
guess:
I'm
I'm,
Ben
moss
I've
been
working.
I'm
working
for
the
company,
pivotal
I
was
looking
to
get
we're.
Exploring
overlay
networking
on
Windows
and
I've
been
trying
to
work
through
what
resources
are
out
there
and
you
know
definitely
looking
for
help.
If
anyone
else
out
there
has
had
any
luck
with
it.
So
far,
I've
yeah
not.
C
D
No
so
my
my
guide
host
gateway
mode,
excellent,
so
and
right
now,
flannel
on
overlay
or
overly
support
and
final
easel
using
the
excellent,
is
still
really
experimental
and
has
a
lot
of
limitations.
So
we
still
they're
still
like
development
active
aldeman
going
on
there
just
to
be
able
to
support
it
or
not.
The
functionality.
D
C
A
D
That's
right
so
with
Kubb,
1.11
I
think
there
were
some
API
changes
that
caused
my
guide
to
break.
Unfortunately,
I
did
add
a
PR
with
fix
for
dad
I'm
still
kind
of
testing
if
it
works,
I
did
get
it
to
work
a
couple
times,
but
I
wanted
to
make
sure
that
it
does
work
like
it's
stable
before
I
merge
it,
and
otherwise,
if
you
use
like
older
1.10
few
binaries,
the
guide
should
still
work.
Okay,
thank
you.
I'll
be
looking
forward
to
that.
Those
changes
as
well.
Okay,
so
I
can
I
can
add.
D
D
A
B
Hi
I
already
go
thingie,
see
sangeun
I
think
we
talked
last
week
when
we
were
speaking
about
the
DNS.
Is
it
possible
to
move
out
the
h,
NS
metal
creation
outside
of
the
CN
I?
Think
you
said
that
you're
going
to
lock
up
if
you
have
some
kind
of
provisioning
for
PCs
I'm,
not
sure
if
you
had
time
for
that,
Patrick.
A
A
A
B
And
also
one
more
question:
when
you
enable
extension
over
the
Janus
Network,
it
actually
does
not
get
reported
back
as
being
enabled
and
as
far
as
I
remember
Dinesh
said,
that's
fixed
in
RS
5.
Is
it
possible
to
back
port
it
back
to
1803
because
it
still
not
working
1803,
so
the
extension
is
running,
but
it
says
that
it's
enabled
false.
B
A
A
A
Filed
two
issues
earlier:
I
guess
as
towards
the
end
of
last
week,
so
I
started
looking
at
my
deployments,
that
I
was
doing
on
on
Azure
and
I
was
seeing
a
lot
of
repetitive
errors
that
were
being
logged.
You
know
like
once,
every
minute
or
two,
and
so
that
was
sort
of
just
one
area,
I'm
going
to
be
looking
at.
A
It
looks
like
there's
some
similar
patterns
in
the
cube
proxy,
but
I'm
a
lot
less
familiar
with
those
as
well.
So
as
you're
doing
you
know,
setups,
you
would
like
to
encourage.
Everyone
to.
You
know,
take
a
look
at
what
sort
of
errors
are
seeing
on
the
system,
especially
for
things
that
are
flooding
the
logs
and
make
sure
to
get
bugs
filed
on
the
kubernetes
repo.
A
For
that,
because
I
think
that,
as
we're
pushing
for
to
exit
out
a
beta
like
if
I
look
right
now,
I
see
26
open
issues
and
it
in
the
repo
but
a
whole
bunch
of
these
are
things
that
have
been
open.
Since
you
know
for
more
than
a
more
than
a
couple
months
there,
and
so
what
I
want
to
do
is
I'll
start
pinging
people
on
some
of
those
older
issues
to
see
whether
or
not
those
are
still
relevant.
A
A
At
some
point,
during
1.10,
we
were
up
up
at
a
point
where
we
had
only
seven
failures
and
we've
sent
some
sort
of
backtracked
a
bit
on
that
with
the
with
the
way
that
we're
setting
things
up
so
the
way
the
tests
are
running
right
now,
breezing
a
tool
called
ACS
engine
and
I
I.
Guess
it's
not
really
the
best
comparison,
but
if
you've
ever
seen
how
cops
it
deploys
things
on
AWS.
You
know
where
automates
setting
up
the
VMS
getting
the
cubelet
configure
and
all
of
that
acs
engine
is
a
little
bit
similar.
A
A
A
D
A
A
D
A
D
A
D
D
The
two
consistent
consisting
of
two
pods
encapsulated
by
a
service
and
I,
had
to
like
pseudo
clients,
client
pods
that
are
interacting
with
it,
and
the
second
test
I
had
was
just
having
two
pods
scheduled
or
I.
Think
I
had
for
like
two
on
each
node
I'm,
just
just
to
try
out
like
four
at
once,
and
they
were
trying
out
external
DNS
resolutions
to
like
bing.com
every
a
couple
minutes.
Every
three
minutes.
A
D
A
very
basic
test
so
to
speak,
but
I
wanted
to
see
if
I
could
reproduce
these
issues
that
a
lot
of
people
have
been
reporting
where
DNS
resolution
seems
to
fail
after
specific
number
of
days
of
the
pods
running
and
I
had
tried
out
for
about
nine
days
and
I
couldn't
find
any
issues
or
reproduce
problems.
No.