►
From YouTube: Kubernetes Office Hours 20200318 (West Coast Edition)
Description
Office Hours is a live stream where we answer live questions about Kubernetes from users on the YouTube channel. Office hours are a regularly scheduled meeting where people can bring topics to discuss with the greater community. They are great for answering questions, getting feedback on how you’re using Kubernetes, or to just passively learn by following along.
For more info: https://github.com/kubernetes/community/blob/master/events/office-hours.md
A
All
right
buddy,
it's
the
third
Wednesday
of
the
month.
That
means
it's
time
for
kubernetes
office
hours
are
monthly
livestream,
where
we
take
your
questions
from
the
audience
and
try
to
get
as
many
kubernetes
answers
as
we
can
in
an
hour
via
our
panel
of
experts
who
are
going
to
it's
troduce
themselves,
then
we're
going
to
show
you
how
it
works,
and
then
we
will
get
to
helping
so
introductions.
Let's
go
pier
B,
be
tuned
Brandon
and
then
Dave
I.
C
D
I
don't
know
if
I
was
supposed
to
be
next
go
I'm
Brandon
I
am
I,
am
a
longtime
user,
a
first-time
caller
I've
been
using
kubernetes
since
28
2015,
7,
prototype
and
stuff,
so
I
don't
operate
clusters,
but
I've
moved
a
lot
of
applications
into
kubernetes
from
like
legacy
stuff.
So
I
can
potentially
help
you
with
that.
E
A
All
right
and
someone's
asking
the
link
to
join
DevOps
day
me
or
do
you
want
to
hop
in
the
zoom?
Do
you
want
a
volunteer
your
time?
You
could
do
it
on
the
spot.
This
is
Brandon's
first
time
Dave.
This
is
what
your
second
time
tuned
to
a
second
time,
yep,
all
right,
let
me
know,
and
if
Jeff,
if
you're
listening,
if
you
can
pace
I
got
it
I
got
it
yeah,
we're
about
to
take
questions
angels,
sorry,
one!
A
Second,
someone
asked
to
join
the
panel
so
DevOps
day
manager
pink
just
say
if
you
want
to
hang
out
on
the
zoom
and
answer
questions
as
well,
because
this
is
open
for
everyone
all
right.
So
let
me
get
started
while
we're
figuring
stuff
out,
welcome
everybody
to
the
kubernetes
office
hours.
The
way
this
works
is
this
is
a
kubernetes
event,
so
the
code
of
conduct
is
in
effect,
so
please
be
excellent
to
each
other.
I'm
gonna
go
over.
The
rules
feel
free
to
introduce
yourself
in
the
channel.
A
Tell
us
where
you
work
on
what
you're
working
on
who
you
work
for
what
you're
doing,
how
you're
coping
working
from
home?
If
you're
working
from
home
and
where
you
are,
we
like
to
have
a
lot
of
audience,
participation
during
this
makes
things
a
lot
a
lot
of
fun,
so
you
can
feel
free
to
just
start
asking
your
questions
just
type
them
in
there
do
like
question
:
or
something
to
make
it
obvious
for
us.
A
What
we
do
is
we
paste
these
into
a
hack,
empty
document
and
then
read
off
of
those,
so
we
can
keep
the
show,
notes
and
stuff
going
so
and
if
we
miss
a
question,
just
go
ahead
and
repeat
it
again
and
then
we'll
go
through
it
that
way.
So
here
the
ground
rules
is
judgment-free
zone.
Everyone
had
to
start
from
every
us
from
somewhere.
So
there
are
no
dumb
questions.
There's
no
questions
that
are
a
waste
of
anybody's
time.
A
Whatever
it
is,
you
need
to
do
we're
gonna,
try
to
answer
it,
and
while
we
will
do
our
best
to
answer
your
questions,
the
panel
doesn't
have
access
to
your
cluster.
So
we
can't
really
like
live
debug.
You
know,
connect
your
cluster
or
anything
crazy
like
that.
But
what
we
can
do
is
at
a
minimum
at
least
get
you
moving
you
down
in
the
proper
direction
in
the
next
step.
