►
Description
Recently, the Kubernetes community announced the deprecation of Docker support. Specifically, the deprecation of dockershim. What does that mean for you? How can you prepare to migrate? Join us for a discussion and some demonstration.
How Red Hat Enterprise Linux (RHEL) users and admins can benefit their organizations and improve their careers by learning how to use containers, Kubernetes, and Red Hat OpenShift.
Learn more at https://red.ht/leveluphour
A
Good
morning
good
afternoon
good
evening
and
welcome
to
the
first
edition
of
level
up
hour
here
in
2021,
I
am
chris
short
host
and
executive
producer
of
openshift
tv
wishing
you
a
very
happy
new
year,
I
am
joined
by
two
wonderful
red
hatters
today,
the
illustrious
langdon
white
and
the
always
brilliant
scott
mccarty
langdon.
You
want
to
talk
about
what
we're
talking
about
today.
B
Yeah
so,
as
we
said,
I'm
lagging
white
we're
doing
the
level
up
hour,
which
is
a
show
about
kind
of
you
know,
giving
you
some
experience
with
containerization
from
a
kind
of
non-developer
perspective.
You
know
we
also
cover
some
development
stuff
as
well,
but
you
know
kind
of
if
you
haven't
really
played
around
with
containers.
B
So
so
that's
why
we're
here
today
and
we
brought
scott
and
I
let
him
introduce
himself,
but
you
know
mostly
because
he
has
the
deep
dive,
scary
internals
of
containers,
expertise,
particularly
on
rel-
that
we
think
will
be
useful
to
this
discussion.
It's
useful
to
a
lot
of
discussions,
but
normally
it's
hard
to
you
know
deal
with
the
energy
that
scott
has
all
the
time.
So.
B
C
Hey
how's
it
going
langdon,
I
guess
I'll,
introduce
myself.
My
name
is
scott
mccarty.
I
am
a
product
manager
for
container
tooling
and
images,
and
so
I
drive
the
road
map
for
like
red
hat
universal
base,
image
and
pod
man
and
scopio,
and
build
a
and
cryo
inside
of
openshift,
and
then
I'll
say
this
like.
So
we
went
through
this.
This
whole
kubernetes
thing
where
everybody
kind
of
got
nervous.
We
went
through
this
like
two
years
ago,
so
like
red
hat
in
particular,
moved
away
from
docker.
C
You
know
to
cryo
and
podman,
with
the
launch
of
rel
eight
and
with
the
launch
of
openshift
four,
and
so
I
published
probably
five
articles
about
that
around
that
time
and
elated
a
lot
of
people's
concerns
at
that
point
and
so
yeah.
I
have
a
lot
of
it's
funny.
We
already
have
the
experience
that
the
rest
of
the
world
is
now
starting
to
go
through
and
I'm
like.
Oh.
A
B
A
B
That's
kind
of
like
that's
kind
of
one
thing
right
like
yeah,
I
mean
this
is
the
idea
behind
oci
right
is
like
there.
There
was
an
organization
that
developed
a
spec
so
that
we
wouldn't
you
know,
have
to
worry
about
this
kind
of
problem
right
is
that
we
can
have
kind
of
competitive
container
run
times.
That
will
all
run
the
same
kinds
of
things
and
you
know,
and
it
in
in
a
lot
of
ways
it's
it
seems
like
it
has
been
very
successful
and
so
that's
kind
of
a
little
bit.
B
B
Yeah
exactly
now,
I
have
to
find
the
slides,
which
is,
oh,
you
know,
always
a
always
a
challenge
and,
of
course,
because
it's
been
a
few
since
we
last
had
the
show
nothing
is
going
to
be
logged
in
and
I
changed.
I
changed
providers
and
all
that
jazz.
Let's
see
how
about
present
oh
right,
but
we
have
the
fun
you
can't
present
in
the
same
way
in
a
different
window
anymore.
It's
really
annoying
all
right.
Let's.
B
B
B
So,
let's
see
just
by
way
of
background,
as
we
mentioned,
this
is
kubernetes
and
docker
duplication,
I'm
thinking
about
changing
the
episode
titling
because
I'm
still
struggling
with,
like
what
does
season
mean
right
in
in
a
live
streaming
thing.
So,
instead
I
went
with
the
super
clear
and
understandable
year
month,
day,
format,
because
then
at
least
it
will
order
automatically
by
you
know,
with.
B
Yes,
so
so
I'm
toying
with
that-
and
actually
these
are
just
the
show
notes
from
last
time.
I
apparently
I
forgot
to
update
that
piece
of
text,
but
the
link
is
correct
to
the
last
episode's
show
notes
so
check
those
out.
If
you
like,
and
you
know
we
we
try
to
recap
the
whole
show
and
talk
about
you
know
kind
of
what
I
try
to
do.
Is
I
try
to
capture
everything
that
happened
as
well
as
like
capture
like
hey,
you
want
to
learn
more
about
this
particular
subject.
B
Go
check
out
the
further
reading
on
this
I
try
to
pull
out
some
fun.
Video
highlights.
I
am
more
successful
than
others,
depending
on
how
funny
we
were
that
particular
episode.
So
remember
more
funny
is
better.
That
way.
I
can
find
video
clips,
but
that's
it
for
the
slides
for
the
moment,
because
otherwise
we
would
be
giving
away
the
sweet,
sweet
internet
points
and
we
want
to
hold
that
until
later.
