►
From YouTube: Kubernetes SIG Windows 20220111
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
january
11th
2022
iteration
of
the
kubernetes
sig
windows,
kubernetes
community
kubernetes
windows,
community
meeting.
As
always,
these
meetings
are
recorded
and
uploaded
to
youtube
so
be
sure
to
get
here
to
the
cncf
code
of
conduct.
A
Let's
get
started
biggest
announcement
is
the
124
release.
Schedule
has
been
finalized.
A
The
enhancements
freeze
date
was
pushed
back
a
week
from
what
I
mentioned
last
week
after
a
number
of
leads,
including
myself,
said:
it's
a
longer
release.
Let's
push
enhancement
freeze
back
by
week
and
the
release
team
didn't
and
the
enhancement
team
did
not
have
any
issues
with
that.
So
now
enhancements
freeze
is
february
3rd
of
this
year,
so
it
just
gives
everyone
a
little
bit
more
time
to
get
any
enhanced
like
the
kept
reviewed
and
and
merged
with.
B
A
Rest
of
the
early
schedules
linked
here
take
a
look
if
you
have
any
questions
the
next.
Actually,
I
think
this
is
probably
we'll
just
make
that
an
agenda
item
does
anybody
else
have
any
announcements
to
make
now
we
can
get
into
the
schedule.
A
Okay,
oh
james
you're,
writing
something.
A
We
knew
it
started
a
different
thing,
yeah
first
thing.
Actually
james.
Do
you
want
to
talk
about
this,
or
should
I
talk
about
this?
The
github
container
registry
go
ahead.
A
Okay,
so
in
the
past
there's
been
a
kind
of
sig
windows
maintains
this
docker
hub
group
segment
is
tools
and
things
like
the
gmsa
web
hook
and
a
couple
of
the
images
used
for
the
cube.
A
Adm
implementations
were
all
hosted
here
and
recently,
or
maybe
not
even
recently,
but
we
recently
noticed
that
the
kubernetes
six
org
has
enabled
the
packages
for
that
org
and
the
container
the
github
container
registry
or
repository
functionality
has
been
kind
of
stable
for
a
while
now,
so
we
james
took
a
look
and
decided
that
our
like
this
is
probably
ready
to
use.
A
So
I
think
in
the
future,
we're
planning
on
updating
all
the
tools
from
sig
windows
tools
to
push
their
images
and
and
any
other
repositories
that
we
have
to
push
their
images
to
the
kubernetes
sigs
packages
repository.
A
If
anybody
has
any
like
concerns
or
issues
about
that,
please
let
us
know
this
does
seem,
has
a
pretty
nice
ui.
You
can
show
it
here
and
james
pushed
one
new
image
for
an
eventful
logger
container
image
that
he
built
here
to
kind
of
test
this
out.
This
is
coming
from
the
sig
windows
tools
repository
as
highlighted
here.
A
You
can
click
in
to
this
and
manage
the
versions
and
see
all
the
tags
and
everything,
and
so
for
like
a
multi-arch
image,
you
can
see
the
different
images
in
here
and
everything
with
the
os
versions
and
the
shaws
of
all
the
images,
and
this
is
open
to
the
public.
So
this
is
like
totally
public.
The
container
images
management
is
tied
to
the
permissions
on
the
repository
that
pushed
the
images.
So
that's
also
a
pretty
nice
user
experience.
A
We
don't
need
to
manage
separate
groups
or
like
docker
hub
accounts
and
all
that,
so
this
is
kind
of
an
fyi
that
this
is
something
that
we're
going
to
be
pursuing
kind
of,
as,
as
we
have
time
moving
this
out
here
and
if
you
weren't.
A
Already
aware
of
the
github
container
registry
and
you're
looking
for
an
alternative
to
docker
hub,
please
take
a
look.
It
looks
pretty
slick.
C
Yeah,
so
I
just
had
a
couple
things
that
so
the
the
gmsa
image
we're
actually
gonna,
be
moving
into
the
k8s
dot
io
container
registry,
because
it's
an
official
image
that
we're
going
to
be
supporting
and.
A
C
Yep
so
jamie's
been
helping
with
that
and
then
this
event
flow
logger.
It's
a
project
that
I
put
together.
I
needed
to
get
the
logs
from
container
d,
it's
a
it's
a
host
process
container
that
listens
to
egw
and
then
boards
that
can
console
out
and
it's
completely
configurable.
So
you
can
listen
to
any
type
of
etw
event
and
then
it'll
forward
it
to
standard
out,
and
it
could
also
forward
it
on
to
other
outputs.
C
So
you
can,
I
think
you
can
put
it
to
some
of
the
different
azure
pin
points
some
google
endpoints,
depending
on
where
you
want
to
put
where
you
want
to
push
that
that
data
for
your
own
logging
purposes,
but
container
d,
only
logs
etw
events
and
so
by
default,
and
so
this
is
a
quick
way
to
be
able
to
configure
and
turn
those
on
and
off
so
check
it
out.
If
you're
looking
at
login
continuity.