Whatever
it
is,
you
need
to
do.
Panelists
you're
encouraged
to
expand
on
your
answers
with
your
experiences,
your
pro
tips.
A
You
know
your
production
experience,
developer
experience,
anything
that
might
help
audience.
You
can
also
help
us
out
as
we're
answering
these
questions
pacing
in
links
to
official,
Docs
blogs
or
anything
that
might
be
relevant
to
the
topic
at
hand.
It
feels
like
every
time
we
have
a
topic,
someone
from
the
audience,
always
plops
a
URL
of
a
tool
that
we've
never
heard
of,
and
that's
always
like
awesome.
A
So
that's
it's
always
fun
to
share
the
knowledge
and
then
what
I
do
is
we
collect
all
the
URLs
and
we
put
them
in
the
show
notes,
and
then
we
publish
those
in
the
forum
thread
so
that
we
have
them
looks
like
a
sari
getting
over
bronchitis
here
this
panel
is
made
up
entirely
volunteers.
So
if
you
want
to
rotate
into
this
panel-
and
you
just
let
me
know-
or
if
you
answer
a
bunch
of
questions
in
chat,
I
might
just
ask
you
to
hop
on
like
here
and
start
answering
questions.
A
I,
don't
want
to
say
help
people
get
the
knowledge
that
they
need
or
something
sorry.
The
chat
is
hopping
right.
Now,
it's
pretty
awesome.
We
have
people
from
Austin,
hello,
Josh
burkas
is
here
he
will
be
answering
the
database
question
that
you
will
all
always
ask
every
single
session
and
we
will
link
you
to
his
blog
post
about
running
databases
and
kubernetes,
actually
I
think
we
might
have
pinned
it
in
the
channel
by
now
and,
lastly,
we're
DevOps
people
too,
so
we
like
to
measure
our
metrics.
A
A
It
makes
it
easier
for
us
to
generate
things
that
are
useful
for
the
community,
so
we
really
appreciate
that
and
thank
you
very
much
panel
for
joining
us
would
also
like
to
thank
the
following
companies
for
supporting
and
the
community
wood
developer
volunteers
on
this
panel
Giants
warm
stock
ex
pusher
calm,
we've
worked,
vmware
university
of
michigan
red
at
Spectrum,
Daioh,
American,
Airlines,
utility
warehouse
and
Microsoft,
and
special
thanks
to
the
CNC
F
for
sponsoring
the
teacher
giveaway.
So
it
will
do
is
if
we
answer.
A
If
we
address
your
question
on
the
air,
we'll
put
you
in
a
little
little
contest
and
I
give
away
two
t-shirts.
Every
session
I'll
give
you
a
code
I'll
PM
you
after
I'll
give
you
a
code,
you
can
go
to
the
CNC
f
store
and
where
the
wonderful
kubernetes
shirt
that
we
all
forgot
to
wear
today.
I
am
not
wearing
I'm
wearing
a
different
kubernetes
shirt
today,
but
yeah
it's
like
the
official
kubernetes
shirt.
A
A
Certain
aks
ingress
the
three
works
as
expected,
using
the
nginx
ingress
controller
in
the
fourth
ingress,
doesn't
work
at
all.
I
only
get
time
out
to
using
the
public
IP
address.
I
can
access
the
service
using
port.
Forwarding,
though
so
I
know,
the
service
is
working
properly.
At
this
point,
I'm
really
stuck
on
what
to
look
at
to
figure
out.
What's
misconfigured
or
broken,
the
clusters
were
all
spun
up
with
terraform
and
the
English
controllers
were
installed
via
helm.
A
A
C
The
way
I
would
go
debugging.
This
is
to
check
first.
If
the
packets
are
coming,
the
requests
from
outside
world
are
coming
into
your
cluster.
That's
the
first
thing
I
will
check.
The
second
aspect
is
if
the
ingress
is
rightly
connected
to
the
service
that
he
is
talking
about.
So
if
it
is
connected,
the
best
thing
is
to
do
a
cube.