A
B
B
B
C
Yeah,
I
was
thinking
I
was
thinking.
I
could
probably
walk
through
the
five
main
points
I
made
about
like
what
the
five
pieces,
what
does
kubernetes
support
even
mean,
and
I
kind
of
break
it
down
into
five
things
that
we
could
probably
like,
walk
through
each
one,
even
though
I
wrote
them
in
text,
you
know
how
it
is
verbally
like
once
you
start
talking
about
something
you
highlight
things.
You
know
anecdotes
and
things
that
you
hadn't
thought
about
and
things
you
didn't
capture
sounds
great.
C
C
About
the
magic
all
the
magic
yeah,
it
is
technology
used
sufficiently
such
that
it
appears
to
be
magic
right
right.
So
so,
like
I
wrote,
obviously
an
article
on
this,
and
then
I
was
a
little
click.
Baity
I'll
fully
admit
with
like
kubernetes
is
removing
docker
support.
Kubernetes
is
not
removing
docker
support
like.
C
But
you
know
like
I
and
I
and
it's
funny,
because
I
talked
about
all
this
like
two
years
ago,
when
we,
when
we
moved
openshift
to
cryo,
and
I
was
prepared
that
there
would
be
a
ton
of
questions
I
mean
if
you
go
back,
we
had
started
working
on
this,
even
in
openshift
3-7.
So
really
it's
more
like
three
years
ago.
I
don't
know
three
and
a
half
four
issues
I
it
was.
C
I
can't
remember
the
dates
like
that's
how
long
ago
it
was
probably
2016-17,
and
actually
I
just
even
filled
in
a
history
of
all
this,
so
we
were
already
thinking
about
like
once.
We
saw
the
the
once
once.
Obviously,
oci
is
one
piece
of
this
and
then
cri
in
general,
like
having
the
container
runtime
interface
that
kubernetes
end
up
building
right
like
so
so
you
you
know,
I
started
thinking
about.
How
could
we
talk
about
this
and
like?
How
could
we?
C
How
can
we
explain
what
the
difference
actually
means
to
people
in
something
that
they
care
about
right,
and
so
you
know
this.
This
version
that
I
published
recently
is
basically
the
culmination
of
like
three
and
a
half
years
of
thinking
about
about
this.
So
I
think
it's
a
lot
more
crisp
now
I
can
explain
it,
but
you
know
there's
basically
five
different
things
when
you
say
the
word
docker,
you
know
like
what
is
docker
support.
There's
like
five
different
buckets
of
things.
C
You
think
through
the
first
one
being
like
container
images
right,
like
I
mean
dan
walsh,
talks
about
this
all
the
time,
and
I
still
fundamentally
believe.
Like
paths
already
existed,
you
know,
so
we
already
had
highly
automated
ways
to
deploy
code.
We
had
that
like
and
for
some
reason
that
a
lot
of
people
can't
get
their
brain
around.
We
ended
up
going
backwards
right.
C
C
Five
stars
would
register
server,
five
stars
with
runtime
format,
you're
getting
like
four
stars
with
command
line,
interface
and
you're,
maybe
getting
like
zero
stars
for
like
api
like
there
is.
That
is
truly
probably
the
biggest
change
you
know
is
around.
That
is
around
the
api
like
what
apis
are
available
and
whether
you
should
have
even
been
using
those
apis
anyway
in
the
first
place,
but
that's
kind
of
beside
that
we
can
get
into
that
deeper.
B
Yeah
I
was
going
to
kind
of
comment
on
on
some
of
that
right.
So
one
of
the
examples
of
of
kind
of
like
should
you
have
been
using
those
apis
right
is
like
jerome
who
I
could
never
pronounce
his
last
name
but
starts
the
p.
You
know
has
talked
several
different
occasions
about
things
that
you
can
do
that
are
really
really
cool
like
can
you
do
doing
docker
inside
docker,
for
example,
you
know
and
dan
walsh.
B
I
actually
remember
specifically
doing
something
along
the
same
lines,
particularly
with
contain
with
sorry
with
systemd
inside
docker
and
then
docker
running
in
that,
and
then
virtual
machines
like
these
are
all
very
cool
things.
Cool
and
production,
usually
not
something
you
should
put
together.
You
know
so
like
there
are
specific
use
cases
where
some
of
these
things
are
like
important
and
necessary
or
whatever,
but
the
vast
majority
of
cases
you
probably
are
trying
to
build
something.
A
little
crazy.
C
Yeah,
I
agree
and
what
happens
is
people
I
don't
know.
If
we
want
we
could
we
don't
have
to
tackle
them
in
order,
but
we
could
probably
jump
into
apis
because
it's
the
hairiest
one,
but
but
people
usually
end
up
in
that
scenario,
not
on
purpose.
They
like
they
have
some.
They
were
like
running
a
kubernetes
cluster
like
kubernetes,
1.11
or
something
right,
and
they
run
it
with,
like
you
know,
with
docker,
because
that
was
all
that
there
was
you
know
or
whatever
at
the
time
and
then
one
day
they
go.
C
Oh,
we
need
to
scan
our
images
and
they
go.
Your
security
team
came
and
beat
us
up
said
we
gotta
scan
these
images,
but
we
gotta
do
it
in
production.
