►
From YouTube: Kubernetes Office Hours 20210119 (EU 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://k8s.dev/events/office-hours
A
And
it
looks
like
we
are
live,
let's
give
it
a
second,
let's
see
if
there's.
If
there's
anybody
watching
wants
to
say
how
the
audio
sounds,
that
would
be
fantastic.
A
Welcome
everybody.
It
is
the
third
wednesday
of
the
month
in
2021.
I
know
it
feels
like
we
just
got
back.
Everyone
already
looks
tired,
but
thanks
everybody
for
listening
in
how's,
the
audio
sound
there.
If
you
can
give
us
a
a
quick
shout
out
on
the
on
the
slack
channel,
if
you're
listening
in
that
would
be
useful.
A
So,
let's
get
started,
I'm
gonna
do
a
little
intro
about
how
this
works.
We're
gonna
introduce
the
panel
and
then
we're
gonna
move
on
to
your
questions.
I
know
we
got
a
lot
going
on
in
the
world
today.
So
if
we
run
out
of
questions,
we're
going
to
go
ahead
and
just
talk
about
our
favorite
parts
of
kubernetes
coming
up
dan
says
that
the
audio
sounds
great.
Thank
you
all
right.
So
here
we
go
welcome
to
today's
office
hours,
where
this
is
our
live
stream.
A
Where
we
answer
your
user
questions
live
on
the
air
with
our
steamed
panel
of
experts,
you
can
find
us
in
office
hours
on
slack
if
you're
on
youtube
follow
that
text
there
down
below
all
right.
Before
we
begin.
Let's
start
by
introducing
ourselves,
I
can't
even
figure
out
an
order
because
there's
so
many
wonderful
volunteers
today,
so
let's
let's
go
with
our
regulars
first
and
then
our
new
people,
so
we'll
go
chris,
mario
pov,
rachel
cyan.
B
Hopefully,
I'm
out
of
practice
on
this
one,
so
my
name
is
chris
google
customer
engineer
for
google
cloud
canada's
public
sector
team
based
out
of
ottawa
ontario.
My
background
is
mostly
with
on-prem,
kubernetes
and
kind
of
the
general
kubernetes
ecos
ecosystem,
so
hope,
looking
forward
to
today,.
A
C
Hey
everybody,
my
name
is
mario
lauria.
I
actually
used
to
be
with
stockx
based
in
downtown
detroit.
I
just
recently
made
a
shift
to
carta
based
out
of
san
francisco
working
on
capital
management
investor
relations.
If
you
have
stock
options,
you
probably
know
us
doing
much
of
the
same
in
the
kubernetes
realm
devops
sre
kind
of
bringing
everybody
together,
focusing
on
developer
experience
and
optimizing
kubernetes
for
our
work
load.
C
So
I
I
have
a
lot
of
background
in
auto
scaling,
kind
of
pod
life
cycle
and
everything
that
comes
into
you
know
the
developer,
experience
of
deploying
from
from
github
to
through
pipelines
and
kubernetes
safely,
so
look
forward
to
a
great
session
today.
D
Cool
hello,
I'm
pablo
swarsatskas,
I'm
from
lithuania.
People
typically
call
me
pav,
I'm
a
certified
kubernetes
administrator
and
I
know
a
lot
about
monitoring.
Observability,
sre
and
devops.
E
Rachel
leakin,
I'm
a
field
engineer,
kubernetes
field
engineer
here
at
vmware.
My
specialty,
I
guess,
would
be
opa
open
policy
agent
and
just
general
kubernetes
and
cloud.
F
G
Hi
I'm
saying
director
of
technical
evangelism
at
sivo
and
I'm
a
cncf
ambassador,
ck
ckd
certified
and
work
mostly
on
kubernetes
and
cloud-native
technologies.
H
I
Hello,
everyone,
I'm
dan
papandrea
people,
call
me
pop.
I
am
the
lead
for
community
and
ecosystem
for
systig,
specifically
the
falco
project,
I'm
a
certified
kubernetes
administrator
as
well
as
I
run.
A
little
show
called
the
podcast.
A
Yeah,
actually,
both
david
and
dan
run
podcasts
and
I'm
gonna
ask
them
to
pop
the
urls
there
in
slack.
If
you
would,
I
enjoy
watching
these
two
and
part
of
the
reasons
they're
here
is
I'm
the
last
one
to
introduce
us
myself.
I'm
george
castro
I'll,
be
your
host.
I've
been
hosting
this
show
for
two
years
working
at
under
heptu
on
vmware.
A
However,
I
have
recently
left
to
go
work
at
unusual.vc
and
with
the
folks
that
are
ricto,
which
you
might
recognize
from
the
kubeflow
community,
so
I'm
bringing
in
dan
and
david
to
check
out
the
show,
because
we
do
want
to
keep
the
show
going.
So,
as
some
of
you
know,
this
show
is
run
entirely
by
volunteers.
A
So
anyone
that
you
see
on
this
panel
it's
a
great
way
to
participate.
It's
a
one
hour
a
month
commitment.
It
doesn't
even
need
to
be
an
hour
a
month.
We
just
need
enough
people
to
do
the
show.
So
if
we
have
more
than
three
or
four
people,
we
just
slap
the
panel
together
on
the
monday
before
and
then
we
get
it
going.
So
this
is
a
great
place
to
start
as
I
was
leaving.
A
I
was
kind
of
summarizing
all
the
things
that
we've
done
in
the
show
and
the
amount
of
contributors
that
have
started
in
this
program
and
and
moved
on
to
do
things
in
other
parts
of
the
kubernetes
project.
We've
got
an
awesome,
batting
record.
So,
let's.
I
A
A
A
And
audience
this
has
all
been
possible
for
you.
We've
actually
had
some
rough
episodes.
We
had
an
episode
where
mario
was
the
only
person
here
and
anna
answered
questions
from
the
audience
by
himself
for
45
minutes,
and
that
was
not
awesome,
but
you
know
we
power
through.
We
make
it
through
speaking
of
poverty,.
A
C
That
meeting
and
how
to
put
it
on
other
people
to
solve
questions
and
problems
for
me
and
so
no
to
your
point,
pop
george,
is
what
brings
it
all
together
like
all
of
our
fancy
titles,
all
the
stuff
we
do.
George
is
the
guy
and
thanks
to
him,
and
everyone
give
him
some
hearts
on
slack,
please
and
I'll.
Let
him
get
to
the
show
now
sorry.
A
A
If
you
have
your
questions,
already
feel
free
to
start
typing
them
in
the
channel
and
what
we'll
do
is
we'll
put
them
in
a
little
queue,
and
then
we
will
answer
them
as
we
get
the
time
at
at
the
end,
we'll
grab
all
the
urls
and
put
it
together
and
it'll
be
nice.
