►
From YouTube: Kubernetes SIG Windows 20210427
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
A
All
right.
Let's
get
started,
I
think
we're
still
doing
some
intros
for
everybody
who's
new.
I
think
I
s
most
of
the
people
on
the
call
I'm
look
familiar,
but
if
there's
anybody
who
just
wants
to
say
hi
or
changing
plans
or
wants
to
work
on
something
different
or
anything
like
that,
please
feel
free
to
just
go
ahead
and
say
hi
or
give
an
update
on.
If
anybody
wants
to
give
an
update
on
what
they're
working
on
anything
like
that,
if
not,
we
can
go
into
the
announcements.
A
All
right,
let's
go
into
the
announcements.
First
thing
is
the
122
release.
Schedule
has
been
announced,
it
was
sent
to
the
kubernetes
dev,
alias,
which
everybody
who's,
an
org
member
should
be
subscribed
to,
but
some
highlights
are,
enhancement,
freeze
is
may
13th
and
the
code
freeze
is
slated
for
july
8th
and
then
the
release
is
targeted
for
that
first
week
in
august.
A
So
they
are,
I
think,
as
discussed
last
week,
they
are
trying
to
or
they
are
settling
on
the
three
releases
a
year
at
least
for
this
year.
So
there's
a
little
bit
more
time
in
between,
but
it
also
means
that
I
guess,
if
things
get
slipped,
it
is
a
little
bit
longer
for
them
to
make
it
in
there.
A
So
just
check
back
on
the
google
groups
and
on
the
the
kubernetes,
the
kubernetes
community
or
release
page,
has
many
updates
to
that
too
kind
of
in
a
similar
note
that
next
cherry
pick
deadline
is,
I
believe,
may
7th.
So
that's
coming
up.
I
think
that
I
know
james
is
working
on
getting
those
fixes
in
for
the
overlay
networking
issues.
If
anybody
else
has
any,
please
feel
free
to
ping.
A
Us
to
help
get
the
release,
manager's
attention
and
we'll
try
and
get
those
in
next
one
is,
I
will
be
taking
some
the
second
part
of
my
parental
leave,
starting
in
may
so
I'll
be
out
and
take
a
little
bit
of
vacation
at
the
end,
so
I'll
be
out
for
about
two
months,
but
I
think
james
and
jay
and
everybody
else
here,
arvin's
been
helping
moses,
been
upping.
Everybody
here
can
help
kind
of
take
over
that.
A
If
anybody
has
any
questions
or
concerns,
yeah
feel
free
to
reach
out
to
me,
but
if
not,
I
think
we're
leaving
everything
in
good
hands.
A
One
last
announcement
is
kubecon
eu
virtual
starting
next
week,
if
anybody's
able
to
or
wants
to
attend,
it's
a
good
way
to
get
some
exposure
from
what
a
lot
of
the
other
cigs
are
doing.
All
of
the
talks
are
always
recorded
and
uploaded
to
youtube
afterwards
and
all
the
slides
are
made
available.
So
if
you
prefer
to
just
catch
up
with
those
on
your
own
time,
feel
free
to
do
that.
A
I
believe
this
sig
windows
maintainer
track
talk,
which
has
some
deep
dive
into
networking,
is
scheduled
for
about
4
10
a.m.
Pacific
time,
so
I
wouldn't,
if,
if
people
want
to
just
watch
that
recording
well
feel
free
to
do
that,
that's
yeah!
If
anybody
has
any
questions
about
that,
also
there's
a
lot
of
good
resources
for
that.
A
So
now
I
think
we
can
get
into
the
agenda.
Was
I
see
you
did
join
the
call
I
I
know
you
messaged
us
internally
yesterday
saying
that
you
had
some
windows,
licensing
updates
that
you
want
to
provide.
Do
you
want
to
go
ahead
with
that.
B
Yeah
yeah,
that
sounds
good,
so
I
think
jay
brought
this
up
and
other
members
about.
You
know
there
are
folks
who
are
trying
to
start
in
kubernetes
community.
Of
course
we
would
love
to
help
and
they
would
need
the
developer
license
for
windows,
server
2019
to
set
up
their
cluster
and
you
know
get
started
in
the
community.
So
I
asked
so
I
went
to
our
you
know
the
windows
leadership
they
all
agreed.
This
is
very
important.
B
We
want
developers
to
come
so
we
are
working
on
a
program
which
is
we
are
calling
a
developer
program
for
community
to
think
about
how
we
can
provide
some
licenses
for
free
or
like
we
could
work
something
out
with
the
with
the
legal
department.
But
that's
gonna
take
some
time,
so
I
wanted
to
just
give
the
some
of
the
options
that
are
available
today
and
the
you
know
that
could
be
readily
used.