You're
like
how
do
we
do
that
you,
google,
search
and
then
you're
like
oh
there's,
this
tool
that
will
talk
to
the
docker
demon
and
we'll
like
scan
the
images
right
like
and
you're.
You
find
it.
Then
you
go.
Oh,
you
know
what
else
we
gotta
do
we
need
to
do.
C
We
need
to
like
snapshot
the
image
because,
like
we
did
this
update
and
then
kubernetes
broke
and
blah
blah
blah,
like
you
end
up
down
some
rat
hole
right
like
where
you're
like.
Oh,
we
try
to
do
an
update.
It
failed
so
now
we're
using
the
snapshotting
tool
that
talks
to
the
docker
api
to
snapshot
it
before
we
do
an
update,
you're
like
wait,
a
minute.
That's
getting
pretty
wacky,
like
you
probably
shouldn't
do
that,
but
you
should
probably
let
kubernetes
handle
that.
C
You
know
temporal,
you
know
it's
always
a
race
condition
right,
like
you,
have
a
business
problem
or
a
technical
problem
and
you're
like
I
gotta
solve
this
thing
and
then
you
end
up
fixing
it
before
it's
fixed
kind
of
the
right
way,
because
you
have
the
problem
today.
Right
and
so
then
what
happens?
Is
I've
seen
it?
You
know
some
of
our
customers,
even
with
openshift
three.
C
They
end
up
in
the
scenario
where
they
have
two
three
four
five
tools
inside
the
kubernetes
cluster
like
running
on
the
nodes
talking
to
the
docker
demon,
and
that's
probably
the
hairiest
place
like
like
that's
the
hairiest
place,
you're
like
oh
man,
ripping
out
docker,
does
kind
of
suck.
In
that
scenario,.
B
Well-
and
I
was
gonna
say
like
one
of
the
things
that
you
kind
of
need
to
remember
about
docker.
Well,
so,
first
of
all,
you
know
I
have
a
little
bit
of
a
demo
that
we
can
walk
through
in
a
little
bit
but
like
when
you
say
docker.
Usually
people
are
referring
to
the
thing
that
launched.
I
don't
know
what
seven
years
ago
or
something
now
that
thing
and
what
docker
the
company,
which
wasn't
even
called
docker
the
company
at
the
time
right,
are
actually
quite
different
right
and
you
know
so.
B
The
thing
that
is
is
necessary
to
remember
is
that
when
docker
was
originally
developed,
it
wasn't
really
meant
to
be
this
thing.
That
would
run
in
an
automated
kind
of
environment,
and
you
know
specialized
world.
You
know
in
a
sense,
it
wasn't
even
really
trying
to
replace
like
paths
or
something
like
that
per
se,
not
in
the
automation,
side
of
it,
etc.
It
was
really
a
ux
tool,
you
know,
or
a
user
experience
tool
for
developers
or
administrators
who
worked
for
this
particular
company.
You
know
which
now
I
can't
remember
the
name
of
cloud.
B
That's
it
that
you
know
they
were
using
this
in
their
production
environment
to
basically
host
other
people's
software,
and
then
it
just
kind
of
took
off
by
storm
because
of
how
awesome
that
user
experience
was.
But
what
it
didn't
do.
A
great
job
of
was
disconnecting
the
kind
of
client-side
tooling
of
it.
You
know,
so
what
it
like,
a
developer
uses
or
even
an
administrator
uses
and
the
production
runtime
and
what
were,
and
so
what
kubernetes
has
been
kind
of
fighting
with,
is
accessing
that
production
aspect
right.
B
Their
production,
runtime
part
of
it
by
way
of
the
client,
ux
runtime
or
whatever,
and
that's
what
docker
shim
does
right
is
basically
that
hook,
and
so
it's
not
that
there's
anything
inherently
wrong
with
the
components
underneath,
but
rather
that
it's
like
how
they're
linked
together
and
how
you
have
to
access
them
and
in
a
sense
that's
actually
why
docker
has
gone
on
to
create
container
d,
for
example,
and
pull
that
out.
So
you
don't
have
to
access
the
apis
in
a
non-public
way.
C
And
I
view
it
as
a
it's.
An
api
battle
right
like
it
was
an
api
battle
like
heroku,
probably
came
out
with
pads
for
heroic,
is
the
first
company
I
knew
about
that.
Did
pass
right,
like
all
my
friends
that
were
in
startup
trees
in
heroku
back
in,
like
I
don't
know,
2008
or
something
I
don't
remember
a
long
time
ago,
like
2011
2008,
that
kind
of
time
frame,
and
you
go
okay,
so
that
api
started
to
kind
of
win
right
like
there
was
like.
Oh,
I
can.
C
I
could
just
type
these
few
commands
or
maybe
wired
into
my
ruby
code
and
have
like
this
deployment.
They'd
have
these
deployment
scripts
and
ruby
that
would
just
go
out
to
heroku
and
deploy
the
script.
You
know
deploy
like
an
entire
application
right
and
so
you're
like
oh,
instead
of
and
at
the
same
time
you
had
config
management,
also
competing
right.
So
that's
another
api
for,
like
configuring,
things
and
deploying
things
and
like
there's
just
this
like
wild
west
of
deployment.
C
Apis
is
what
I'd
say,
and
then
you
know
then
again
something
like
kuroku
and
then
even
openshift,
like
2,
was
really
highly
automated,
like
it
was
really
good.