So
before
we
start
here
the
ground
rules,
this
is
a
kubernetes
event,
so
the
code
of
conduct
is
in
effect,
so
please
be
excellent
to
each
other.
This
is
also
a
judgment-free
zone.
A
Everyone
had
to
start
from
somewhere,
so
you
can
help
out
your
buddy
by
having
a
supporting
supportive
environment
in
the
channel,
and
we
don't
consider
any
question
too
easy
or
too
simple,
or
things
like
that.
So
please
just
remember
that
everyone
has
different
skill
sets.
This
is
a
very
complicated
piece
of
software
and
the
ecosystem
around
it
can
be
very
complicated,
so
everyone
has
to
start
to
somewhere.
A
So
we
try
to
have
a
positive
environment
here
and
while
we
will
do
our
best
to
answer
your
questions,
the
panel
doesn't
have
access
to
your
cluster.
So
live
debugging
is
going
to
be
off
topic,
but
we
will
do
our
best
to
get
you
moving
down
the
next
step
that
you
want
to
be
so
we're
not
going
to
ssh
into
your
cluster
and
go
dig
around
and
things
like
that.
A
Yeah
and
panelists,
you
encouraged
to
expand
on
your
answers
with
your
experiences
and
pro
tips
as
we
answer
questions
you'll
all
dive
in
I'll
kind
of
emcee,
the
whole
thing,
but
what
people
want
to
hear
is
your
production
experiences
your
failures,
just
you
know
anything
about
a
given
topic.
That's
why
we
have
you
know
we
have
this
we're
not
here
to
kind
of
show
off
how
great
we
are
things
like
that,
we're
here
to
talk
about
the
real
stuff
that
people
deal
with
day
to
day
audience.
A
You
can
help
us
out
by
pasting
urls
to
your
official
docs
blogs.
You
might
have
seen
or
anything
that
might
be
relative
to
the
topic
at
hand.
One
of
the
best
things
about
being
on
this
show
is
every
single
week,
someone
from
the
audience
when
we're
talking
about
something
people
are
whacking
in
urls,
the
github
repos
of
stuff,
and
we
find
a
new
tool
just
about
every
month
that
someone
else
hasn't
heard
of
that's
awesome,
that's
powerful!
That's
really
great!
What
I
do
is
at
the
end
of
the
show.
A
I
grab
all
the
urls
and
we
stick
them
in
the
show
description
with
the
hope
that
someday
you
know
someone
who's
stuck
with
this
problem
might
run
into
a
tool
that
helps
them
out
so
audience
this.
Your
participation
is
crucial.
We
love
to
see
all
these
hearts
and
things
like
that
scrolling
in
slack.
So
please,
please
help
us
out
by
giving
us
the
urls
and
resources
that
you
use.
You
can
always
post
your
questions
on
discuss.kubernetes.io.
That's
where
I'll
be
sticking
all
the
show
notes
and
things
like
that.
A
A
If
you
want
to
retweet
that
that'll
help
us
out
and
then
lastly,
this
panel
has
made
entire
volunteers
if
you
want
to
rotate
in
just
let
us
know
we
love
to
have
new
people
rotate
in
and
help
us
out.
Okay,
with
that,
let's
get
started,
I'm
going
to
check
out
the
questions
here,
who's
doing
the
questions,
doc
how's
everyone
doing
today,
I
see
someone's
pasting
it
in,
so
I'm
just
going
to
read
off
of
that.
A
So
usually
what
we
do
is
someone
from
the
panel
just
whacks
all
the
questions
into
a
document,
and
then
I
read
those
if
I
have
skipped
over
your
question
just
at
me
right
in
there
in
the
channel-
and
I
will
get
back
to
you
all
right.
The
first
one
comes
from
mustafa
al
bawami.
A
H
Yeah,
okay,
it's
such
a
tough
question.
You
know
bare
metal
clusters
have
a
lot
of
challenges
that
you
won't
get
from
other.
I
don't
want
to
say
traditional,
that's
probably
the
wrong
word,
but
you
know
other,
maybe
managed
provider
clusters
or
virtualized
clusters
like
I
guess.
One
of
the
major
challenges
is
that
is
the
std
component.
You
know
you
are
running
a
stateful
workload
on
bare
metal
and
it
requires
tender,
loving
care,
it
needs
hugs.
H
It
needs
to
be
nurtured,
it
needs
to
be
backed
up
and
you
really
need
to
make
sure
you're
drilling
home
those
disaster
recovery
scenarios,
and
if
you
lose
your
ncd,
you
don't
have
a
cluster
anymore,
so
you
know
really
put
in
the
time
and
effort
to
make
sure
that
that
at
least
that
one
component
is
steady,
you
can
recover
from
that
with
that,
at
least
so
that's
that's.
The
first
bit
of
advice.
G
Yeah
and
let's
add
on
the
the
bare
metal
cluster
setup
should
go
as
as
normal
as
you
do
the
cube
adm
setup,
so
so
that
that
should
be
fine
and
there
are
a
lot
of
blogs
where
people
have
done
bare
metal,
clustering
and
the
setup,
and
they
have
also
a
few
of
the
blogs,
will
also
mention
I'm
trying
to
find
them
also
mention
their
experience
over
running
kubernetes
in
production
on
bare
metal
and
how
that
works.
G
So
monitoring
is
another
very
big
piece
that
you
have
to
take
care
when
you're
running
on
bare
metal
and
not
the
managed
services,
because
that
is
where
your
things
go
wrong
and
you
need
to
think
like
what
happens
when
and
you
need
to
predict
the
failures.
So
all
these
things
play
a
very
important
role,
so
you
have
to
have
proper
monitoring
on
your
atcd
and
on
the
cluster,
and
the
setup
should
be
fairly.
G
I
mean
we
should
not
say
simple,
but
it
should
be
fairly
in
in
the
steps
that
they
are
in
the
cube.
Adm
and
the
networking
should
be
taken
care
of
very
precisely
in
the
bare
metal.
H
Yeah,
I
guess
there's
a
couple
more
challenges
worth
mentioning
with
a
bare
metal
cluster.
If
you're
going
to
be
doing
a
highly
available
control,
plane,
you're
going
to
have
to
be
comfortable
with
you
know,
virtual
ips
and
been
able
to
distribute
your
traffic
to
your
the
leader
of
the
control
plane.
Csi
is
obviously
a
challenge.
H
People
generally
want
to
run
stateful
workloads
on
to
give
a
nice
cluster,
whether
that's
advised
or
not,
but
you
know,
look
at
the
static
provisioner
from
the
kubernetes
six
repository,
which
provides
a
way
to
expose
your
nvme
and
other
super
fast
disks
from
the
bare
metal
to
kubernetes.
That
way,
and
if
you
want
to
go
with
something
replicated
distributed,
you
know,
you've
got
really
great
projects
like
open
ebs,
rook,
cef,
etc.