I've
confirmed
so
so,
starting
with
the
visual
studio
subscriptions.
B
I
think
there
was
a
question
from
james
last
time.
He
was
asking
if
they're
mtn
subscription.
So
yes,
I've
confirmed
with
if
you
have
a
visual
studio
subscription,
it's
very
common
with
companies.
If
you
have
one,
you
could
download
and
use
a
windows
server
for
deaf
purposes,
so
it
comes
with
your
visual
studio
licensing
and
if
you
have
any
questions
there,
let
me
know
happy
to
answer.
B
B
So
everybody
have
it
for
for
right
now,
just
one
more
option
that
I
was
told
was
there
is
a
partner
program
that
if
I
think
it
is
only
for
microsoft
licenses,
but
if
somebody
is
working
for
one
of
the
microsoft
partners
they
have,
you
know
they
can
get
the
windows
server
for
like
extremely
cheap.
I
think
it's
like
under
100
or
something
still
it.
C
B
You
know
well,
there
is
some
money
involved,
but
there's
that
option
and
the.
D
B
Is
the
evaluation
version
should
work
for
180
days
and
I
think
from
a
cluster
creation
perspective
like
you
know
that
that
should
work
for
180
days,
so
these
are
like
kind
of
four
options
that
I'll
post
in
this
ignored
windows
as
well.
That
could
be
used
right
now,
as
we
are
playing.
You
know,
thinking
about
how
we
can
provide.
You
know
licenses
to
developers
whoever
wants
to
contribute
in
the
community.
A
I
think
I
missed
this
last
week,
but
I
will
also,
I
think
I
can
also
reach
out
to
one
of
our
pms
over
here
in
azure,
who
manages
our
the
azure
subscriptions
that
we
donate
to
the
cncf
org.
A
I
know
in
the
past
we've
given
virtual
machines
for
running
like
ci
tests
that
require
windows
nodes
to
individual
projects.
That
may
be
another
option
too.
B
Yeah,
that
would
be
great.
I
yeah,
I
think
you
you
were
out
marked
last
week
so
last
week,
jay
and
and
a
couple
of
folks
brought
up
that
you
know
their
folks
is
starting
trying
to
trying
out
to
contribute
to.
You
know,
sick
windows
and
of
course
you
would
need
to
set
up
a
cluster
and
you
would
need
a
windows,
server,
2019
license
and
that's
their
hurdle,
and
it
shouldn't
be
so
it
would
be
great
if
you
could,
but
from
a
windows
container
perspective.
B
A
Yeah,
I'm
not
sure
how
quick
the
last
option
that
I
mentioned
is
either,
but
if
anybody
is
interested
try
reaching
out
to
james
or
myself,
we
can
connect
people
with
the
right
folks.
For
that,
thanks
moz.
A
Craig
on
the
call
or
caico:
do
you
either
of
you
have
any
would
have
any
comments
for
this
too.
I
don't
know
if
you
guys
would
be
able
to
help
provide
some
azure
resources
for
development
purposes.
E
E
Yeah,
you
know
I
could
go
and
ask,
but
currently
I
don't
have
any
contact
per
se
but
yeah.
Let
me
let
me
go
and
ask
about
this.
F
A
B
I
think
jay
I
mean
I
think
jay
is
not
today
here
in
the
meeting,
but
I
think
he
he
brought
it
up,
as
I
think
he's
working
with
our
team
start
working
after
hours,
some
of
the
folks
wanted
them.
It
wasn't
a
general
question.
It
was
like
specific
but
also
kind
of
general.
B
I
mean,
because
you
know
we're
trying
to
set
up
that
after
hours
and
like
make
it
easier
for
folks
to
contribute-
and
I
think
going
through
like
setting
up
instructions
or
whatever
james
and
g
are
working,
which
is
pretty
awesome.
That's
one
of
the
things
they
want
to
include
there
that
hey.
You
know
you
can
set
up
the
environment
this
way
and
if
you
need.
G
A
Yes,
I
think
that
would
be
that
eval
version
that
was
mentioned.
B
Way
like
there
is,
you
know,
like
you
know,
it's
all
good
to
use
it
that
way,
but
we'll
find
other
ways
as
well
to
provide
licenses.
A
All
right
is
there
any
other
discussion
topics
for
this.
A
Sounds
like
you're
ready
to
move
on
this
has
been.
I
know
this
has
been
brought
up
a
couple
of
times,
but
I
wanted
to
re,
highlight
the
pspv
next
work
for
for
windows.