It
took
like
one
command
to
deploy
something
you
one
could
argue.
It
was
actually
every
bit
as
easier
as
docker
or
easier,
but
what
it
was
it
was
visual
basic
like
it
was
great
as
long
as
what
you
were
doing
was.
C
A
C
That's
what,
in
my
view,
that's
what
paz
was.
It
was
amazing
if
all
you
want
to
do
is
deploy
ruby,
but
as
soon
as
somebody
said-
oh,
we
have
this.
We
have
this
actor
pattern
model
thing
where
the
ruby
things
have
to
go
fire
up,
other
ruby
things
they're
like
well,
you
can't
do
that
with
pads,
like
that's
crazy
or
like
any
time,
you'd
end
up
in
one
of
these
edge
cases,
getting
out
of
pass
became
really
panty
ass.
Then
docker
came
out,
like
you
said,
as
like
a
management
tool
right
like
right.
C
B
C
B
A
C
A
C
Building
heroku,
you
know
plugs
or
whatever
slugs
or
whatever
they
were
called
at
the
time
or
building
open
shift
two
slugs.
They
were
freaking
hard
like
that
was
a
I
got
I
got
rebuffed
by
that
like
10
times
like
I
sit
down
and
go
to
dude,
I'm
like
I
don't
know,
I
don't
feel
like
doing
this
like
it
was
just
hard
right.
B
Right,
I
did
want
to
make
a
little
shout
out
there
with
the
visual
basic
comment,
though,
because
back
when
I
was
doing
visual
basic,
we
would
do.
We
would
actually
literally
have
strings
of
assembly
in
our
code
that
we
would
execute
in
ram
to
rewrite
v
tables
so
that
we
could
get
what
we
wanted
to
work.
Oh,
so
you
could
you
could
do
anything
and,
oh
you
could
do
it.
You
try
hard
enough.
C
Yeah,
it
was,
you
might
say,
like
the
other
problem
is,
then
the
api
battle
went
to
okay,
docker's,
really
good
at
running
a
few
services
on
a
single
node,
and
then
we
all
saw
that
wait,
a
minute
at
scale
you
need
like,
like.
I
don't
know
what
dot
cod
cloud
had
internally,
but
I
assume
they
were
already
they're
like.
Oh,
this
makes
a
single
node
really
easy,
but
they
probably
had
some
extra
software
layer
that
went
out
and
talked
to
the
docker
api
and
scheduling
like
if
you're
a
hosting
provider.
C
It's
not
the
same
as
a
cloud
provider.
Let's
be
honest
like
like
the
spin
up.
Shutdown
is
like
five
percent
of
your
workload
and
the
running
is
like
95
percent
a
cloud
provider
like
35
40.
Is
your
spin
up
shutdown
right,
like
your
your
actual
runtime
capacity,
is
probably
lower
and
that's
the
problem
with
the
configuring,
the
networking
configuring,
the
workload
see
you
know,
allocating
cpus,
deallocating
blah
blah.
C
It's
the
allocation
deallocation
that
ends
up
eating
so
bad
and
so
like
kubernetes,
was
way
better
at
that
right,
like
and
so
as
soon
as
you
went
to
do
that
at
scale.
You're
like
oh,
but
of
course,
docker
didn't
want
to
give
up
and
and
and
docker
was
the
thing
that
kubernetes
got
built
on
so
that
talking
to
docker
was
hard-coded,
and
then
we
went
through
that
period.
We're
like
oh,
let's
break
this
out.
Let's
call
it
the
docker
shim.
That
code
was
already
part
of
kubernetes.
They
basically
just
severed.
C
B
B
Yeah
right
there's
that
yeah
there's
too
many
c's
and
ends
in
all
of
them,
and
I
get
I
get
them
except
all
the
time.
I
need
like
a
cheat
sheet.
You
know,
along
with
my
vi,
cheat
sheet
so
yeah,
but
so
so
I
mean
you
know
it's
like
every
other.
You
know
kind
of
nascent
technology
that
we
see
right.
It
goes
through
a
lot
of
churn
on
you
know,
kind
of
how
it
works
and
how
it
connects
together,
because
you
know,
particularly
in
the
open
source
world.
B
C
But
some
people
get
pincered
and
that's
right.
The
people,
the
people
that
had
stuff
that
was
talking
to
both
the
docker
api
and
the
kubernetes
api
in
a
big
cluster
do
get
pincered,
but
that
is
a
minority
right
like
that
should
be
a
very
small
minority.
It
it
hit
vendors.
You
know
like,
like
ride-along
vendors,
like
you
know,
there
were
vendors
that
would
go
and
talk
to
the
docker
api
to
scan
images.
There
were
vendors
that
would
talk
to
the
docker
api
to
do.
C
Backups
of
data
like
there
was
all
kinds
of
experimentation
and
the
problem
was,
is
experimentation
in
a
marketplace
once
I've
bought
it.
I'm
annoyed
right,
like
like
I'm
going
to
be
very
annoyed.
If
I
have
money
and
time
you
know,
labor
and
capital
invested
in
this,
and
now
and
now
the
the
carpet
gets
moved
and
that's
that's
probably
the
biggest
downside
of
the
docker
deprecation.