E
A
H
Nct
on
dedicated
hardware
outside
of
your
cluster,
it's
entirely
up
to
you.
I
do
generally
prefer
to
keep
it
as
part
of
the
control
plane
and
trying
to
stick
to
the
same
type
of
deployment
that
the
cluster
api
would
give
us.
I
think
what
they're
promoting
and
the
tooling
that
they're
providing
is
what
we
should
be
aiming
to
replicate,
if
not
using
the
cluster
api
and
then
adopting
those
same
patterns
as
well
for
our
own
bare
metal
clusters.
I
Shout
out
to
jason
d
tiberius
on
that
amazing
resource
in
the
community,
but
yeah
again
again,
I'm
kind
of
my
two
cents
on
them
and
I've
seen
a
lot
of
implementations
and-
and
I
think
again
it's-
I
definitely
agree
with
the
staple
workloads
there.
You
know
your
storage
paradigms
have
to
be
figured
out.
Your
monitoring
paradigms
definitely
have
to
be
figured
out
in
terms
of
high
ability.
I
You
know,
networking
decisions
are
key.
If
you're
doing
something
incorrectly,
you
can
basically,
you
know,
kill
a
bare
metal
cluster
versus
if
you
had
a
managed
cluster
through.
Like
one
of
the
you
know
the
hyperscalers,
you
know
they're
going
to
think
of
that
for
you
right,
and
so
you
know
definitely
those
considerations
need
to
be
there.
A
F
Yeah
metal
ob
seems
to
be
the
most
for
on-prem
implementations.
H
H
Bgp
supported
load,
balancing
and
virtual
ips,
which
is
a
really
really
cool
project,
works
really
well
and
worth
checking
out.
H
And
I
guess
one
more
kind
of
option
for
people
running
bare
metal.
You
know
if
you're
not
familiar
with
the
tinkerbell
project,
that's
the
cncf
sandbox
project
for
provision
and
bare
metal
machines,
and
that's
going
to
help
you
along
the
way.
There's
also
cap
t,
which
is
a
work
in
progress
which
is
a
cluster
api
implementation
on
top
of
tinker
bell.
So.
A
H
A
Awesome
so
good
to
know
lots
of
bare
metal
stuff
now
metal
lb
is,
is
that
I
should
be
okay.
If
I'm
using
it
right.
Is
it
just
one
of
those
things
where
people
are
moving
on?
Is
it
like
moving
on
from
the
old
nginx
default
configs
to
like
new
stuff,
or
is
there
still
going
to
be.
H
So
many
llb
and
cuba,
they
just
provide
like
a
single,
a
single
external
addressable
ip
address
for
your
cluster,
and
then
you
can
adopt
your
regular
and
graduates
from
there.
If
you're,
using
llb,
now
you're
still
saying
like
okay,
okay,
nothing
really
is
going
to
change.
It's
just
not
really
moving
forward.
Okay,
anymore.
Okay,
all
right.
A
A
H
A
Yeah
yeah
from
I
I
I'm
a
basic
user,
so
I
don't
have
you
know
multi-regional
concerns
or
anything
like
that
and
I'm
finding
qubit
q
badminton
is
just
perfect
for
me
between
cube
admin
and
kind
and
mind
runner
says
this:
is
the
year
of
kubernetes
bare
metal?
Was
it
ever
not
that
year?
A
All
right
next
question
comes
from
dinesh
shanmugem
says
I
perform
manually
certificate
renewal
on
my
kubernetes
master,
using
the
cube
cuddle
alpha,
renew
all
which
all
which
did
update
all
the
certificates
post,
that
I
did
a
restart
on
my
cubelet
service
but
looks
like
my
cluster
went
down.
I'm
able
to
see
my
pods
using
cube,
cuddle
get
pods,
but
I'm
not
able
to
schedule
any
new
pods.
A
I
checked
the
kate's
api
server
and
docker
container
logs,
and
I
see
the
following
issue:
authentication
error,
x,
509
certificates,
signed
by
unknown
authority
certificates
and
possibly
because
of
qtr
rsa
verification
error,
while
trying
to
verify
candidate
authority
certificate.
A
This
error
is
in
the
channel
if
you
all
screw
up
any
anyone
have
an
idea
on
this
one.
I
kind
of
asked
cardi
to
take
a
look
before
we
started,
but
I
don't
know
if
we
did
this
is
this:
is
this
one's
queued
up
from
over
the
break?
Anyone
have
an
idea
on
this.
One.
H
Let's,
let's
so
the
api
server's
healthy,
I
think
they've
said
that
the
fact
that
they're
running
cube
control
and
they
can
get
the
pots
is
fine.
The
cubelet
is
not
healthy.
It's
complaining
about
an
insecure
certificate
that
usually
means
that
the
ca
ser
is
not
provisioned
properly
on
the
machine,
or
maybe
it's
expired.
Maybe
it's
not
part
of
the
bundle.
Maybe
it's
not
been
passed
through
as
a
couple
parameters.
H
So
you
know
the
first
place
to
start
would
be
probably
looking
at
the
systemd
unit
file,
making
sure
your
cubelet
has
the
correct
parameters
that
you
need
to
pass
in
the
the
ca,
the
certificate
and
any
flags
that
are
needed.
I
don't
remember
them
off
the
top
of
my
head,
but
I
hope
that
gives
them
enough
to
get
started.
F
Also,
if
you
just
updated
just
the
api
cert,
he
should
probably
consider
making
sure
that
all
of
the
other
components
search
or
updated-
you
know
the
scheduler
and
all
those.
A
D
Typically,
when
I
do
this,
I
if
you'd
apply
everything
through
cube
adm,
then
I
think
basically,
that
command,
which
cube
qbdm,
I
think,
like
rotate
certificates
or
something
like
that,
which
will
basically
go
and
recreate
everything
for
you
and
make
sure
that
your
certificates
are
up
to
date.
I
A
A
Let's,
let's
cover
let's,
let's
cover
whatever
whatever
certs
y'all
may
have
so
ckas.
F
Oh,
yes,
I
have
a
cka
and
a
ckad.
I
got
those
searched
to
pretty
much
confirm
my
understanding
of
the
technology.
It
doesn't
necessarily
add
benefit
for
my
current
job
outside
of
giving
me
a
little
bit
more
credibility,
but
I
use
it
as
a
confirmation
to
under
make
sure
that
I
understand
the
technology.
I
can.
F
I
took
the
ckad
first
and
then
the
cka
afterwards
and
I'm
currently
working
on
the
cks.
So
I'm
hoping
to
have
that
before
the
end
of
the
year
at
minimum.
C
Yeah
I
just
want
to
mention
that
the
the
certified
kubernetes
exams
are
unlike
many
other
exams
which
are
basically
multiple
choice,
questions
or
you
have
to
type
this
exact
statement
in
a
cisco,
router
terminal
sort
of
thing.