Cut'
will
describe
the
ingress
and
see
if
you're
disconnected
or
check
if
there
backends
are
healthy.
You
know
these
are
the
three
aspects
have
first
people
that
see
things
are
okay,.
F
A
Any
other
advice
here,
I
see
the
person
I
asked
a
question
is
typing
in
the
audience
feel
free
to
also
type
your
solutions
in
the
chat
we
like
to
get
as
much
opinions
from
people
as
we
possibly
can.
So
it
looks
like
they're
typing
I
do
want
to
run
a
question
by
the
panel
from
this
morning
that
we
were
not
able
to
answer.
A
All
right,
Simone
asked
us
a
question
this
morning
that
we
were
kind
of
stuck
on,
says
I've
been
creating
setting
up
Samba
inside
kubernetes
containers
and
I
wanted
to
ask
if
anyone
has
tried
or
has
tips
on
how
to
configure
samba
as
an
active
directory
clients
running
inside
of
kubernetes
and
most
of
the
answers
we
realized
this
morning
is
we
know
people
are
doing
Active
Directory
in
containers
I'm,
just
not
quite
sure,
with
Samba
on
Linux
workloads.
So
any
of
you
have
any
expertise
with
that.
That
would
be
a
good
one.
A
A
A
A
D
No
it's
silly,
but,
like
you
know,
you
can
try
deleting
the
ingress
controller
pod
and
seeing
if
it
comes
well,
yeah
like
I
I
would
I
would
do
like
normal
whack-a-mole
stuff
to
it.
For
a
few
minutes
before
debugging
it
I
know,
I
mean
I.
Think,
like
you,
gotta
figure
out.
If
traffic
is
getting
through,
the
load
balance
turn
it
into
the
ingress
controller,
and
then
that
is.
G
Okay,
part
of
the
question
I
understand
is
like
they
have
four
controllers
now
rest
running
with
English
and
helm
right,
so
the
three
are
running
and
the
fourth
one
is
not
running.
Is
it
the
question
that
we
are
taking
here?
Yeah,
yeah,
okay,
so
I
think
a
by
default.
All
the
I
mean
the
question.
Maybe
all
the
others
are
also
public
IP
or
all
the
other
three
are
doing
an
internal
IP.
G
So
if
other
three
are
doing
an
internal
IP,
they
might
be
working
fine
and
if
coming
through
the
AKS
load
balancer-
and
that
is
exposing
to
public
IP-
and
you
say
that
it's
not
working
as
expected,
then
it
must
be
outside
the
Cuba
notice,
and
that
is
the
a
case
load
balancer
which
you
are
using,
is
not
being
tied
directly
into
it
into
a
service
properly
or
something
inside.
But
it's
not
out
it's
outside
the
cube
of
it
is
it's
not
inside
the
cube
is
basically
you
miss
your
English
service
is
hitting.
G
Maybe
you
you
log
it
into
English
pod.
What
I
would
do
is
log
into
your
English
pod,
execute
to
the
part
and
try
to
hit
the
service
that
you
are
internal
to
the
kubernetes
and
if
you
are
getting
it
from
the
English
from
the
English
control,
pod
and
you're,
trying
to
read
it
properly.
That's
fine,
but
the
AKS
is
not
routing
the
public.
Ip
routing
may
not
work
properly
to
ingress
so
check.
There
are
two
things
that
routing
from
the
AKS
load
load
balancer
to
the
public.
G
A
All
right,
so,
let's
on
this
one,
a
little
bit
as
you
keep
typing
and
then
we'll
come
check
check
back
up
on
you
regularly.
We
have
a
question
from
Peter
welcome
Peter,
first
time
it
says
any
recommendations
for
doing
local
micro
service
development
in
kind
but
needing
to
access
resources
in
a
cloud-based
cluster
for
like
app
authentication,
authorization,
etc.
I
know
it's
a
vague
question.
A
A
A
So
we'll
go
I'm
type.