C
You
know
ahead
of
time
so
that
most
like
because
openshift
hat
is
probably
the
biggest
you
know,
kubernetes
distribution
that
is
independent,
that
people
install
themselves
right
like
it
already
believes
the
trail
here
in
that,
like
all
the
vendors
that
work
with
kubernetes
are
like,
I
want
to
work
with
openshift
and
when
we
moved
to
cryo
they
went.
Oh,
we
got
to
use
the
cri
interface.
C
We
can't
just
go
directly
to
the
docker
demon,
and
so
I
at
a
minimum
you
know,
there's
hundreds
of
vendors
that
we're
either
looking
at
they're
at
some
stage
of
the
pipeline
of
either
looking
at
moving
the
cri
in
process
of
moving
to
cri
or
have
moved
to
cri,
and
so
like
data
dog
and
anchor
and,
like
you
know,
black
duck,
and
like
I
mean
I,
I
talked
to
probably
like
a
hundred
vendors.
You
know
in
the
last
two
years
of
like
how
are
they
gonna
move
from
the
docker
api
to
cri?
C
B
Although
I
will
I
mean
you
know,
like
I
said,
I
think
largely
we're
trying
to
kind
of
say
you
know
we,
we
really
hope
for
the
vast
majority
of
use
cases.
This
is
not
that
big
a
deal.
I
was
a
little
surprised
to
discover
that
the
removal
is
actually
sooner
than
I
thought
it's
actually
the
end
of
now
this
year.
B
Yeah
I
thought
it
was
going
to
be
more
like
a
two-year
thing.
You
know,
but
you
know
going
back
to
that
kind
of
you
know.
Innovation
thing,
two
aspects
of
that
right,
one
is
that
you
know
it
does
churn
a
lot,
so
things
do
move
fast,
but
on
the
flip
side
of
that
one
of
the
things
we
have
learned
over
over
the
years
with
highly
you
know
innovative
things
whatever
is
to
actually
have
you
know
for
lack
of
a
better
term.
B
Lts
ish,
you
know
versions
right
so
so
there
are
versions
of
kubernetes
that
are
sticking
around
longer
right.
You
know,
I
think
you
know
we're.
We
have
an
open
shift
out.
For
example,
that's
got
a
longer
term
of
service
than
traditional
right,
so
so
you
can.
You
can
probably
delay
that
adoption
without
losing
too
much.
You
know
one
of
the
things
that
was
the
big
driver
for
something
like
rel
for
adoption
right
is
that
the
hardware
was
changing.
So
you
you
couldn't
delay
it.
B
B
Kubernetes
or
whatever,
like
the
delay,
is,
is
less
or
like
it's
less
forced
to
kind
of
move
on
you
know.
Obviously
you
want
to
maintain
support
your
you
know.
Whoever
your
vendor
is
providing
support,
but
aside
from
that,
there's
not
a
big
driver
to
move
aside
from
being
able
to
take
advantage
of
features
and
features
are
often
a
good
driver
to
remove,
but
at
the
same
time
it's
not
because
it
won't
work
otherwise
right.
So
I
think
that's
yeah.
C
B
B
A
B
A
B
C
And
this
is
probably
a
good
time
to
then
dig
into
some
of
the
other
ones
like
go
like
like,
for
example
like
this
before
you
get
worried
about
the
date
think
about
whether
you
actually
need
it
or
not,
whether
you'll,
actually
notice
or
not
like.
If
all
you're
worried
about
is
container
images
like
we
have
a
bunch
of
docker
images,
they're
saved
into
registry
we'll
be
able
to
use
those
does
that
have
to
change
no
like
that,
doesn't
have
to
change
at
all.
That's
easy
right,
like
like
cryo
container
d,
any
other
runtime.
C
C
It's
the
exact
same
container
in
linux,
like
I
used
to
do
this
demo,
where
I
would
have
podman
and
actually
I
think
I
had
cryo
and
podman
and
docker
side
by
side
and
I'd
fire
up
three
different
containers.
I'd
use,
cri
ctl
to
talk
directly
to
to
cryo,
and
I
would
fire
up
a
container
locally
and
I
would
be
like
hey
like
now
tell
me,
which
one
was
started
by
what
demon
you
know
like
like
look
at
them
and
and
they'd
be
like.
Oh
look
at
the
sc
linux,
I'm
like
the
s.
C
Linux
is
the
same.
Like
the
esvert
code
that
we
wrote,
you
know
that's
basically
asvert.
It's
called
s,
pert,
a
technology
that
basically
dynamically
generates
a
cube.
You
know
or
a
seo
linux
label.
That's
the
same
in
kvm,
it's
the
same
in
it's
the
same
in
podman,
it's
the
same
in
docker.
It's
the
same
in
everything
because
run
c
uses
it
so
like
run
c
calls
the
same
thing.
B
B
Could
have
we
could
have
named
our
database
sql
right
and
our
web
server
internet
information
server.
You
know
if
we
wanted
to.
You
know,
play
with
the
big
boys
so
yeah,
actually
that
so
I
do
have
a
little
demo
of
that.
I
don't
do
we
want
to
talk
about
that,
because
I
thought
it
was
interesting,
particularly
seeing
as
I
don't
do.
Multi-Machine
vagrant
files
very
often-
and
I
always
like
to
show
off
vagrant
as
well
so
I'll
just
share.
B
All
right,
you
should
see
vs
code
is
that
what
y'all
are
seeing?