So
I'm
someone
who
actually
failed
the
ccna,
which
is
the
cisco
certified
network,
associate
exam
back
in
the
day
and
got
got
my
aplos
and
a
few
other
ones
like
that,
and
those
those
really
aren't
practical
right,
they're,
they're,
more
theoretical,
they're
more.
What
could
happen?
They're
more,
you
know
pick
one
of
these
four
options.
C
The
cka
ckad
cks
are
all
hands-on
interactive.
You
are
solving
problems
dynamically,
and
so
I
I
think
in
terms
of
the
realism,
what
you
get
out
of
them
is
so
much
more,
so
I
actually
have
my
cked
I'm
working
on
my
cka
right
now.
I
don't
know
why
I
waited
so
long,
but
again,
it's
actually
more
of
like
a
fun
challenge
to
make
sure
I'm
kind
of
at
a
level
and
learning.
C
Sometimes
it's
you
never
like
feel
like
you're,
really
good
at
something,
and
then
you
go
back
and
read
a
book
about
it
like
a
beginner
book
about
it,
you
start
at
zero
and
you're.
Just
like
you
know
what
it
was
good
to
go
over
that
actually
right.
It
kind
of
reaffirms
things
in
your
mind,
that's
kind
of
what
I'm
doing,
and
hopefully
the
cks.
After
that
I
love
the
exams
and
I
really
highly
recommend
them
to
everyone.
There's
lots
of
discounts
that
we're
just
kind
of
around
with
the
holiday
season.
I
Here
he
and
waleed
who's.
Another
walid
is
one
of
the
greatest
resources
out
there
in
terms
of
just
you
know,
putting
together
all
this
data
about
all
the
exam
questions
where
to
study,
and
all
that
I
think
everybody
who
probably
passed
their
ck
on
this
panels
probably
use
a
walid.
You
know
resource,
but
sayon
and
waleed,
like
literally
attacked
every
part
of
it.
It
is
one
of
the
most
useful
youtube
like
you
know,
things
that
present
webinars
I've
ever
seen.
So
am
props
on
that,
then.
F
Also,
when
you're
taking
this
the
test,
make
sure
that
you
pay
very
close
attention
to
the
questions
and
follow
the
context
and
suggestions,
because
you
could
end
up
troubleshooting
on
the
wrong
cluster
and
that's
not
something
that
you
really
want
to
do.
I
And
they
provide
the
context
in
the
in
the
thing,
so
you
could
basically
you
know
copy
paste
it
and
the
other.
You
know
useful
thing
is
the
cheat
sheet
right:
the
cheat
sheet
having
aliasing
you
know
again
really
super
useful.
Instead
of
having
to
do
cube,
ctl
having
to
type
k,
it's
much
more
simple,
especially
when
you
have
a
you
know,
time
time
restraint
so
yeah.
F
Ensure
that
your
environment
is
stable,
also
that
your
desk
is
clear
and
all
that
stuff
there
they
have
used
showing
your
room,
so
you
know
make
sure
all
that's
there
make
sure
that
you've,
you
know,
taken
care
of
everything
prior
to
starting
the
test.
H
You
don't
just
clear
your
desk.
You
literally
have
to
remove
everything
I
put
it
in
a
drawer,
get
out
of
the
room,
be
prepared
to
move
your
laptop
camera
about,
if
you
can
to
show
them
the
environment
and
and
remember
you're
allowed
the
kubernetes
documentation
to
open
it's
a
invaluable
piece
of
advice.
Hopefully,.
I
F
I
F
Other
thing
it
takes,
they
have
all
the
yellow
icons
across
the
bottom.
They
force
you
to
get
rid
of
them
all,
except
for
like
two,
the
browser
and
your
one
additional
one.
So
you
be
prepared
to
have
your
workstation
slightly
disrupted
yeah.
B
A
B
I
found
another
thing
I
found
really
helpful
was
spend
a
lot
of
time
in
the
doc
docs
beforehand,
because,
while
you
do
have
access
to
it,
you
want
to
know
where
those
answers
are
generally
ahead
of
time.
So
you
don't
spend
all
your
time
searching
through
the
docs
to
find
the
answer
you
need.
So
if
you
understand
the
structure
of
the
kubernetes
documentation
makes
it
a
lot
easier
to
get
the
answer
and
not
waste
time
through
the
search.
G
J
G
I
Invaluable,
like
you
know
again,
the
the
cloud
guru
classes,
the
the
udemy
stuff-
really
really,
because
what
you
know
the
mock
test,
fantastic
again.
No,
you
know
I
have
no
affiliation
to
any
of
those,
but
like
they
were,
the
udemy
class
was
fantastic
shout
out
to
the
crew
crew
over
there.
F
A
All
right-
and
I
got
a
piece
of
advice:
if
you're
going
to
make
your
own
cheat
sheet
and
stuff,
do
not
post
that
on
the
kubernetes
official
forums,
because
we
can't
see
that
and
the
cncf
could
see
that
and
they
will
catch
you
cheating.
I
hate
that
I
have
to
say
that.
But
please
don't
you.
A
Yeah
yeah,
so
we
had
to
get
that
blocked.
Yeah
we
used
to
go
in
there
and
be
like
wow
the
entire.
You
know
people
used
to.
I
don't
know
what
people
do.
Pavel
has
a
follow-up
question.
Then,
what's
the
opinion
and
thoughts
on
the
cks,
anyone
finished
that
one.
Yet
that's
the
newest
one
right:
the
security
one.
B
Yeah
I
just
started
studying
for
it
the
course
that
goes
with
the
cert
is
really
nice.
I'm
about
two
chapters
in
so.
G
Yeah
cps
cks
is
pretty
heavily
focused
on
on
obviously
security
and
follow-up
of
cka,
so
you
have
to
have
cka
first
and
then
only
you
can
sit
for
cks.
It
has
third-party
tools
that
that
the
questions
will
come
from,
for
example,
says:
take
falco
trivia,
cis
benchmark
q
bench.
So
all
these
tools
are,
I
mean,
questions
related
to
these
tools
will
also
be
there
and
you
have
to
I
mean
the
documentation
is
allowed.
G
I
mean
you
are
allowed
to
open
the
docs
for
these
tools
as
well,
so
you
are
allowed
kubernetes,
plus
falco
dogs,
cystic
stocks
and
the
other
ones
which
are
which
are
there
from
which
the
questions
will
be
there
and
the
resources
again.
The
valid
has
a
very
good
repository
kim
has
very
good
course,
plus
killer
dot,
sh
with
a
lot
of
questions,
and
there
is
a
series
on
medium
that
he
has
started
as
well.
Oh.
A
A
A
A
A
C
I
I
linked
her
her
blog.
