►
From YouTube: Kubernetes SIG Windows 20210316
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
Hello,
everybody
and
welcome
to
the
march
16
2021
iteration
of
the
sig
windows
kubernetes
community
meeting.
As
always,
these
meetings
are
recorded
and
uploaded
to
youtube
so
be
sure
to
adhere
to
all
of
the
cncf's
code
of
conduct.
All
right.
Let's
get
started,
let's
open
this
up
to
kind
of
some
introductions.
If
there's
anybody
who's
new
on
the
call
who
wants
to
kind
of
introduced
themselves
and
ask
any
questions,
I
see
a
couple
of
new
people
on
the
call
today,
but
no
no
obligations
to
do
so.
B
B
I'm
really
excited
to
get
more
involved
in
the
sig
and
have
some
chats
with
you
and
members
here
and
see
what
I
can
do
to
help
out.
C
B
You
welcome,
thank
you.
No,
so
what
I
wanted
to
do
was
take
my
skills
with
kind
of
windows
and
linux
interoperability
and
get
more
involved
in
the
cloud
native
kubernetes
space,
and
this
opportunity
opened
up
at
rancher
and
I'm
super
excited
about
it
and
we're
building
out
a
team,
and
we've
got
some
cool
ideas
and
yeah
we'll
take
it
from
here.
A
All
right
that
stream
might
be
a
good
opportunity
to
discuss
kind
of
what
the
topic
that
jay
was
added
last
week,
but
we
got
delayed,
so
it
might
be
a
good
meeting
for
you
to
have
here.
Is
there
anybody
else?
If
not
we
can
go
through?
I
can
just
quickly
go
through
a
couple
of
introduce
the
leads,
since
there
is
a
bunch
of
new
folks
here
again
yeah
I'll.
A
Do
that
I'm
mark
I'm
an
engineer
in
in
azure
at
microsoft
and
I'm
currently
the
co-chair
of
sig
windows,
james
or
jay.
Do
you
want
to
introduce
yourselves
real,
quick.
C
Sure
I'm
jay
and
I
work
with
mark
and
james
on
some
stuff,
currently
focusing
a
lot
on
andrea
cni
windows,
network
policy,
type
stuff
and
the
overall
cluster
api
kind
of
user
experience.
A
Yep
all
right
cool
thanks
since,
unless
anybody
else
wants
to
introduce
themselves,
we
can
go
on
in
terms
of
announcements.
I
don't
think
that
there's
any
big
ones.
I
believe
that
test
freeze
is
the
24th
I'll
have
to
double
check
that,
but
that
might
be
an
important
date
for
folks
here.
C
A
A
Is
the
I
believe
it's
the
24th?
I
need
to
double
check
on
that,
so
we've
passed
code
freeze
at
this
point.
There
is
a
documentation
or
a
deadline
today
to
have
documentation,
open
up
or
prs
open
up,
but
I
believe
that
that
only
applies
to
new
or
advancing
features
where
they
like
to
see
placeholders.
They
don't
need
to
be
complete
yet,
but
they
want
to
know
who's
going
to
be
responsible
for
opening
up
the
pull
requests
for
the
documentation
updates,
so
they
can
ping
people
if
they
need
to.
A
Does
anybody
else?
Have
any
announcements
at
this
point,
I
think
we're
getting
kind
of
late
in
the
release
cycle.
So
there's
not
a
whole
lot
of
kind
of
important
dates
left
here.
A
E
So
sorry
for
disturbing
yeah,
I
just
have
a
question
regarding
any
bending
brs,
so
I
have
one
pr
which
yeah
was
stuck
because
of
the
milestones
of
1.1
milestone.
So
what's
the
process?
Look
like
yeah
going
after
that?
When
does
it
open
and
do
I
need
to
do
any
manual
intervention.
A
Is
it
if
it's
labeled
sig
window,
if
it
has
the
label
sig
windows
on
it,
we'll
probably,
we
should
pick
that
up
when
we
do
the
backlog,
refinement
that
depends
on
how
the
pr
was
labeled.