A
Yeah
and
here's
an
example
of
kind
of
the
config
file
for
that
you
can
name
the
etw
session
well,
say
if
it's
an
etw
session
name
the
etw
session
list
of
the
etw
providers
that
you
want
and
the
output
type.
So
this
might
be
a
great
way
for
people
to
get
logs.
Oh
also,
so
this
is
set
up
to
run
as
a
host
process
container,
but
I
think
it's
mentioned
in
the
readme
that
this
is
something
that's
very
easy
to
run
as
a
standalone
binary
as
well.
A
A
Okay,
next
agenda
item
is
back
in
december
we
had
kind
of
talked
about,
and
I
think
there
was
a
lot
of
agreement
for
figuring
out
a
time
to
have
a
shared
space.
To
do
some
of
the
docs
updates.
I
asked
a
few
people
and
it
seems
like
that.
A
I'm
going
to
propose
that
we
for
a
couple
of
weeks
have
a
bi-weekly
meeting
to
kind
of
as
a
touch
point
for
some
of
the
docs
updates
that
we
want
to
do
as
segwindows,
and
I
wanted
to
propose
that
we
all
use
alternating
thursdays
for
the
thursdays
at
9
30
a.m.
Pst.
So
the
same
time
as
this
meeting,
but
the
weeks
that
we
don't
do
the
backlog,
grooming
or
refinement
to
kind
of
use
the
touchpoint,
and
then
we
can
sync
up
see
what
kind
of
updates
we
would
like
to
do.
A
Maybe
assign
some
owners
to
that
and
then
follow
up
in
the
following
two
weeks.
Does
that
I
guess
does
anybody
have
any
issues
with
that?
Is
that
does
that
time
not
work
for
people
who
would
want
to
be
part
of
that
effort
or
or
anything
like
that
I'll
also
post
in
slack
sometime
today,
to
get
some
more
feedback
on
that?
A
A
I
see
jamie
joined,
I
wanted
jamie.
Oh,
we
had
just
talked
about
one
one
thing
that
we're
doing
kind
of
in
parallel
with
the
efforts
that
you're
doing
to
publish
the
gmsa
container
image
to
the
official.
You
know,
cades.gcr
that
I
o
we
are
investigating
looking
at
moving
the
pushing
container
images
that
are
built
out
of
sig
windows
tools
and
some
of
the
other
ones
that
we
wouldn't
really
consider
as
an
official
in
an
official
capacity
to
the
github,
credential
or
github
container
registry
for
the
kubernetes
org.
A
E
A
But
I
think
jay
jamie.
Do
you
have
the
next
agenda
item.
B
Yeah,
I
think,
there's
like
100
threads,
going
on
about
csi,
proxy
and
stuff
like
that
and
different
channels
and
stuff
I
figured
we
should
all
just
talk
about
it.
So
what's
the
latest,
I
haven't
really
tested
it
on
vsphere.
Yet
to
be
honest,
so
I
I'm
I'm
just
curious
to
learn
more
about
where,
where
you
all
are
at
jamie
luther.
E
E
Since
you
know,
rke
rancher
rk2,
like
you
know,
depends
on
wins
to
do
a
lot
of
functionality,
so
we
started
booting
csi
proxy
and
as
part
of
that,
I
have
to
provide
qa
with
a
workflow
to
test
which
then
I'm
having
to
read
and
see
where
all
the
implementations
are
and
then
luther
kindly
went
to
the
csi
sig
meeting
yesterday
for
csi
proxy
and
it
looks
like
the
smb,
the
vsphere,
the
azure
ones,
and
the
gcp
ones
are
the
most
mature
implementation,
wise
with
smb
kind
of
being
the
one
of
the
better
ones
like
like
you
know.
E
Alpha
is
v
series
alpha,
but
we're
going
to
start
testing
that
too,
so
we
can
give
some
feedback
there
for
anybody,
that's
looking
to
use
it,
but
yeah.
I
have
a
scenario
between
all
the
different.
The
smb
csi
repository
by
the
way
has
the
best
most
complete
gamol
that
I've
found
so
far
for
like
if
you're
trying
to
do
something
local.
E
E
So
but
it's
all
there.
It's
in
the
it's
in
the
v
sphere
chart
and
it's
in
the
smb
chart.
So
the
official
csi
smb
driver
chart
has
the
option
to
enable
windows,
support
which
means
you'll,
get
pulled
down
all
the
right,
csi
proxy
bits
and
the
vsphere
charts
the
same
way
so
just
know
that
those
both
ship
with
that
stuff
there,
I'm
pretty
sure
the
azure
one
probably
has
a
chart
too.
E
That's
got
that
enabled,
but
so
there's
there's
some
stuff
floating
around
there
that
we're
kind
of
digging
into
and
testing,
because
we
know
you
know,
customers
want
storage
and
csi.
Proxy
is
probably
one
of
the
better
choices
at
the
point
for
for
getting
it
even.
B
A
E
I
I
think,
it'd
be
a
pretty
easy
thing
to
add.
Like.
E
Csi
proxy
is
pretty
simple.