Okay,
cool,
okay,
so
what
I
did
was.
Obviously
I
have
a
lot
of
experience
with
podman,
so
that
is
first
in
my
list
here.
But
so
we
have,
you
know.
Basically
I
just
you
know
I
set
up
live
verts
that
it
has
four
cpus
and
some
ram
and
whatever,
but
basically
I
just
installed
podman
scopio
and
builda.
B
Yeah
and
like
I,
I
love
the
origin
story
of
like
basically
making
it,
because
I
had
this
problem
all
the
time,
because
I
was
a
web
developer
for
a
long
time.
Is
that,
like
you
know,
how
do
I
make
it
on
a
mac
so
that
I
can
have
multiple
running
web
servers
without
having
to
know
the
ins
and
outs
of
virtualization,
and
you
know
and
like
routing
and
all
that
other
stuff
and
it
just
works,
and
it's
it's
really
really
good.
I.
B
B
A
B
B
B
You
know
so
I
don't
mean
freely
like
as
in
not
open
source
but
freely
available
engine
that
will
run
a
docker
container
and
if
you
notice,
when
I
install
it,
I
actually
install
container
excuse
me,
I
install
container
d
and
docker,
and
so,
as
far
as
I
can
tell,
what
it's
doing
is
that
ux
that
we
know
and
love
is
actually
talking
to
container
d
which,
from
earlier
in
this
discussion,
you
know
that
what
kubernetes
is
talking
about
doing
is
exactly
this.
So
you
know
so.
C
B
Yeah,
okay,
so
long
story
short
there's.
So
when
we
say
docker
we
could
mean
this
one,
but
there
is
also
their
enterprise
version,
which
is
what
they
you
know,
sell
support
on
which
I
don't
I
didn't
have
a
copy
of
handy.
So
I
assume
it's
the
same
as
this
or
very,
very
similar.
So
you
know
your
mileage
may
vary.
It
might
be
worth
a
follow-up
just
to
test
it
out,
but
I
imagine
anything
that
works
in
the
kind
of
freely
available
version
will
work
in
the
fully
supported
version.
A
A
A
B
B
So
so
you
know,
but
so
basically
what
I
wanted
to
do
here
is
you
know
this
is
not
an
official
installation,
but
it's
you
know
it's
pretty
close.
It
was
on
fedora
magazine,
which
is,
you
know,
published
widely.
If
it
was
horribly
wrong,
it
would
have
had
you
know
a
wall
of
comments
and
would
have
gotten
cr.
You
know
corrected,
but
this
is
how
you
get
moby
on
fedora
and
this
particular
article
was
written
for
32,
so
I
adapted
it
a
little
bit
for
33.
B
C
B
C
B
So
that
this
basically
this
line
here
should
become
unnecessary,
soonish
right,
but
you
know
it
kind
of
installs.
The
way
you
normally
do.
I
was
a
little
confused
about
why
it
installed
docker
compose
by
default,
but
probably
because
doctor
compose
is
super
useful,
so
you
know,
and
we've
actually
we've
covered
that
in
another
episode
as
well.
You
know
you
know
trying
like
trying
to
replicate
some
of
the
docker
compose
experience
with
podman
and.
B
B
Because
I've
been
doing
it
with
yaml
and
I'm
all
set
to
switch
to
sigma
so
yeah,
like
you,
know
doing
it
in
yaml,
well
what
I
kind
of
make
the
point
of
it's:
it's
forward
compatible
if
you
want
to
move
to
something
like
kubernetes,
which
is
super
nice
rather
than
docker
compose,
which
is
kind
of
like
a
dead
end.
But
docker
compose
has
a
great
user
experience
much
like
docker
itself.
That's
really
hard
to
find.
C
There's
a
bunch
of
code
written
in
it.
So
like
one
of
the
one
of
the
migration
paths,
we're
hoping
to
support
is
you
can
use
the
docker
compose,
get
some
of
your
compose
files,
basically
running
in
podman
and
then
do
the
coupe
generate.
You
know,
generate
kube
and
then
generate
the
kubernetes
ammo,
and
so
then,
like
we
see
a
path,
there
will
be
able
to
basically
import
your
compose
files.
C
C
B
You
know
basically
the
install
script,
so
I
think
I
shared
the
whole
screen,
so
you
should
be
able
to
see
this,
but
what
I
wanted
to
kind
of
make
the
point
of
is
like,
and
you
know
I
will.
I
will
cheat
with
the
you
know-
hey
I
I
did
this
in
advance
so
that
it
would
actually
work,
but
I
promise
I
didn't
make
any
changes.
Besides
what
is
checked
into
the
repo
it
is.
B
Actually,
if
you
want
to
look
at
the
repo
it's
there
under
the
branch
called
2021
0106.,
it's
just
I
don't.
I
don't
like
to
actually
push
it
to
master
until
I
have
the
show
notes
and
everything
in
there,
but
it
is
pushed
so
if
you
want
to
see
this
vagrant
file
or
the
docker
file,
I'm
about
to
show
you
it
is
there
the
the
only
thing
that
maybe
I'll
go
one
of
these
days
and
write
a
patch
for.
A
C
B
Yes,
yes,
so
I
would
really
like
vagrant
ssh
to
instead
of
just
yelling
at
you
for
not
choosing
which
server
you
want.