Do
you
know
if
it
was
a
feature
or
if
it
was
labeled
as
a
feature
or
a
bug
if
it
was
labeled
as
a
bug,
there's
still
a
chance
to
get
it
into
the
121
release?
A
But
if
it's
labeled
as
a
feature,
then
it's
probably
going
to
have
to
wait
or
wait
until
122
opens
up
and
for
that
process
it's
usually
pretty
early
in
the
that
the
milestone
or
the
in
that
release,
though,
either
release
t
or
team
or
members
of
sig
release
will
kind
of
declare
send
a
mail
to
the
kubernetes
developers
when
they're
allowing
those
types
those
new
pr's
in,
but
for
the
last
couple
of
releases
they've
been
trying
to
change
things
up
a
little
bit.
A
So
there
hasn't
really
been
a
consistent
process
for
that,
but
either
in
120
or
121.
They
had
kind
of
a
phased
approach
where
certain
like
bugs
and
test
fixes,
were
allowed
to
go
in
before
the
before
any
kind
of
new.
Just
code
was
allowed
to
go
in,
but
that'll
all.
E
E
Marked
as
a
bug,
so
I
added
it
on
the
day
when
the
freeze
started,
but
the
issue
is
yeah,
it
requires
adding
a
milestone
label
on
it
and
when
I
try
to
do
so,
it
mentioned
yeah.
Only
the
release
management
group
is
the
one
allowed
to
add
milestone.
1.1
label.
A
Okay,
in
that
case,
could
you
either
post
it
in
slack
and
my
apologies,
if
you
already
did,
I
haven't
been
on
slack,
I
was
out
of
office
yeah.
A
The
the
chairs
and
tech
leads
can
add,
milestones
to
it,
so
we'll
take
a
look
and
if
it
meets
the
bar,
we
can
try
and
get
that
in
for
this
for
121.
If
it
is
a
bug,
fix.
F
I
I
think,
last
time
mark
you
talked
about
the
kubecon
cfp,
open
until
april
11.
I
know
a
couple
of
folks
are
working
on
a
couple
of
proposals,
just
a
reminder
that
we
need
to
submit
anything
any
interesting
sessions
about
sick
windows
before
oh.
A
Yes,
correct
yeah,
so
the
kubecon
north
america
cfp
is
open.
There's
a
link
in
I'll
I'll.
Just
move
that
up
too.
If
anybody's
interested
in
submitting
a
talk,
there's
a
link
there
and
there's
also
all
of
the
dates
for
when
kubecon
north
america
is
scheduled.
A
I
also
believe
that
it's
still,
if
anybody's
interested
in
attending
kubecon
eu
virtual,
you
can
buy,
you
can
purchase
tickets
for
up
until
I
think
up
until
pretty
close
to
when
qcon
e
virtual
is,
which
is
may
the
first
week
of
may
so,
if
anybody's
interested
in
joining
that,
that
might
be
a
good
opportunity.
F
A
Yeah,
we're
always
open
to
taking
kind
of
advice
and
either
figuring
out
so
each
q
con
each
state
gets
it
used
to
be
two
now
it's
down
to
one
maintainer
track
talk
that
we
can,
if
there's
a
lot
of
enough
kind
of
appetite
from
the
community
to
deep
dive
into
different
issues.
We
can
go
into
that
for
the
q
for
this
year's
kubecon
eu
virtual
we're
doing
a
deep
dive
into
networking
with
david
shot
and
kalya.
F
Yep
one
more
thing
mark,
I
don't
know
if
it's
too
early,
but
usually
at
the
end
of
the
you
know
the
last
release
cycle.
We
start
talking
about
the
next
release
like
122
and
the
major
items
committed
and
if,
like
you
know,
the
whole
sick
windows
is
on
page
on
like
the
the
priority
you
want
to
do
that
today,
or
maybe
some
day
later.
Let's
do
that.
F
F
A
F
A
C
Yeah
so
it's
open
conversation,
but
I
guess
the
only
real
thing
here
was.
I
was
just
thinking
well,
our
old
groups.
We
don't
use
them
anymore
and
I
think
the
real
goal
to
cut
all
the
you
know,
sort
of
like
bike.