This
is
from
years
ago
that
she
did
this.
A
lot
of
it
depends
on
the
system
you're
running
on
tapping
into
the
correct
device,
but
I
think
it
might
be
a
little
out
of
scope
because
it's
not
generally
something
you're
going
to
do
inside
kubernetes,
but
it
is
definitely
as
cool.
It's
not
going
to
work
the
best
there's
just
definitely
some
things
you're
going
to
need
to
tune
with
relation
to
how
those
containers
are
using
system
resources,
and
things
like
that.
C
So
I
I
would
really
venture
that
you
kind
of
ask
the
question:
what
am
I
really
trying
to
solve
and
what
am
I
really
trying
to
do
with
with
this
option?
You
know
a
lot
of
browsers
have
more
isolation,
features
and
security
features
nowadays
that
it's
not
necessary
to
run
a
browser
there
or
you
know
it's
easier
to
just
kind
of
download
the
zoom
binary
and
get
it
going.
So
it
really
depends
on
what
you're
trying
to
do,
but
yeah
there's
definitely
resources
out
there.
C
You
can
search
github
for
other
people
who
have
four
just
for
sales
containers
and
then
made
other
versions
as
well.
So.
A
Yeah
yogi
says
gooey
containers,
probably
misconfigured
x,
environment
variables
in
port
mapping
for
sure,
okay,
moving
on
more
questions
on
on-prem,
see
david.
Everyone
knew
you're
coming
today.
Mustafa
asks
what
is
the
best
practice
to
set
up
kubernetes
on-premise
for
an
h-a
cluster
of
four
to
eight
nodes.
A
This
sounds
similar
to
what
we
were
recommending
before
and
let
me
let
me
expand
their
question
here.
A
little
bit
at
what
point?
Do
you
go
past
quote:
unquote
normal
cube
admin
like
when?
Do
you
determine
okay,
I'm
gonna
do
full-on
cluster
api
onto
something
like?
Is
there
a
number
in
your
brain,
or
is
it
just?
A
H
Yeah,
so
I
I
don't
think
we're
in
a
position
where,
for
bare
metal
at
least
maybe
adoption
cluster
api
is
the
default.
Yet.
A
H
There
is
a
provider
for
equinix
metal.
There
are
some
caveats
about
the
way
that
cluster
api
works.
That
makes
again
stateful
workloads
a
little
bit
of
a
challenge,
and
that
is
that
the
way
the
cluster
api
handles
upgrades
is
by
destroying
the
old
machines
and
bringing
up
new
machines.
So
any
any
databases,
any
workloads
or
states
whatever
disk
wise
is
on
those
old
machines
is
going
to
disappear,
which
means
you
have
to
have
all
of
this
tooling
in
place
to
either
make
sure
that
it's
replicated
and
depending
on
the
size
of
that
data.
H
You
know,
data
has
gravity
it's
difficult
to
move
that
could
take
a
lot
of
time.
Your
upgrade
process
could
become
rather
painful,
and
so
it
may
be
that
you
use
cluster
api,
mostly
for
the
ephemeral
and
stateless
workloads,
queue
workers
and
a
whole
bunch
of
other
things,
and
then
maybe
run
a
node
pool
or
just
a
separate
cluster
for
your
stateful
workloads
or
run
them
outside
of
kubernetes.
That's
entirely
up
to
you.
Try
make
the
right
call
yeah
if
you
get
it
wrong,
fix
it
like
there's,
there's
no
right
and
wrong
there.
H
A
Okay,
that
brings
us
on
a
mind,
runner's
question
semi-related
here,
they're
doing
on-prem
with
console
as
a
service
mesh
and
asking
for
load
balancer
recommendations.
So
we
have
metal,
albeit
we
we
talked
about
cube
vip
is
is:
are
we
missing
any
or
those
are
two
major
ones?
Is
it
worth
looking
at
at
any
other
two,
any
other
thing
other
than
those
two.
I
guess
is
my
question.
A
H
I
think
cuba
is
driven
of
a
requirement
for
something
to
replace
my
llb
now
that
it's
not
really
actively
maintained
the
challenge
with
the
bare
metal
load.
Balancing
is
that
you,
you
typically
need
a
load
balancer
that
speaks
bgp.
It
has
to
be
able
to
advertise
to
a
local,
asn
or
global
esm
where,
where
the
ip
address
routes
to
so
it's
not
like
you
can
just
pick,
you
know
nginx
or
aj
proxy
or
anything
else
that
you
might
be
using
in
that
fashion
and
deploy
it
as
your
load
balancer
for
the
cluster.
H
I
I
You
know
after
the
fact
and
stuff
like
that,
so
that
might
be
of
use
again,
they're
very
fledgling,
but
there's
a
lie:
seeing
a
lot
more
projects
that
are
doing
more,
like
load,
balancing
functions
or
at
the
cni
level,
like
looking
to
include
a
little
bit
more
of
that
functionality.
So
I
mean
it's
sky's
kind
of
wide
open,
almost
like
evp
functions,
a
lot
of
people
just
looking
into
it
and
growing
and
growing
and
growing
right.
I
Tonight
can
I
do
the
time
with
lannister
there's
a
tool
for
every
task
and
there's
a
task
for
every
tool
at
the
end
of
the
day,
if
you're
deploying
to
you
know
something,
that's
centric
to
you
know,
you
know
that
run
time.
If
it's,
you
know
container
d
again
props
to
file
estus,
I
think
they
do
an
amazing
job.
It's
kind
of
the
default
for
a
lot
of
the
implementations
that
are
out
there
and
yeah.
I
From
a
performance
perspective
and
all
of
that,
but
again
no
allegiant
cryod,
you
know,
is
another
kind
of
runtime
that
that's
useful,
for
you
know
if
you're
running
workloads
in
red
hat
or
or
beyond
right,
it's
like.
G
C
One
of
the
options-
probably
yes,
so
eks,
also
is
at
least
providing
the
ability
to
do
container
d.
I
don't
know
if
they're
doing
it
by
default,
yet
I'm
not
sure
what
they're
doing
with
the
latest
cluster
versions.
But
I.
F
I
Correct
and
then
like
ibm
another
they,
I
think,
they're
the
first
to
default
to
container
d,
but
others
I
think,
have
that
option
and
it's,
I
think
it's
either
going
to
be
by.
I
think
aks
by
default,
is
using
container
d,
so
I'm
pretty
sure
yeah,
because
I
think
we
had
to
compile
that
for
remember.
The
docker
shim.
A
A
G
A
Container
d
right:
what
about
rangers
rancher.
G
But
you
have
the
option
to
switch
to
docker,
so
if
mirantis
has
said
that
they'll
support
docker,
so
the
all
the
downstreams
who
are
trying
to
match
up
with
the
upstreams
they'll
also
have
to
support
docker.