Why
can't
it
just
prompt
you
for
which
server
you
want?
I
don't
know
I
have
to
type
it
in
again,
because
I'm
lazy,
so
theoretically
that's
running
and
that's
cool.
So
what
what
we
can
do
here
is
you
know
I
kind
of
like.
B
You
know
this
is
worthwhile
to
explore
more
complex,
docker
files,
and
you
know
I
just
wasn't
sure
the
value
versus
time
kind
of
trade-off
for
it.
So
you
know
feel
free
to
play
around
with
this
yourself.
That's
part
of
why
I
published
the
vagrant
file.
You
know
feel
free
to
go,
grab
it
and
you
know
use
it
any
way
you
like,
but
so
one
thing
I'll
point
out.
You
know
we
have
to
do
it
with
sudo.
B
You
know
one
of
the
tricks
on
at
least
most
of
the
linux
is,
I
don't
know
about
doing
it
much
on
windows
or
mac.
You
know,
because
I
mostly
use
linux.
One
of
the
tricks
is
actually
to
add
your
user
to
the
docker
group,
which
will
let
you
do
it
without
sudo,
but
be
aware
that
you
are
running
it
as
root.
You
just
happen
to
have.
Basically
it's
a
no
password
pseudo
kind
of
scenario,
so
you
can
just
run
the
command
rather
than
having
to
actually
choose
it.
B
B
Docker
run
I
t,
and
this
actually
yeah
I
t
I
don't
want
to
remove
it,
so
I
want
to
make
sure
that
this
is
going
to
stick
around
to
finish
the
demo
and
that
should
just
work,
and
it
does
look
at
that.
So
now
I
can
ping,
and
one
of
the
cautions
that
I
read
in
one
of
the
articles
was
that
the
networking
may
be
slightly
different,
so
that
is
some
place
where
your
mileage
may
vary,
and
let
me
see
if
I
can
find
my.
A
B
C
C
It
could
end
up
being
slightly
different,
but
this
is
another
one
of
those
ones
that
I
do
not
understand
how
people
end
up
down
this
rat
hole
like
this
is
one
that
I'd
never
understood
like
I've,
never
changed
the
networking
ever
because
you're
running
web
apps
like
why
do
you
ever
need
to
even
muck
with
it,
but
somehow
from
data
and
talking
to
a
lot
of
people,
people
always
go
down
the
rabbit
hole.
Well,
I
got
to
change
ip.
I
got
to
do
this,
I'm
like.
Why
do
you
have
to
do
that
like?
C
C
B
Actually,
really
funny
when
my
when
I
was
going
to
set
up
this
demo,
this
is
the
first
time
in
a
long
time
that
I
went
to
go
put
together
a
demo
that
was
in
vms
versus
inside
a
container.
But
you
know,
as
you
say
that
I
actually
I
don't
know
that
I've
mucked
with
networking
like
like
you're
talking
about,
like
kind
of
you,
know,
rejiggering
how
the
dhcp
works
or
any
of
that
stuff.
I
mean.
C
B
B
Yeah
at
least
30
years
ago,
yeah
yeah,
so
all
right,
so
we
just
created
this
new
container.
So
you
know,
I
don't
think
this
gets
enough
play,
but
you
know
I
still
think
that
this
feature
is
one
of
the
coolest
ones,
that
there
is
I'm
actually
not
gonna
name
it
that
I
want
to
make
it
something
that
I
will
remember,
and
so
what
we
do
is
I'm
just
committing
it.
B
I
can
not
using
a
registry
any
of
that
stuff
just
moving
the
tarball
around,
because
I
thought
that
would
be
like
the
truest
test
in
a
sense
like
I
know,
nothing
has
changed
that
container
image
because
I've
literally
just
tarted
off
right.
So
now
we
are
going
to
and
one
of
the
nice
things
about
you
know
using
vagrant
for
this
stuff
is
just
you
can
do
something
like
this
and
you
always
know
where
your
home
is.
B
B
A
B
Yeah,
probably
yeah,
which
you
know
thanks
tar
all
right
so
now
we're
going
to
jump
to
the
docker
machine,
and
so
you
know,
as
you
saw
in
the
vagrant
file,
this
is
just
set
up
with
the
docker
ce
per.
You
know,
docker's,
installation
instructions,
and
we
are
going
to
try
our
scp,
which
is
probably
going
to
take
a
little
bit
longer.
Yep
that
looks
better
and
then
now
we
conveniently
can
still
just
run
docker
import
and
then.
B
C
B
A
B
C
B
C
B
C
A
B
We'll
have
to
take
that
up
with
the
fedora
people.
If
it's
got,
I
think
you
should
file
a
bug
that
the
ping
response
time
is
a
lie
and
see
how
well
that
goes
all
right.
So
you
know
networking
works.
You
know
it's
kind
of
like
I
don't
know
what
else
to
kind
of
show.
You
know
the
fact
that
it
runs
at
all
is
basically,
you
know
we're
pretty
we're
in
pretty
good
shape.
Yeah.
B
C
B
Two,
yes
yeah,
it's
really
amusing,
so
let's
just
cheat
and
copy
and
paste
okay
and
then
we
can
say
sudo
nope,
podman
import,
and
what
I
love
is
that
it's
exactly
the
same
command
with
you
know
whatever
six
different
letters
except,
I
don't
need
to
do
sudo,
because
I
can
do
this
as
me
now
and
then
I
can
run
and
yeah.