Shutting
out
of
the
whole
thing
like
what
the
real
goal
here
is:
let's
get
everyone
engaged.
Let's
give
people
areas,
they
can
own.
Make
mistakes,
try
stuff
become
leaders,
you
know
and
like
that's
the
that's
the
goal.
C
Some
of
these
sub
areas
might
allow
us
to
scale
this
whole
windows
thing
so
that
we
can
keep
up
with
all
the
work
that
we
have,
because
a
lot
of
us
are
super
busy
with
with
things,
and
so
we
don't
have
as
much
bandwidth
to
track
every
little
thing
that's
going
on,
but
maybe
if
we
had
focus
areas,
we
could
make
a
bigger
impact
without
having
to
do
more
work.
That's
kind
of
my
thought,
so
it's
kind
of
a
two-part
conversation,
the
names
of
the
groups
to
me
are
arbitrary.
C
The
thing
that's
really
interesting
to
me
is
who's
interested
in
really
stepping
up
and
and
leading
one
or
one
one
of
these
and
and
and
sort
of
owning,
the
overall
sort
of
interlock
with
the
broader
sig.
If
we
make
new
groups
right-
and
I
think
we
were
talking-
some
people
expressed
some
interest
sort
of
it
seemed
like
this.
C
I
think
arvind
was
maybe
interested
in
some
of
them,
and
I
I
I'm
interested
of
course,
in
the
networking
side-
and
I
don't
know
we
got-
we've
got
other
people
who've
come
on
board.
There's
community
work
to
do.
There's
test
grid
stuff
to
do
so.
A
I
was
going
to
say
we
can,
we
can
go
through
what
some
of
these
were.
I
it's
a
little
bit
strange
like
that.
These
are.
These
are
actually
the
github
repositories
that
we
own
and
they
don't
necessarily
tie
directly
to
different
sub
projects,
but
we
can
certainly
add
some
clarification
or-
or
everything
here
and
set
up
new
github
kind
of
repos
under
kubernetes,
as
we
see
necessary.
C
C
A
So
I
think
that
some
of
them
were
subprojects
and
some
of
them
are
misleading.
I
can
go
over
some
of
these
real,
quick
and
give
kind
of
my
what
I
remember
from
history
here.
A
lot
of
these
are
set
up
by
users
like
by
people
who
are
no
longer
part
of
the
community,
so
the
windows
gmsa-
that
was
I
that
was
a
sub
project
to
support
group,
managed
service
accounts
in
for
windows
workloads.
A
There
were
two
kind
of
big
enhancements
around
this
one
was
gmsa
itself
and
another
one
was
the
adding
that
run
this
username
kind
of
analogs
to
run
as
user
for
for
windows
workloads.
I
believe
that,
right
now
this
is
mostly
a
repository.
I
don't
know
why
that
links
to
the
owner's
file
with
some
documentation
and
also
some
like
an
admission
web
hook.
That's
used
to
as
kind
of
a
sample
for
configuring
group
managed
service
account
functionality
in
windows.
A
This
is
linked
to
from
some
of
the
official
documentation
in
the
kubernetes
website.
So
this
is
something
that
I
don't
think
we
want
to
shut
down,
but
we
can
definitely
either
wind
down
or
kind
of
you
know
say
this
is
in
maintenance
mode.
Here,
let
me
I
can.
D
I
would
definitely
say
we
can:
we
need
to
keep
it
around.
There
is
some
work
that
needs
to
be
done.
Some
of
the
pod
specs
need
to
be
migrated
to
the
latest
versions
and
the
web
hooks
and
that
kind
of
stuff
from
a
maintenance
perspective,
but
for
the
most
part
it's
pretty
stable
at
this
point,
so
I
don't
know
if
there's
anybody
that
wants
to
step
up
and
take
on
some
of
that
work
would
be
a
good
first
thing
to
get
involved
in
and
pretty
scoped.
C
So
we
are
scaling
up
over
here
on
the
vmware
side
and
we
have
a
new
engineer
in
china.