Here
they
are
planning
on
moving
that
to
alpha
this
release,
and
I
think
the
current
state
of
things
are
is
they're
planning
on
including
mechanisms
or
settling
on
a
mechanism
to
exclude
windows
pods
from
a
lot
of
the
the
policy
enforcement.
A
So
that
is,
I
think,
that's
helpful,
but
it
would
be,
I
think,
even
better
if
the
this
pspv
next
would
be
able
to
go
in
as
enough
implementation
with
windows
with
some
windows,
specific
kind
of
policies
and
implementations
too,
and
I
think
currently
looking
for
contributors
to
help
with
that.
So
if
anybody's
interested
you
know
feel
free
to
either
review
the
cap
comment
on
the
cap
contact
james
and
myself,
we've
been
involved
in
some
discussions
with
that
and
everything.
A
So,
there's
that
I
don't
see
anything
else
on
the
agenda,
so
I
don't
know
if
we
want
to
open
it
up
for
just
general
topics
or
maybe
go
into
the
after
hours
a
bit
early
or
if
anything
has
anybody
else
to
add.
If
anybody
has
anything
else
to
add,
feel
free
to.
D
D
A
Sure
can
you
add
them
to
the
agenda
too
so
that
we
can
open
up
the
links.
F
D
For
the
first
one
I
I
mentioned
this
one
once
in
our
meeting,
it
is
the
pod
start
in
the
termination,
so
I
go
back
and
gather
more
events
from
the
kill
bullet.
So
it
seems
that
I
I
I
feel
I
feel
this.
This
seems
a
little
bit
familiar
because
it
seems
like
in
the
past
some
darker
side
have
some
similar
issue
as
well.
It's
like
they
tried
to
delete
a
a
log
file,
but
it
was
not
able
to
usually
if
it
has,
it
is
being
called
by
something
some
other
processes.
D
So,
but
I
do
see
some
aks
side
to
fix
some
issue.
I
think
some
racing
improves
the
racing
conditions
yeah.
I
mentioned
the
aks
one
in
the
first
one.
First.
C
D
So
I'm
not
sure
if
this
is
still
repro
in
aks
or
if
it
is
related
to
certain
dog
version,
so
or
I'm
just
not
sure
whether
or
not
this
fix
is
really
really
take
the
impact
for
all
the
clusters
or
because
the
fish
seems
to
be
very
general.
I
feel
it
should
be,
but
so
the
first
thing
I
want
to
know
is
whether
or
not
this
still
repro
in
aks.
A
Keiko,
do
you
have
any
comments
on
that?
I'm
not
too
familiar
with
what
all
like
what
all
issues
have
been
reproducing,
which
one
time
the.
A
It
looks
like
from
the
commented
stalker.
E
Why,
like
windows,
no,
not
ready
happen
and
some
of
them
again
it
does
relate
to
the
container
networking
yeah.
Let
me
let
me
check
back
on
this.
Oh
yeah
yeah,
so
I'm
trying
to
fix
some
intensity
yeah.
So
this
one
is
because
this
is
exactly
the
good
of
that.
G
E
Yeah
there
was
an
hms
issue
that
happened
on
the
os
that
caused
a
lot
of
issues
and
we
had
to
postpone
it
and
provide
tennessee.
Instead
of
working
with
the
container
networking
team
on
the
windows
side,.
A
Yeah,
so
I
think
that
this
issue
was
there
with
the
october
security
updates.
There
were
some
issues
with
the
container
networking,
but
I
believe
that's
been
resolved.
It's.
D
E
Do
is
we
receive
the
windows
operating
system
with
the
patches
on
every
month
and
we
do
update
them.
So
it's
a
monthly
update
that
customers
will
need
to
do
manually
and
let
me
post
you
the
the
release,
note
information,
so
that
you
know,
like
you
know,
what's
been
released
recently.
D
Yeah,
actually,
I'm
sorry.
We
also
do
that.
We
it's
like
a
tuesday
patch,
but
so
do
you
mean
what
I
was
asking
is
because
you
know
because
we
we
also
release
the
note
images.
I
just
want
to
be
clear
which
which
reach
which
patch
so
I
can
check.
You
know
when
this
cluster
reproves,
whether
or
not
that
node
image
has
a
patch.
A
D
Yes,
but
the
problem
is
the
this:
doesn't
repro
consistently,
so
if
we
we
want
to
suggest
someone
to
upgrade,
we
cannot
suggest
you
know
upgrade
it
immediately
to
the
latest.
Usually
it's
like
to
the
closer
virgin.
So
that's
why
I
I
want
to
make
sure.
I
Just
but
if
this
is
just
or
I'll
go
ahead,
perry
sorry.