Next
question
is
from
angel:
welcome
angel,
we're
gonna
my
questions
around
HP,
a
I
found
this
article
about
horizontal
pot
Auto
scale,
but
I'm
still
missing
something
on
it
say:
I
want
to
set
up
HPA
and
there's
some
pre
setup
that
I
need
to
do
before
the
container
start
serving
requests
properly,
because
the
service
is
a
servicing
request
and
working,
but
I
need
some
extra
data
like
a
large
set
of
data
and
the
server
will
be
using
to
properly
serve
the
requests.
D
Yeah
I,
don't
know,
I,
think
I'm.
A
little
confused
about
I
mean
it's
something
like
generally
pre
setup
stuff.
You
can
do
via
an
init
container
and
stuff,
but
I
don't
know
I,
don't
know
how
that
interacts
with
this
is
stuff,
if
at
all,
like
I
feel
like
the
init
container
stuff.
Just
kind
of
like
goes
and
then
is
not
really
part
of
the
the
calculation.
D
A
G
Mean
if
you
can
elaborate
the
question,
because
HPA
is
or
you're
already
doing,
yeah
and
that
the
thing
is
you
want
to
run
some
scale,
some
data
inside
it
before
the
parts
come
up
or
what
is
that
they
want
to
do
because
services.
The
question
mentioned
that
the
services
are
coming
up
and
then
he
wanted
to
yes.
C
The
way
I
understand
the
problem
is
this
right:
I
think
they
want
to
load
some
data
before
this
container
stacks
up.
This
spot
starts
up
now.
One
of
the
ways
trying
to
do
is
through
using
the
unit
container,
but
it
doesn't
help
that's
what
he's
suggesting
so
so,
how
do
we,
how
you
should
get
it
done,
except
questioning.
H
D
G
G
Yeah,
what
to
name
mentioned?
That's
right
right
if
you
want
to
load
something
into
the
container
right,
something
external
to
the
volume
somewhere.
Some
like
some
use,
some
volume
like
class
reference
or
any
other
thing
which
on
to
the
right
to
the
volumes
and
before
your
continue,
comes
up
just
access
the
volume.
Some
title
right
that
that
or
load
that
that
rental
is
containers.
E
G
Yeah
I
think
he
came
back.
He
said
that
at
that
so
take
the
example
of
fingers
say
the
app
is
deployed
already,
but
the
app
depends
on
large
set
of
data
that
is
mounted
in
the
container,
but
the
data
is
in
the
physical
node
I
need
to
copy
the
data
into
the
notes.
So
he
says
he
got
some
ideas
now.
Yeah.
A
That's
good
and
we've
successfully
tricked
Bob
to
joining
us
to
turn
against
her.
The
question
so
he's
on
his
way:
quick
feedback
on
Peters
question,
which
was
any
recommendations
for
doing
local
micro
service,
dev
and
kind,
but
needing
access
to
resource
in
a
cloud-based
cluster.
So
burkas
says:
let's
see
so
Peter
needs
it
for
secrets
and
calls
to
other
micro
services,
and
that
says
I
know
it's
all
theoretical
I'll
need
to
set
up
ingress
public
ip's,
but
is
there
a
checklist?
A
Josh
says
there
really
isn't
but
Jorge
other
Jorge
not
me
recommends
that
for
something
a
setup
like
that,
you
might
want
to
look
at
something
like
scaffold
from
Google
and
josh
says
for
off.
I
could
see
VP
ending
to
the
cloud
to
use
the
same
off.
Ingress
will
be
tricky
and
then
Jorge
goes
on
to
say
exposing
your
kind
clusters
in
general
not
recommended
to
the
Internet.
Let's
see
yep
that
can
so
that
come
conversation
is
still
going
they're
talking
about.
A
Basically,
they
want
that
setup
for
end-to-end
tests
of
ingress
for
quickly
testing
local
changes,
so
yeah
I
keep
I'll,
keep
an
eye
on
that
conversation.
Those
of
you
just
joining
us
welcome
to
the
kubernetes
office
hours
feel
free
to
just
type
your
questions
in
slack
and
our
steam
panel
of
experts
will
try
to
answer
it.