Her
name
is
wendley
and
she
is
testing
out
the
gmsa
stuff,
as
we
speak,
pere
has
been
training
her
on
some
stuff
recently
and
so
she's,
just
starting
to
get
involved
she's
in
the
asia
time
zone,
though
so
she
may
not
be
able
to
show
up
to
these
sig
windows
meetings
as
much
but
I'll
ping.
C
Her
about
this
today,
she's
really
interested
in
this
this
stuff,
which
is
kind
of
cool,
because
I
don't
know
a
lot
of
people
that
are
really
interested
in
gmsa.
But
she
is,
you
know
so.
Yeah.
A
C
Yeah
she's
running
those
as
well
she's,
starting
to
run
those
as
well.
D
I
guess
yeah
that
would
be
a
couple
for
those
tests,
so
they
should
be
getting
merged
or
promoted
pretty
soon.
A
So
the
nexus
windows
samples-
I
actually
don't
know
too
much
of
the
history
of
this
one.
You
can
take
a
look.
I
think
that
this
was
meant
to
be
more
or
less
kind
of
okay,
it
doesn't
look
like
there's.
A
whole
lot
here
looks
like
patrick
has
something
I
think
this
was
meant
to
be
kind
of
a
just
like
a
collection
of
kind
of
more
interesting
deployments
that
utilize
either
windows,
workers
themselves
or
windows
and
linux
workloads
here.
A
This
is
definitely,
I
think,
an
opportunity
that
anybody
can,
or
anybody
can
kind
of
contribute
to
this
and
to
make
this
better.
I
think
that
things
have
kind
of
stabilized
now,
but
I
know
kind
of
early
on
when
windows
workloads,
g8
there
was
tons
and
tons
of
questions
about,
can
clusters
even
support
windows
and
linux
nodes
in
the
same
cluster.
How
do
I
deploy
things
and
targets
windows
notes,
specifically?
A
How
do
I
do
all
of
that
stuff?
And
I
think
this
was
an
attempt
to
kind
of
demo
some
of
that
functionality
yeah.
I
see
that
this
has
got
some
linux
workloads
in
there
and
I'm
guessing
some
windows
workloads
in
there
too.
So
this,
I
think
we
could
maybe
talk
about
later,
because
the
other
two
are,
I
think,
more
interesting.
A
So
windows
testing
is
where
sig
windows
has
all
of
their
or
most
of
the
test
content.
So
the
kind
of
definitions
that
actually
drive
the
proud
jobs
live
in
the
kubernetes
test
info
repository,
but
windows
testing
has
a
lot
of
things
like
the
cluster
definitions
for
setting
up
those
clusters,
at
least
on
azure
and
gce.
A
This
is
probably
the
most
active
sub
project.
I
would
guess
today.
I
know
a
lot
of
folks
who
are
interested
in
testing
kind
of
contribute
and
keep
this
up
to
date.
There
are
probably
are
some
opportunities
to
clean
this
up.
I
don't
know
if
we
still
need
to
have
all
of
these,
so
these
are
all
the
different
test
images
that
are
used
or
potentially
used
by
the
ede
tests.
A
Here
I
believe
most
of
these
are
now
just
being
built
as
with
their
analogs
and
windows
or
in
linux,
so
we
could
probably
go
and
either
start
deleting
some
of
these.
I.
G
G
A
lot
of
those
images
have
been
centralized
into
the
agnost
image,
so
they
are
no
longer
relevant
because
the
last
time
they
were
being
used
was
kubernetes,
115,
116
or
something
like
that.
So
they're
definitely
not
relevant
anymore.
There
might
be
a
couple
of
few
images
that
might
be
interesting
for
other
than
e3
conformance
tests.
A
Sounds
good
and
then
we
also
do
have
documentation
for
how
to
build
and
run
the
ede
tests
on
run
them
against
nodes
with
window
or
like
windows
nodes
in
a
cluster.
I
believe
that
this
is
fairly
up-to-date.
I
know
I've
made
a
couple
of
updates.
Jay's
done
a
lot
of
work
to
keep
to
update
this
recently.
A
So
this
is
a
good
guide,
but
anybody
is
more
than
welcome
to
open
up
a
pr
or
an
issue
if
these,
if
they
have
any
kind
of
questions
about
these
documents
or
this
documentation
here,
so
folks
can
kind
of
just
look
at
that,
as
they
have
time
now.