This
looks
very
similar
to
an
issue
that
we've
had
that
me
and
james
have
seen
at
some
point
on
container
d,
but
where,
because
it
looks
like
the
lot,
the
logs
saying,
access
denied,
which
means
it's
still
open
and
then
the
finalizer
can't
delete
it
and
then
it
kind
of
just
sits
in
this
never-ending
state.
But
I
don't
know
where
that
issue
got
to.
I
don't
know
if
js.
G
Yeah,
it
looks
like
a
different
set
of
error
messages.
I
was,
I
was
thinking
the
same
same
thing,
but
it
does
look
like
a
different
set
of
error
messages.
A
A
long
time
ago,
well
a
long
time
ago
in
kubernetes
time
that
should
have
addressed
that
issue
too.
So
it's
possible
that
it's
a
different
issue.
As
caico
mentioned,
there
are
a
lot
of
different
ways
that
you
can
kind
of
see
the
see
these
same
symptoms
where
you
can't,
where
you
get
stuck
terminating.
D
So
doesn't
mean
that
it's
like
both
both
the
virgin
is
like
kubernetes
virgin
matters
and
also
the
windows.
Patch
version
also
matters
right.
It's
like
you
needed
to
upgrade
it
to
certain
kubernetes
version,
and
also
you
need
to
apply
certain
windows
os
patch.
A
D
A
There
was
that,
just
from
reading
this,
it
seems
like
there's
something
that's
stuck
getting
cleaned
up,
it
seems
like
the
os
fix
would
prevent
that
from
leaving
those
bad
files
around
and
the
coop
the
cubelet
fix
would
either
work
around
the
issue
or
try
and
more
forcefully
clean
up
those
resources.
A
D
Oh
okay,
I
can,
I
can
run
the
command
and
fold
it
out.
I
will
put
it
in
the
lock
later
yeah.
A
Okay,
yeah,
I
think
this
one.
It
looks
pretty
generic
too
unable
to
attach
mod
volumes.
A
D
D
So
that
is,
that
is
the
secret
volume,
and
it
seems
that
the
the
they
both
they
have,
the
empathy
there
and
the
secular
volume
failed
to
mount,
and
but
I
feel
I
feel
it
it
already
it.
It
failed
in
create
the
pod
sandbox.
D
So
and-
and
this
happens
when
there
were
two
parts
you
know
they're
both
trying
to
trying
to
it-
seems
that
there's
some
concurrent
operations
going
on
and.
A
Yeah,
do
you
know,
are
you
able
to?
I
guess
a
couple
of
things
maybe
to
try
are:
are
you
able
to
start
the
two
pods
if
you
space
their
starting
times
out
and
like?
Do
you
only
hit
this
if
you
try
and
start
two
parts
in
the
same
note,
at
the
same
time
or,
and
things
like
that,
that
might
be
good
yeah.
D
I
think
it
was
happen
when
you,
when
you
suddenly
start
start
a
lot
of
parts,
and
I
also
is
something
similar
when
I
kill
some
pod
and
then
risk
try
to
restart
it
immediately.
So
it's
very
frankly
killing
the
pod
and
the
restarting
and
we're
also
saying
sometimes
saying
hey.
You
failed
to
amount
to
the
volumes
so
julie.
B
Okay,
yeah
this
we
haven't
seen
this
on
docker.
There
was
something
about
secret
volumes
on
container
d
we
have
seen,
but-
and
these
concurrent,
when
you
are,
when
you
mean
handling
concurrent
aux
actions,
you
mean
like
you're,
trying
to
scale
you're
going
to
spin
up
multiple
parts.
At
the
same
time,.
D
B
D
Let
me
see
I
I
didn't.
I
didn't
watch
the
watch
that
when
let
me
maybe
take
a
look
at
the
log,
I
think
let
me
see
if
they
also
set
up
the
limit
and
the
request
correctly
yeah.
Let
me
then
take
a
look
at
the
products
back.
B
A
Yeah
for
this
one,
it
might
be
interesting
to
to
comment
on
what
the
node
configuration
is
like
the
node
size,
what
the
system
reserves
are
and
like
how
many
pods
you
see
spinning
this
up
with,
I
don't
think
we
haven't
seen.
This
doesn't
seem
to
be
too
widespread.
Otherwise,
I
think
I'd
see
more
issues
in
kubernetes
about
this,
so
it
could
just
oh.
H
G
D
A
I
think
we're
kind
of
getting
to
the
top
of
the
hour.
Does
anybody
have
any
thing
else
to
discuss
otherwise
I'll,
probably
hand
it
over
for
the
after
hours
pairing
session.
A
All
right,
thanks
all
I'm
going
to
end
this
meeting,
everybody
have
a
good
day
and
good
week
stay
healthy.