A
Yeah
yeah
and,
of
course,
the
actual
docker
shim
implementation
is
now
maintained
by
mirantis.
So
if
you
really
want
to
stay
on
that,
good
luck,
all
right
next
question
comes
from
pavel
says:
what
do
you
think
about
the
elastic
license?
Does
moves
like
this
affect
the
kubernetes
users
or
a
future?
I
don't
think
it
has
any
effect
at
all.
What
is
what
does
the
panel
think?
I
don't
care,
what
license
your
software
on
top
runs
this.
F
L
F
H
Yeah
the
sl
sspl
license,
I
think,
is
really
targeted
at
people
that
want
to
take
mongodb
or
elasticsearch,
I'm
not
familiar
with
any
other
projects.
I've
adopted
it
yet,
but
anyone
who
wants
to
take
those
two
projects
and
sell
them
as
a
commercial
service
or
run
them
as
a
sas
business.
It
shouldn't
affect
anyone.
That's
running
elasticsearch.
A
A
H
I
know
a
lot
of
the
slack
simon
people
are
starting
to
have
these
conversations
and
look
at
alternatives
to
elasticsearch,
and
I
think
that's
it's
a
mailed
over
reaction.
I
think
you
know
elastic
as
a
company
in
mongodb
is
a
company
are
just
trying
to
protect
their
their
income
stream
and
the
large
companies
they
probably
don't
need
it,
but
at
the
same
time
you
know
we're
moving
away
from
this
generation
of
enterprise
licenses
towards
sas
based
businesses
and
fluxdb
is
a
good
example
like
invoicedb
enterprise,
there's
no
influx
db.
H
Clouds
elastic
are
down
to
managed
cloud.
Mongodb
have
atlas
like
this
shift
means
that
they
can't
really
get
away
as
open
core
anymore,
so
instead
they're
going
to
fill
open
source.
With
this
caveat
that
says,
hey
we're
going
to
give
this
all
away
for
free.
You
just
can't
run
it
and
sell
it,
and
I
think
it's
all
right,
there's
obviously
arguments
about
whether
is
it
still
open
source
because
it's
not
osi
approved,
but
I
think,
as
an
end
user
of
those
products
keep
using
them.
A
A
What
is
your
workflow
with
opa
pre-validation
of
yamls
before
they
can
get
submitted
to
the
api
server
continuous
auditing
of
existing
workloads?
How
do
we
get
the
json
submitted
by
our
workloads
to
opa
for
validation,
I'm
just
reading
up
on
opa
and
prep
for
cks,
but
this
looks
like
a
necessity
for
enterprise
kubernetes.
I
know.
Last
year
we
talked
a
little
bit
about
more
and
more
people
are
using
opa
and
we
were
talking
about
it
replacing
psps
at
some
point
I
don't
know
the
status
of
this.
Is
that
still
a
thing.
E
So
yeah
for
the
latter
part
of
the
question
I
think
at
the
time,
so
oprah
wasn't
being
replaced,
wasn't
replacing
psps,
but
I
believe
I
just
saw
some
email.
Last
week
I
was
on
vacation,
but
I
think.
E
That
said,
psps
were
getting
deprecated,
so
maybe
that
is
yes,
maybe
oprah
will
most
likely.
Oppa
will
come
in
and
take
in
that
place.
But
at
the
time
when
we
were
having
this
discussion
right
oppa
you
could
use
both
psps
and
oppas
when
necessary.
It
wasn't
either
or
but
now
we'll
see.
E
Psp
is
getting
deprecated
then
yeah
oppa
is
will
be
the
preferred
way
to
go
when
you
need
to
kind
of
put
up
some
guardrails.
As
we
say,
policy
enforcement
things
like
that,
I
heard
there's
another
tool.
That's
coming
out
that
that's
kind
of
out
I
haven't
played
around
with
it.
Yet
it's
also
in
the.
E
I
think
it's
in
the
stand
box
stage
for
cncf
caverno
yeah,
not
to
confuse
anyone
else,
but
but
just
another
thing,
that's
out
there
that
I
have
to
go
and
check
out
regarding
that
policy
enforcement,
so
that
was
it
for
yeah
latter
half.
I
do
see
a
lot
of
companies
using
oppa,
it's
just
it
comes
up
every
time
I
have
a
customer
meeting
regarding
that
space.
It's
like
opa
oppa.
I
was
like
yeah.
E
Throw
down
some
plates,
but
so
yeah.
I
definitely
think
I
I
like
using
it.
I
think
it
works
well,
and
I
personally
liked
it
better
than
psp's
to
accomplish
what
I
needed
to
accomplish,
but
that's
just
a
personal
preference.
E
I
think
that
the
workflow,
I'm
still
I'm
trying
to
figure
out,
do
they
mean
like
oppa's
workflow,
or
they
said
your
workflow.
So
I
was
wondering
which
in
develop
there's
for
me
my
brand
new
and
say
yours.
I
thought
maybe
how
I
develop
policies,
but
I'm
assuming
they're
leaning
more
towards
how
oppa
works
in
a
kubernetes
cluster.
E
Not
yet
I've
seen
some
of
my
colleagues
do
it.
I
haven't
had
the
chance
to
get
my
hands
on
it
and
try
it
out.
Okay,.
B
Yeah,
I
just
want
to
add
another
fun
thing
about.
Oppa
is
that's
quite
flexible
in
that
you
can
use
it
outside
of
kubernetes
like
if
you
were
using
terraform
or
say
cappy
policies
or
any
so
like
the
azure
templates
or
google's
krm.
You
could
use
opa
to
validate
those
for
your
infrastructure
as
well.
So
it's
not
just
kubernetes
specifics,
so
you
could.
If
you
have
a
team,
that's
good
at
open.
You
can
apply
those
resources
across
the
board
to
different.
E
D
Yeah
just
start
some
information,
so
basically
from
one
dot
to
one
yeah
part,
security
policies
are
being
depleted
and
they're
gonna
be
removed
from,
I
think
one
dot.
One.
D
D
A
Nick
minus
there,
once
I
had
conf
test
plus
opa
for
pipeline
validation,
is
great.
We
have
that
in
place.
The
policies
are
very
close
to
what
you
have
in
the
admission
plus
super
quick
to
run
language
and
quirks
language,
and
quirks
takes
a
bit
getting
used
to
thanks
for
that
feedback
nick
and
then
I've
tossed
the
url
for
conf
test
there,
mine
runner,
says
contest
was
also
parts
of
the
recommendations,
thanks
rachel.
G
Yeah,
that's
that's
pretty
nice
video
on
opa
and
also
if
you
want
to
try
out
opa,
you
can
directly
install
opa
via
rancher,
so
you
can
just
install
rancho
onto
your
cluster.
You
can
enable
opa
and
you
can
write
the
policies
they
have
basic
structure
over
there.