Bob
is
here
for
one
question:
only
you
might
be
busy
for
other
stuff,
but
we're
hoping
if
you
had
opinion
around
angels,
HP
a
question
and
hopefully
have.
I
A
A
G
The
question
is
summarizing
is
that
they
have
an
HP
they're
leaking
their
setup
for
the
HP,
a
horizontal
part
of
scaling.
But
the
thing
is
that
before
the
parts
come
up,
they
have
some
local
data
to
be
written
to
the
parts.
So
is
there
any
way
to
make
it
faster
or
something
before
the
HP,
a
scales
up
or
something
like
that?
I.
I
A
A
D
I
A
J
D
I
I
The
other
thing
is
like
regarding
the
node
ideas
that
you
could
probably
taint
the
node
and
still
have
a
demons.
That's
scheduled
to
it.
So
that
way
that
prevents
scheduling
until
you
know
the
demon
set
or
something
is
done,
and
then
its
job
would
be
to
remove
the
taint
once
it
is
essentially
reached
a
happy
state.
Mm-Hmm.
A
A
A
See
what
everyone's
David
and
why
leaner
typing
looks
like
they
have
opinions
we'll
see.
Let's
move
on
to
the
next
HP
question
here
so
B:
let's
just
talk
HP,
let's
just
keep
talking
HP
a
I've
deployed
an
application
kubernetes
cluster
with
HP
a
enabled,
if
it's
more
than
60%
expected
to
scale
it's
working
fine,
but
if
there's
any
pending
pod,
the
scale
down
is
not
working.
A
Like
the
60%
expected,
are
you
saying
that
you
got
60%
of
the
things
that
we're
supposed
to
show
up
showed
up,
so
it
only
goes
to
60.
It
doesn't
go
past
the
60%
and
the
Lord
even
or
the
0%,
even
when
the
load
is
0%.
Also,
how
did
you
get
more
information
hold
on?
Let
me
see
Sarah.
Is
there
a
threat
all
right,
so
even
if
the
load
is
0%,
okay,
so.
B
A
A
I
A
C
Scaling
up
I
would
say
there
is
a
pending
pod,
which
probably
is
due
to
the
resource
limits.
The
scheduler
is
not
able
to
schedule
one
of
the
nodes.
There
is
a
race
condition
that
is
happening.
The
scale
up
is
it's
trying
to
scale
up
and
there's
a
pending.
The
bot
is
one
of
the
bot
is
pending,
because
there's
no
resource,
probably
and
that
kind
of
screws
the
whole
scale
down
process
as
well
seems
like
a
race
condition
that
is
happening
here.
G
Configuration
right
there
is
a
horizontal
court,
autoscaler
downscale
stabilization
time.
So
do
you
know
what
time
did
you
set
up
on
or
is
already
preset
on
that?
Because
the
down
when
you're
downsizing
the
parts
are
when
you're
scaling
down,
maybe
that
setting
maybe
for
five
minutes
or
so
that
is
being
said.
So
we
just
need
to
wait
for
at
least
five
minutes
or
so
to
just
scale
the
scale
it
down.
G
A
A
A
Yeah,
but
we
you
know,
we
can
definitely
run
those
questions
by
the
sig
windows
folks,
because
we
are
due
to
do
another
before
the
virus
and
kind
of
knocked
everything
off
schedule.
We
were
planning
on
having
a
cig
window
specific
office
hours,
so
stay
tuned.
For
that
we
might,
you
know
having
them
around,
we'll
be
able
to
have
that
expertise,
because
we
do
not
because,
as
I
learned
last
time,
just
because
someone
works
at
Microsoft
doesn't
mean
they
know
anything
about
Windows,
right,
Dave,
I
know
zero.
A
All
right,
our
next
question
is
from
hold
on.
We
didn't
put
that
in
the
sig
windows
is
where
I
learned
about
GM
si
you
know
I
would
I,
would
love
to
like
sit
down
with
them
and
actually
like
ask
them
and
talk
to
them
about
about
that
stuff
and
your
Nass,
your
fans,
anyone
actually
running
windows
containers
in
production
for
real,
because
I
run
into
people
who
say
they
are
and
when
I
asked
really
they're
all
like
no,
not
really.