The
one
big
issue
with
csi
proxy
as
it
sits
right
now,
is
there's
only
two
official,
I'm
gonna
air
quote
official
there's
only
two
people
publishing
binaries
for
it
and
that's
microsoft
and
google,
so
you
can
pick
which
one
you
want
to
choose
we're
pulling
the
ones
from
microsoft,
because
that
was
the
first
person
to
respond
to
be
like.
Oh,
we
got
pre-compiled
binaries
that
you
can
pull
so,
but
I
also
know
that
google
does
compile
them
and
have
them.
E
It
would
be
nice
to
maybe
get
those
officially
like
being
built
and
published
as
part
of
the
project,
so
it's
easier
just
to
know
where
to
go
pick
them
up,
but
I
think
their
direction
is
changing
a
little
bit
with
with
host
process
containers
coming
and
stuff.
But
I
I
personally
don't
see
if
you're
worried
about
using
csi
proxy
now
my
personal
opinion
so
far
is
I
don't
think
it's
going
to
go
anywhere
anytime
soon,
because
there's
lots
of
work
from
the
csi
plugins.
E
F
F
E
Yeah,
so
it's
I
don't
know
it's
interesting,
it's
new!
That's!
What
I
didn't
realize
is
how
new
it
was.
They
said
1.0
I
was
like
okay,
it's
ready
and
then
that's
like.
I
keep
forgetting
not
everybody,
every
project's
caught
up
to
what
they're
doing,
and
so
you
know
it's.
It's
like
trying
to
get
cni's
as
host
process
containers
like
well,
they
got
to
be
making
windows
cni's
and
they
got
to
be
making
the
images
and.
A
Yeah,
I
think
that
there's
currently
like
the
current
stages,
that,
like
my
understanding,
is
that
there's
the
csi
proxy
today,
which
it
can
be
compiled
and
run
on
the
node
and
that
needs
to
be
run
on
the
node.
And
then
the
different
plugins
can
connect
to
you
have
to
in
your
container
images
or
in
your
in
your
pods.
A
You
have
to
mount
in
the
named
pipes
that
are
exposed
by
that
and
then
there's
kind
of
an
intermediate
step
that
I've
tested
not
very
extensively
and
that's
running
the
csi
proxy
that
binary
in
a
host
process
container
instead
of
the
setup
as
a
service
on
the
host,
and
I
do
plan
on
making
an
image
a
container
image
for
that
and
putting
it
inside
windows
tools
and
publishing
it
to
the
that
same
kubernetes,
github
container
registry
in
the
near
future
as
and
then
the
long-term
strategy.
And
that
will
operate
the
same
way.
A
So
that
will
expose
named
pipes
on
the
host
that
the
different
provider
like
csi
plug-in
providers,
can
mount
in
via
volume
mounts
in
their
deployments.
And
then
the
long-term
strategy
is
to
not
require
csi
proxy
to
be
running.
On
the
host,
but
instead
use
it
as
a
binary
and
then
the
whole
thing
will.
The
whole
plugin
will
run
as
a
host
process
container
and
just
call
into
the
call
in
the
csi
proxy
library
which
will
do
all
of
the
operation
storage
operations
as
possible
as
needed.
E
Yeah
I'll
post
the
link
to
how
you
enable
it
if
you're
using
wins.
So
if
you
got
the
latest
version
of
wins,
you
can
just
pop
in
some
config
and
he'll
start
up
csi
proxy,
for
you
does
it
and
download
it.
It
doesn't
really
manage
it,
but
it's
a
way
to
to
kind
of
get
started
and
testing
it
out.
If
you
want
to
pretty
easily
so.
D
D
Change
that
right
so
csi
proxy
itself
is
ga,
so
the
api
is
v1
already
and
so
for
smb,
gcpd
and
azure.
Maybe
they
are
all
considered
as
those
drivers.
A
Yeah,
I
think
that
jamie
was
specifically
looking
for
ones
that
could
be
like
all
local,
so
you
could
set
up
a
smb
server.
E
E
Yeah,
that's
the
hard
thing
for
like,
especially
if
you're
doing
development
right
like
running
as
much
things.
Local
is
faster
feedback
and
a
little
easier
from
my
experience.
So
that's
kind
of
where
I
focused,
I
know
the
gce
one
and
azure
ones
are,
are
being
used
right
now
and
and
seem
to
work
really
well
so.
B
B
B
Yeah,
if
anybody
wants
to
put
some
some
some
some
like
passive,
like
you
know,
like
leverage.
A
A
lot
of
times
when
I've
talked
to
some
of
the
reviewers,
like
I
always
try
to
you
know,
say
like
hey.
I
don't
want
to
be
annoying,
but
a
lot
of
times
they're
like
sometimes
they
just
have
so
much
inform
like
different,
interrupt
like
so
many
different
notifications
that
they
just
ignore
everything
and
like
to
be
kind
of
interrupted.
A
B
Sure,
yeah
and
probably
paying
a
meme
on
it
too,
he's
the
it's
his
pr.
Actually,
okay,
yeah.