That's
weird.
C
C
By
default
like
something
that
people
might
not
realize,
if
they're
not
deep
in
this
world
like
when
you
did
it
the
way,
you
did
it
because
you
went
from
docker
to
everything
else.
Those
are
actually
docker
specific
image
formats.
It's
like
99.8
the
same
as
oci,
but
it
is
technically
slightly
different.
B
Didn't
work
right
if
it
says
imported
from
tarball,
I
don't
think
I've
seen
that
before
but
yeah
I
mean
you
know
you
can
kind
of
go
and
inspect
it.
I
think
you
can
kind
of
see.
You
know.
I
wonder
it's
probably
this
docker
io
part
here
that
indicates
that
it's
a
docker
image,
but
this
would
be
a
interview
with
dan
walsh
to
know
for
sure
yeah.
C
B
B
Yeah
yeah
exactly
so
that
was
my
little
demo
like
kind
of
the
point
of
the
demo.
Is
that
there's
not
much
to
it?
But
you
know
there's
that
that
I
wanted
to
share
with
y'all,
and
I
noticed
it
is
56
after
the
hour
at
least
in
eastern
for
narendev.
Is
that
36,
probably
or
anybody
else
who's
on
a
a
non-clean
hour
break?
Let
me
share
this.
B
So
we've
had
a
little
bit
of
movement.
You
know
we
had
a.
We
had
a
new
person
show
up
and
and
see
a
couple
of
episodes
over
the
brick.
I
will
say
I
have
a
few
kind
of
administrative
points
that
I
give
out
so
like
initial
show
and
like
some
filing
some
issues
that
are
not
represented
here,
so
I
just
pulled
it
before
that
data
was
added.
So
these
are
a
little
shy.
B
B
And
then
joe
fuzz,
who
you
know
holding
on
at
1800
points
and
then
jp
dave
you
got
to
go,
you
got
to
go,
fill
out
those
forms.
We
know
you're
here
a
lot
so,
but
if
you
want
to
collect
sweet,
sweet
internet
points
which
have
the
value
of
being
listed
on
this
list
of
people
who
have
collected,
sweet,
sweet
internet
points,
you
can
go
and
at
the
bottom
of
the
page
there
you
can
just
go
to
that
form
and
enter
that
code
or
the
other
link.
A
B
Yes,
if
I
can
find
this
stupid
window
right
exactly
so,
and
then
one
of
the
things
that
I've
been
trying
to
do
recently,
but
don't
always
recall
to
do,
let
me
just
throw
this
in
the
chat
first
and
so
to
someone's
question
in
the
chat.
What
are
sweet?
Sweet
internet
points
they
are.
They
are
unfortunately,
not
like
the
35
000
bitcoin.
They
are
a
way
that
we
like
to
encourage
participation.
So
basically
there's
a
bunch
of
different
things.
You
can
do
to
collect
points.
B
You
can
watch
the
episode
and
fill
out
the
form
like
this.
You
can
file
a
pr
which
is
actually
how
netherlands,
hack'em
has
pulled
ahead,
and
you
can,
you
know,
there's
a
bunch
of
other
activities.
So
if
you
go
to
the
episodes
repo,
which
I
just
realized,
I
don't
think
we
ever
put
a
link
to
that
in
the
chat.
B
Let
me
just
grab
it.
If
you
go
to
the
rep
episodes
repo
on
the
activities,
doc,
there's
a
whole
list
of
things
that
you
can
do
to
collect
sweet,
sweet
internet
points,
and
one
of
the
things
I
was
saying
I
I've
been
doing
lately
is
to
try
to
collect
some
people
who
are
new
to
the
show
and
say
hey.
You
know,
I
hope
you
came
back
and
we
would
love
you
to
continue
watching
the
show.
B
Please,
you
know
bring
up
your
points
and
thanks
so
much
for
for
coming
to
the
show.
So
are
we
do
we
want
to
kind
of
wrap
it
up
here?
Do
we
have
any
other
kind
of
points
to
make?
Are
there
any
other
questions
in
the
chat
that
we
haven't
answered?
We
are
just
about
the
top
of
the
hour,
but
I
don't
think
we
have
a
hard
hard
stop
today.
No.
C
B
A
B
A
A
B
Cloud
native
kind
of
crazy,
if
you
are
unfamiliar
with
what
a
service
mesh
is
please
come
to
the
show,
but
it's
you
know
we
kind
of.
In
short,
it
provides
a
bunch
of
kind
of
administrative
features
that
are
hard
to
do
cross-cutting
into
an
application
without
massive
application
change,
but
with
zero
code.
So
it's
it's
pretty
cool.
So
definitely
come
come
to
the
show
next
week
and
chris.
What
else
do
we
have
going
on
today
or
coming
up
this
week?.
A
Coming
up
here
at
11
a.m,
eastern
is
the
open
shift
administrator
office
hour
with
andrew
sullivan,
and
I
I
forget
what
he
has
planned
today.
As
his
discussion
point,
I
think
he
was
still
choosing
between
one
of
the
other
yesterday,
but
we'll
be
there
to
answer
your
questions,
including
you
know,
any
problems
with
hpe
equipment,
so
yeah,
let
us
know
at
11,
am
eastern.
That
is
now
1600.
Utc.