A
We'll
see
about
that
all
right,
all
right,
we've
got
about
15
minutes
or
so
left.
So,
let's
power
through
some
of
these
questions
here,
sorry
Simone.
Hopefully,
I
can
get
one
of
them
to
come
onto
the
show
and
help
help
answer
these
questions,
but
I
will
definitely
pass
them
pass
that
along
to
them
all
right
David
as
says,
hey,
guys,
strange
question:
have
you
seen
a
docker
registry
run
inside
a
cluster
that
needs
to
pull
images
from
it?
A
A
G
A
C
A
Right
right
is
a
common
question:
cuz
Willie
says
on
the
edge
and
they
use
the
openshift
internal
registry
in
the
default.
Namespace
Nexus
on
the
edge
is
Bastion
host.
Interesting
because
I
could
see
you
running
a
separate
cluster
for
your
registry
right.
If
it's
like
a
really
large
deployment-
and
you
got
a
bunch
of
harbors
in
a
cluster
and
any
other
best
practices
tuna.
Have
you
ever
set
up
like
harbor
for
anybody?
How.
E
Do
we
not
from
ground
up,
but,
but
typically
we
do
see
folks
have
them
separated
from
the
class
there
yeah.
A
A
Interesting
and
Willie
said
yeah,
it's
one
of
those
things.
If
you
have
clusters,
you
can
already
have
to
have
a
registry
somewhere,
so
I
think
at
that
point
it's
the
it's
the
kind
of
optimizing
where
it
should
be,
but
that's
interesting
audience
where,
where
are
you
running
a
registry?
If
you
are
your
own
registry,
I'd
be
interested
to
know
what
people
are
doing,
that
one.
A
What's
the
setup
for
the
back-end
service,
it's
a
default
deployment,
yamo
file
for
the
back-end
service
and
they're
using
git,
lab
and
Josh
says
yeah,
we
kind
of
need.
If
you
could
toss
your
llamo
file
on
App,
a
spin
or
github
gist.
That
would
that
would
help
welcome
back
Brandon
in
the
meantime.
Are
there
questions?
David
says
it's
a
good
idea!
Thank
you
have
used
harbor
outside
the
harbor
outside
the
cluster,
but
it
would
be
nice.
A
There
was
a
way
to
pull
images
on
one
node
and
then
other
nodes
can
get
the
image
from
that
one
node,
rather
than
pulling
from
the
registry.
Oh
I
see
so
they
kind
of
want
a
more
peer-to-peer
as
opposed
to
all
the
nodes
grabbing
from
the
registry.
They
kind
of
want
the
nodes
to
be
able
to
peer-to-peer
an
image
between
themselves,
which
sounds
a
lot
like
yeah
I
was
like
I'm,
pretty
sure
harbor
can
do
this,
but
I'm
not
sure
I
am
not
the
harbor
expert,
but
stand
by
here.
There's
a.
E
A
D
A
A
A
Sorry,
there
goes
I've
got
a
problem
on
one
of
the
clusters:
I
manage
for
cron
job
pods,
occasionally
gets
stuck
in
container,
creating
status
and
never
recover
so
that
job
never
completes.
Has
anyone
seen
this
before
or
know
what
causes
it?
There's
no
information
about
events
that
occurred
on
the
pot
or
anything
it
just
hangs
and
container
creating
indefinitely
all
right.
So
how
can
we
start
debugging?
This
I.
D
E
A
C
C
C
A
A
G
A
E
C
A
F
C
F
You
get
very
weird
behavior
things
just
start
breaking
that
a
lot
of
times
doesn't
give
you
a
good
hint
just
from
the
kubernetes
logs
yeah.
But
if
you
look
into
I
would
look
at
that
OS
disk
performance
there,
especially
you
see
it
a
lot
with
like
cron
jobs,
where
people
are
spinning
up
hundreds
of
jobs
at
once,
yep.