So
you
can
write
the
constraint
and
the
policies
and
yes
in
for
cks
it
is
required.
So
you
might
get
a
question
with
opa
where
you
have
to
play
around
a
bit
on
the
rego.
You
don't.
I
The
opac
you
know
page,
they
also
have
like
a
regal
sandbox.
So
you
can
basically,
like
you,
know,
put
together
policies
and
even
if
you
make
some,
you
can
contribute
them.
I
think
that's
very
cool.
It's
actually
something
we're
emulating
for
falco.
So
it's
you
know.
That's
that's
great.
I
think
you
know,
opens
fantastic
fantastic
tool.
B
Gatekeeper
also
has
a
preset
of
a
library
which
I
think
papa's
referring
to
I'm
going
to
drop
the
link
into
there
for
that
as
well.
F
That
is
excellent,
so
I
think
that
opa
and
kavona
bernard
will
work
together.
Cavern
is
more
like
a
mutation
and
sidecar
injection
type
tool,
so
more
than
likely
to
work
together
with
oh,
I
don't
see
it
replacing
it.
F
L
You
played
around
with
guyverno
chauncey.
F
A
Anything
else
on
oppa
we
cover
up
a
lot
last
year
when
we
were
doing
the
blog,
the
office
hours
blog
post,
that
we
will
hopefully
publish
someday
our
number
one
recommendation
was
most
of
you
need
opa,
you
don't
even
know
it
so
definitely
a
good
place
to
invest
some
time
any
anything
else.
Oppa
caverna
related
anyone
gonna
miss
psps.
A
Since
I
know
someone
tossed
a
link
there.
I
see
a
lot
of.
I
see
a
lot
of
no's.
Okay,
that's
good
vivek,
kumar
sahu.
I
hope
I
get
that
right
says
I'm
a
sophomore
to
be
young
learning.
Docker
and
kubernetes.
Can
you
name
some
beginners
projects
using
this
tool?
This
is
a
question.
No,
we
haven't
really
had
asked.
A
You
know
you,
you
type
cube
admin.
You
set
up
your
first
cluster.
What's
is
there
a
test?
Is
there
like
a
vet,
little
vanity
program
that
you
all
use?
I
know
you
know,
there's
a
lot
of
hello
world.
C
C
Yeah,
so
I
I
have
docker
containers
that
run
certain
video
game
servers,
including
quake,
3,
battlefield,
1942
and
other
things
like
that.
I
also
have
a
repo
and
I'll
link.
All
this
don't
worry.
I
have
a
repo
of
kind
of
sample
services
that
I've
collected
over
over
the
years
that
are
amazing
for
getting
kind
of
a
sample
sort
of
application
up
and
running,
and
it
helps
you
understand
how
things
are
going
to
deploy
what
can
happen.
Looking
at
logs
life
cycle
things
like
that,
I
will.
C
I
will
link
all
these,
but
really
it's
kind
of
a
sample
of
like
how
what
it
looks
like
to
manage
real
services
to
provide
real,
endpoints
or
apis
in
in
a
cluster
which
is
super
helpful,
so
and
and
a
lot
of
them
have.
The
yaml
has
a
lot
of
dense
config,
so
you
can
kind
of
go
through
and
see
what
exactly
is
being
deployed
and
configured,
which
is
helpful
for
learning.
F
Yeah
when
I
approach
this
question,
I
would
recommend
some
training.
There's
a
lot
of
training
classes
out
there.
That
will
give
you
like
little
assignments
that
will
help
you
kind
of
get
kick-started,
creating
docker
containers
and
the
manifest
to
deploy
them
to
kubernetes
cluster
or,
just
you
know,
download
some
helm,
charts
and
start
deploying
those
and
inspecting
how
they
were
put
together,
et
cetera.
A
Oh
looks
like
we're
getting
lots
of
recommendations
here,
let's
see
who's
first,
chris,
you
had
one
too
this
google
anthosconfig
management,
repo
page
thing
constraint.
L
A
A
Nick
recommends
pod
info.
This
looks
really
interesting.
It's
just
a
page
that
has
a
bunch
of
pod
info
in
your
browser.
I
think
that's
cool
I've
seen
a
similar
one
to
that,
but
I
can't
remember
its
name.
Mine
runner
recommends
spring
pet
clinic
microservices
nah.
Now
we're
talking.
I
want
to
see
a
pet
store
for
sure,
and
there
are
tons
of
stuff
in
there
that
that
one
looks
that
one
looks
meaty
because
it
has
like
graffana,
dashboards
and
stuff
there's
an
edx
course.
A
You
link
to
rachel
likes
catacota
and
mario's
tossed
in
hall.
His
quart
is
one
I've
seen
used
in
demos
a
lot
qr.
I
don't.
C
C
A
Yeah,
that's
just
useful
because
it
gives
you
so
much
information
back
in
return,
the
google
microservices
demo
onlineboutique.dev
I've
seen
this
that's
a
good
one.
I.
I
Spent,
I
think,
45
minutes
with
the
creator
of
it.
On
my
show,
just
talking
about
the
hipster
demo.
I
think
it's
a
fantastic
thing.
It
literally
is
all
of
their
services
put
together,
but
also,
like
you
know,
kubernetes,
you
can
run
it
locally
as
well.
You
don't
need
gke
for
it
yeah
it
has
like
dot
net
pieces.
I
mean
it
has
every
like
workable
language
in
the
world,
so
props
to
you
on
that
awesome
stuff
is.
I
Yeah
yeah
I've
hacked
it
like.
I've
used
it
at
multiple
conferences
to
show,
like
you
know,
hacking,
hacking,
pods
and
all
that
stuff.
It's
just
it's
the
most
comprehensive
demo
of,
like
you
know,
cloud
native
technologies,
I've
seen
like
it's
it's
fantastic.
A
F
One
of
the
things
that,
when
you
start
playing
with
these
demos
and
if
you
decide
you
want
to
play
with
it
in
a
cloud
provider,
a
lot
of
these
services
will
kick
start
a
load
balancer.
So
you
want
to
be
careful
that
you
shut
it
down.
Otherwise,
you
incur
unexpected
costs.
A
Actually,
tips,
since
you
know
he's
getting
started
any
you
know:
I've
known
people
who've
been
carrying
around
their
like
cloud
provider,
scripts
with
them
or
whatever
are
any
any
tools
in
this
space.
To
avoid
avoid,
you
know
that
kind
of
unexpected
cost
or
like
at
the
end
of
your
workday.
Do
you
like
religiously
check
your
your
dashboards
like?
Is
there
a
you
know?
Do
you
only
use
spot
instances
like
what
kind
of
any
any
tips
there
for
people
getting
started
to
avoid
that
kind
of
stuff.