Sig
windows
tools,
this
one
is
interesting
because
my
understanding
with
this
was
initially.
A
This
was
used
to
host
kind
of
some
of
the
scripts
and
workflows
and
deployment
files
that
were
being
used
to
test
cube,
adm
support
on
windows,
so
in
116,
maybe
some
folks
who
I
don't
believe
are
on
the
call
ben
moss
and
gab
from
vmware
we're
working
on
getting
cube,
adm
support
for
for
for
windows,
and
this
relied
quite
heavily
on
wins.exe
and
there's
a
lot
of
kind
of
scripts
around
setting.
A
A
This
feature
to
declare
it
stable
until
both
cluster
api
had
kind
of
full
support
for
windows
nodes,
and
also
that
the
windows
networkers
had
windows
had
privileged
containers,
so
that
we
could
not
need
to
have
kind
of
the
kind
of
official
deployment
scenarios
deploy
wins
just
for
a
variety
of
reasons,
mainly
kind
of
security,
oriented
reasons
for
this.
A
This
is
I'm
at
least
considering
this
still
to
be
kind
of
in
a
holding
pattern.
Here
I
know:
we've
the
the
windows
privilege
container
support
that
we
were
hoping
to
have
go
alpha
for
121
has
been
delayed
until
122,
but
all
indications
are
that
we
can
get
that
merged
pretty
early
in
the
122
code
code
base.
A
D
Yeah
I
mean
to
me
once
we
have
privileged
containers.
A
lot
of
this
can
go
away
and
some
of
the
work
there,
so
I
think
there's
there.
We
need
maybe
needed
for
a
team
to
do
all
the
various
work
to
get
the
emo
files
into
the
right
places
so
like
getting
coop
proxy
baked
into
where
the
linux
one
is
so
that
windows
works,
getting
the
yaml
files
defined
out
for
all
the
other
cni's.
D
So
once
we
have
privileged
containers,
we'd
like
to
have
other
cni's
bring
on
support
for
those
instead
of
going
through
kind
of
these
custom
scripts
they'd
each
in
the
individual
cni
would
maybe
support
those
themselves,
and
so
we'd
need
somebody
to
kind
of
take
that
on
and
and
get
it
working.
So
it's
pretty
fun,
exciting
stuff.
So.
A
Yeah,
as
james
mentioned
so
part
of
part
of
what's
also
in
this
repository,
is
the
deployment
file
that
will
deploy
either
q
proxy
or,
I
believe,
flannel
via
a
daemon
set
on
windows
and
eventually,
I
think
the
end
goal
should
be
to
not
need
a
separate
repository
to
host
those.
We
should
just
try
and
get
the
kind
of
the
official
or
the
the
the
standard,
q,
proxy
and
cni
deployments
to
just
check
and
see
hey
other
windows
nodes,
here's
what
the
demon
set
should
look
like
for
in
that
case,
instead.
C
A
I
think
it's
more
of
the
former,
especially
for
some
of
especially
for
some
of
the
the
cni
deployments,
like
a
lot
of
the
documentation
on
how
to
deploy
a
cni.
Is
you
set
up
your
linux
node
you
get
the
cubelet
running
and
then
you
deploy
a
daemon
set
from
that
and
the
demon
sets
like
under
you
know
the
core
os
repository
for
flannel,
but
then
the
documentation
for
or
at
least
the
guidance
for
windows
is
after
you
do,
that
you
go
and
you
deploy
this
deployment
out
of
the
sig
windows
tools
repository.
A
C
A
Yeah
go
on
for
q
proxy.
I
know
like
there's
many
different
ways
of
deploying
q
proxy.
We
could
have
it
so
that
it
is
more
of
if
you're,
using
cube,
adm
or
cluster
api.
It
will
go
and
get
a
baseline
set
up
for
the
for
just
like
a
privileged
aim
and
set
that
gets
the
point
to
the
nodes.