A
Yeah
I'm,
just
wondering
if
it's
like
you
could
just
be
over
subscribing
at
disk
and
kubernetes,
is
actually
helping
you
out.
You
know
you
just
get
an
error
rate,
but
that
would
be
interesting
to
know,
but
hopefully,
pizza
that'll
give
you
some
places
to
look
at
where
you
might
find
some
bottlenecks
or
something
like
that.
A
Yeah
awesome
feel
free
to
keep
on
typing.
If
you
have
any
follow-up
questions
and
stuff,
we
will
try
to
get
to
them
as
fast
as
we
can
next
question
all
right.
This
is
going.
We
have
time
for
two
questions.
This
will
be
the
second
last
question
and
then
one
more
question
and
then
we
will
give
away
the
t-shirts.
So,
let's
see
you
Manuel
asked
us
hi.
How
would
you
proceed
if
a
customer
asks
you
to
deploy
a
kubernetes
cluster
in
a
completely
isolated
environment?
No
internet
connection
is
possible.
A
D
G
Link
so
I
think
someone
also
push
it.
It's
just
that
yeah,
it's
just
a
couple
of
vendors
out
there
who
does
offline
lists
of
Cuban.
It
is
should
it
just
that
you
just
need
to
pull
up
the
repositories
and
the
images
that
would
need
it
for
it
and
then
keep
it
on
from
and
then
try
to
run.
A
couple
of
commands
bunch
of
commands,
so
I
think
we
have
four
you're
provided
the
links,
one
from
the
gravitation
and
one
from
the
Genesis.
So
just
try
to
check
those.
A
A
G
H
A
Right-
and
there
is
what
there
is
one
more
but
we
are,
but
we
are
out
of
time,
so
we
will
just
go
back
to
chat
on
this
authentication
in
a
cluster,
giving
access
to
LDAP
for
my
ad
users
there's
a
thread
there,
so
everyone
helped
me
out,
but
either
way,
DevOps
a
mere
and
keep
mispronouncing
your
name.
You
want
one
of
the
kubernetes
t-shirts
for
the
day,
so
thanks
for
dropping
by
hope
what
we
were
able
to
at
least
like
get
you
started
in
a
certain
direction.
A
I
will
PM
you
afterwards
with
a
code,
and
you
can
do
that
and
the
other
winner
is
angel.
Vera
angel,
if
you're
still
listening,
you've
won
a
kubernetes,
shirt
and
yeah.
Yours
is
I
kind
of
want
to
know
about
Windows
containers,
it's
a
scooby-doo
mystery
for
sure,
but
absolutely
keep
an
eye
out
for
that.
We've
we've
had
people.
This
is
why
it's
important
you.
Let
us
know
how
you
like
on
this.
A
We
have
had
people
ask
for
a
window
specific
QA,
and
then
we
definitely
do
want
to
do
a
cig
one
with
all
of
Signet
work
as
well,
because
there's
a
lot
of
questions
around
everything.
That's
happening,
you're
on
network,
so
my
goal
is:
this
quarter
is
to
try
to
keep
these
going
every
month,
but
have
an
additional
highly
specialized
office
hours
so
that
we
can
say
hey
everybody.
This
is
gonna,
be
the
network
one
or
the
storage
one,
or
something
like
that.
A
So
if
you're
interested
and
stuff
like
that,
we're
always
interested
in
your
information
so
panel
any
last
comments
or
anything
really
glad
everyone
came
out
so
far.
The
West
Coast
traffic
has
been
good,
so
I.
We
think
it's
a
good
time
slot
compared
to
the
previous
one,
so
that
any
any
last
comments
anybody
how
y'all.
D
G
A
A
It's
always
the
third
Wednesday
of
every
month
and
I
always
announce
to
the
kubernetes
and
forums
on
the
kubernetes
users
channel
and
in
here
feel
free
to
stick
around
in
the
office
hours
channel
and
help
each
other
out
we
like
to
have
and
with
that
panelists.
Thank
you
very
much
for
joining
us
and
everybody
have
a
great
week.
You.