F
I
Cetera
microcase,
as
well
as
another
one
that
you
know
contraction
but
again
those
three
absolutely
doing
it
locally
and
then,
if
you
want
to
test
there's
free
tiers,
I
think
you
get
almost
300
bucks
from
are
used
to.
I
don't
know
if
that's
still
the
case
with
google
and
others
like
there's,
there's
free
tiers
for
for
just
kicking
a
a
cluster
around
and
putting
some
spot.
You
know
instances
and
the
other
thing
is
gotta.
I
Give
props
to
kubernetes
docs
there's
example
like
deployments
that
you
can
use
just
to
get
the
idea
of
how
to
you
know,
assemble
your
yaml
and
stuff
like
that.
So
big,
big
time,
props
to
the
docs
team
on
that
and
they've.
Also
by
the
way
they
gotta
they
they've
merged
references
into
the
documentation
which
was
a
herculean
effort.
You
got
to
give
them
props
for
that.
So.
F
F
I
A
Up
and
then
yogi
and
borco
bring
up
good
points.
Please
set
up
budget
alerts
in
your
cloud
provider,
cost.
H
H
I'll
throw
one
more
right
here:
the
catacoder.com
gives
you
like
a
playground
to
play
with
kubernetes
and
docker
and,
like
you,
don't
need
to
follow
the
scenarios
like
I
sometimes
just
spin
up
a
kubernetes
cluster
in
their
playground
and
deploy
my
own
things
to
it
and
play
with
it
for
an
hour
like
that's
a
pretty
good
way
of
avoiding
cost
as
well.
I.
E
Was
just
gonna
code
is
my
my
go-to.
If
I
need
something
quick,
because
I
don't
have
to
care
about
resources
on
my
laptop
or
in
the
cloud.
G
E
I
just
close
my
browser
and
I'm
good
to
go
and
there's
also
you
could
use
it
for
terraform
and
some
other
you
know
other
things
to
play
around
with.
So
it's
not
just
kubernetes.
Oh.
L
A
L
A
Wow
the
amount
of
urls
today
great
job
everybody-
we
got
a
lot
of
lots
of
great
stuff
and
yoga
has
a
pers
as
a
personal
demo,
app
as
well,
which
is
using
a
lot
of
spring.
So,
if
that's
your
wheelhouse,
you
definitely
want
to
check
that
out.
I'm
checking
that
out
now.
That
is
definitely
a
bookmark
all
right
and
with
that
we
are
running
out
of
time,
but
I
told
andre
we
would
check
out
his
question.
He
wants
us
to
check
out
this
pull
request
here.
A
Yeah
are
any
of
you
in
cygnode
you're,
not
on
signode,
right
andre,
I'm
going
to
have
to
follow
up
with
you
on
this
one,
because.
A
Yeah
I
see
don
here.
Let
me
let's
talk
a
little
bit
about
how
to
how
to
get
reviews
stuff,
and
I
know
we're
running
out
of
time.
I
just
want
to
spend
a
real
minute
on.
This
is
sometimes
it's
very
difficult
to
get
changes
into
kubernetes,
because
at
any
given
time
there
are
how
many
about
a
thousand
open
pull
requests.
934.
Today
it's
been
about
a
thousand,
since
I've
been
working
on
kubernetes
for
three
years
now
the
velocity
of
merges
is
always
going
up.
A
So
it's
not
like
the
same
thousand
sitting
there.
However,
there
are
certain
cases
where
some
cigs
get
backed
up
and
some
you
know
some
don't
so
you
might
feel
that
oh,
my
change
is
stuck.
How
do
I
get
eyeballs
on
it,
and
so
I'm
going
to
give
you
some
recommendations
on
on
what
what
I
would
do
here.
The
first
is
the
sig
node
slack
channel.
Each
sig
has
a
slack
channel
in
the
slack.
You
could
say:
hey
here's
a
pull
request.
A
The
second
I
would
do
is
each
sig
has
a
weekly
meeting
that
anybody
in
the
public
can
attend.
If
you
go
to
sigs.yaml.
A
A
He
goes,
I
would
like
to
discuss
the
technical
details
of.
Do
you
think
it
makes
sense,
he's
insignificant,
yeah,
he's
feeling
alone
yeah.
So
the
next
next
recommendation
here
is
try
to
attend
a
meeting.
If
you
can
and
if
not,
we
have
sick
contributor
experience
that
might
be
able
to
help
so
there's
actually
a
channel
called
pr
reviews
in
slack
for
small
one
like
this,
that
you
can
do,
and
you
can
ask
for
pr
reviews.
A
I
will
actually,
after
this
I'm
going
to
see
if
dims
is
around,
and
I
would
like
to
see
if
I
can
maybe
sneak
that
in
sometimes
it's
just
one
of
those
things
where
so
many
pr's.
You
know
nobody
hates
you
on
purpose,
one
of
those
things
it's
just.
We
have
limited
amount
of
volunteers,
and
sometimes
you
just
got
to
be
a
squeaky
annoying
wheel.
Another
good
one
is
the
mailing
list,
because
that's
kind
of
asynchronous
post
as
well
sometimes
posting
on
there
might
help.
A
But
after
this
I
will
definitely
see
if
I
can
get
you
a
review
on
that.
Any
of
you
try
to
do
a
pr
that
gets
stuck
on
something.
A
A
We
had
we
had
a
script
before,
but
then
I
found
it's.
C
C
A
And
andre
I'd
like
to
give
you
a
t-shirt
today,
I
will
dm
both
of
you
tomorrow
or
tomorrow.
After
the
show
and
yogi
says
me
me
me
me
yogi
you've,
given
us
a
lot
of
great
urls,
so
pm
me,
and
I
will
give
you
another
kubernetes
t-shirt
so
with
that
panel.
Thank
you.
So
much
has
been
great
anybody
else.
Anything
to
add
anything.
You
all
have
a
good
time.
A
Yeah
excellent
yeah,
so
the
way
it
works
just
volunteer,
the
monday
of
I
say:
hey
who
wants
to
come
on
and
whoever
shows
up.
If
we
get
more
than
three
people
we
do
a
show
and
then
we
like
we
do
like
to
keep
this
this
group
going
with
new
people
and
get
a
bunch
of
diverse
viewpoints
and
things
like
that.
So,
if
you're
doing
something
out
there,
I
think
pavlos
actually
started.
A
You
know
just
popping
urls
in
the
channel
and
things
like
that
and
eventually
I
just
dared
him
to
come
on,
and
here
he
is,
you
could
pm
me,
I'm
castro
j.o
and
if
not
I'll,
follow
up
with
you
yogi
with
that
we
will
finish
panel
stick
around
everybody
else.
Thank
you
very
much,
and
we
will
see
you
it's
always
the
third
wednesday
of
every
month,
and
we
will
see
you
all
in
february.
Welcome
back,
have
a
happy
2021.