H
A
Or
suggest
a
completely
new
new
one,
the
other
one
that
I'll
call
out
briefly
here
and
I'm
actually
not
sure
why
it's
not
mentioned
here-
is
sig
windows
and
sig
storage
kind
of
co-own,
this
csi
proxy,
which
is
another
project
to
allow
out-of-tree,
csi
plug-ins
to
run
on
windows
nodes
which
don't
have
privileged
containers.
A
That
is
something
else
that
we
should
get
added
here
too,
and
I
know
jing
who's
on.
The
call
is
very
active
there,
along
with
deep
and
kai
and
a
few
other
folks
yeah
exactly
what
what
jay
said.
I
think
that
there's
a
lot
of
opportunities
for
anybody
who's
interested
to
kind
of
take
the
brains
on
some
of
these
and
give
them
some
of
the
love
that
they
haven't
had
in.
Some
of
the
recent
milestones.
C
Yeah
and
a
lot
of
us
have
been
doing
this
stuff
for
a
while,
we
can
help
like
we're
not
this
is
you
do
not
have
to
be
like
a
windows
expert
or
a
kubernetes
expert,
even
it's
more
owning
the
logistics
of
how
we,
how
we
grow
this
community
around
these
various
things
that
are
valuable,
that
we
that
we
curate
right
so
there's
definitely
a
lot
of
deep
technical
work
here.
But
a
lot
of
the
work
is
helping
organize
things
and
helping
us
make
sure
that
we're
doing
things.
H
Yeah,
I
don't
mind
helping
out
with
this
stuff
again.
I
don't
know
where
to
start,
but
I
think
one
thing
we
should
maybe
try
and
do
is
sort
of
prioritize
these
items
so
that
we
can
sort
of
tackle
the
low
hanging
fruit
first
or
tackle
something
that
we
really
need
to
get
and
done
in
the
short
term.
H
A
What
do
you
think
would
be
the
best
way
to
dive
into
some
of
these
our
vendor
jay?
Should
we
just
kind
of
make
one
thread
on
slack
and
seguenos
for
each
one
of
these
and
start
talking
on
there?
Should
we
devote
some
time
to
each
one
of
the
next
couple
community
meetings
to
going
deeper
into
any
one
of
these,
I'm
open
to
suggestions
and
kind
of
whatever
focus.
H
On
I
I
don't
know
about
jay's
touch,
my
suggestion
would
be
is
to
devote
sometime
during
like
10
minutes
just
to
kick
things
off
during
a
community
call
and
then
start
following
up
on
slack
to
see
what
comes
out
of
it.
H
C
I
yeah
I
mean
I
I'm
all
for
slack,
I
feel
like
we
I
feel
like,
though,
what
we
need
is
we
need
we
need.
We
need
owners
right.
We
need
people
to
say
I'm
interested
in
this
topic
and
I
want
to
own
it.
I
feel
like
we
do
do
a
lot
in
slack,
so
I
feel
like
not
to
not
do
that
like
we
should.
We
should
definitely
keep
using
slack
because
that's
like
the,
but
I
mean
I
think
we
I
was
hoping
we
could
kind
of
like
get
some
people
to
say
like
hey.
C
A
Yeah
another
option
is
we
could
kind
of
try
and
kick
things
off
on
slack
and
schedule
like
kind
of
out
of
band
meetings,
especially
for
like
the
gms
a1j,
where
you
mentioned,
there's
people
that
might
not
be
able
to
make
this
meeting
at
their
normal
time.
We
could
say:
hey
we're
looking
for
people
who
are
interested
in
kind
of
taking
ownership
and
curating
this
we're
more
than
hope
or
they're,
more
than
happy
to
help
get
them
started.
F
Yeah,
I
was,
I
was
also
thinking
the
same
thing
mark
I'm
interested
in
some
of
the
areas,
but
getting
started
is
hard.
So
if
we
schedule
like
kind
of
a
hackathon
where,
like
there
are
experts
there,
like
you
know,
if
I
need
to
set
up
my
dev
machine
or
whatever
like
for
that
area,
get
kind
of
getting
started
then
also
like
what
owning
it
means
or
what
it
entails.
What
time
commitment
we're
looking
for,
I
think
those
questions.