►
Description
We now have some containers but what if we want them to work together. A classic example is a data container and a function container. We have also made various containers, how do we get them around the data center?
Learn more at https://red.ht/leveluphour
A
Good
morning,
good
evening,
good
afternoon,
wherever
you're
hailing
from
welcome
to
another
episode
of
the
level
up
hour
here
on
openshift
tv,
I
am
chris
short
the
host
and
you
know,
executive
producer
of
this
wonderful
channel.
I
guess
is
what
they
call
me
now,
and
I
am
a
I
work
here
at
red
hat
on
the
openshift
team,
with
the
one
and
only
langdon
white
langdon
good
morning.
How
are
you
this
week
good
morning.
C
B
B
There's
a
hackathon
put
on
by
boston
university
that
I
just
blanked
on
the
name
of
tech
together,
that's
what
it
is,
which
is
one
of
the
few:
let's
see
non-binary
or
women-only
hackathon,
and
in
advance
of
that
they
actually
do
a
what
this
thing
they
call
prehax
where
they
basically
get
like
high
schoolers
who've,
never
participated
in
a
hackathon
before
give
them
like
a
day
or
two
before
the
hackathon
to
kind
of
learn
how
to
do
a
hackathon
and
then
they
go
and
do
the
hackathon.
B
So
I
I
like
the
program,
so
I'm
I'm
wearing
the
shirt,
and
now
I
can
see
myself
in
your
background.
Oh
I'm
gone
you're.
A
Gone
sorry,
so
yeah
awesome,
that's
great
to
hear.
Is
it
virtual
this
year
like
so
these
hackathons,
like
I've,
seen
a
few
of
them
like
I've
done
some
looking
lace
recently
and
it's
like
I've
seen
some
events
that
have
gone
back
to
in
person
and
it's
kind
of
scary
to
me.
That's
that's
just
my
sense,
but
yeah.
A
B
That's
that
was
finding
finding
technology
but
yeah
november
6th
through
8th,
and
you
know,
if
you
are,
you
can
be
a
coach
and
you
can
be
of
any
gender
or
you
know
like,
like
a
judge
or
a
coach
or
like
a
helper.
You
know.
B
And
actually
they
put
it
as
all-female
femme
and
non-binary
hackathon
and
it's
kind
of
amazing
to
see
you
know
five
800,
you
know
kind
of
women
doing
tech
stuff.
There
are
a
bunch
of
they've
been
a
bunch
of
cool
projects
like
one
year
they
did
the
one.
I
particularly
remember
I
have
this
term.
B
So
if
you
live
in
the
inner
city,
you
take
public
transit,
a
lot
less
so
currently,
but
you
know
in
theory,
and
one
of
the
plagues
of
the
of
public
transit
is
a
knowing
when
your
transportation
is
coming
right
right,
that's
the
easy
one,
so
that's
mostly
solved
by
apps
like
transit,
but
there's
another
plague
which
is
what's
called
a
ghost
bus,
at
least
that's
what
I
call
it.
B
I
think
she
had
another
term
for
it,
which
I
can't
think
of
now,
but
it
was
better,
but
it
was
that's
when
a
bus
is
supposed
to
be
coming,
but
never
actually
doesn't
actually
exist,
and
so
you're
standing
at
the
stop
waiting
for
the
bus
per
the
app.
B
That
tells
you
that
a
bus
will
be
here
any
minute
and
it
never
shows
up
yeah
that
would
suck,
and
so
she
was
doing
an
app
that
was
crowdsourcing
when
whether
a
bus
was
a
ghost
bus
or
not
so
that,
basically
the
stops
before
you
could
report
that
the
bus
never
was
came
and
it
never.
It
never
existed
so
that
you
wouldn't
be
standing
there
waiting
for
the
bus,
because
somebody
prior
on
the
line
would
report
that
it
was
a
ghost
bus,
so
yeah.
So
I
thought
it
was
really
cool.
B
I
thought
the
projects
in
general
really
cool.
You
know
typical
hackathon
they're,
usually
like
a
really
good
first
draft,
you
know,
but
they
also
give
out
an
award
for
biggest
fail,
which
I,
those
are
the
ones
that
are
often
the
most.
B
A
Twitch
shows
yeah,
you
know,
I
wonder
if
we
could
have
a
biggest
failed
twitch
show
everyone.
B
A
A
Right,
like
it's
no
big
deal
right,
like
we've,
had
shows
that
ended
with
stuff,
not
working
and
then
literally
20
minutes
later.
Oh,
it's
this
one
thing:
everybody
twitter
exactly
so
yeah!
That's
not
like!
That
sounds
like
an
awesome
event.
I
would
love
to
tell
you
what
I'll
I'll
do
some
more
research
and
I'll
read
about
it
and
I'll
probably
share
my
newsletter.
Maybe
we'll
see.
B
Well,
I
also
I
I
wanted
to
pitch
one
more
thing,
which
is
dev
confused,
which
is
a
conference
that
red
hat
supports.
There's
you
know
we
being
a
corporation,
it's
kind
of
hard
to
take
money
right
for
for
a
lot
of
things.
B
So
so,
if
it's
not
in
your
kind
of
core
business,
it's
kind
of
weirdly,
you
know
it
may
not
make
sense
to
somebody
who's,
not
versed
in
how
corporations
work,
but
it's
actually
a
little
difficult
to
to
take
money
for
things
that
aren't
actual
like
things
that
you
sell
right,
yeah.
B
C
B
We
do
a
conference
every
year
called
devconf
and
we
actually
do
several
of
them.
There's
a
devconf
czech
republic,
which
is
the
first
one
and
that
was
started
11
years
ago.
I
think
this
will
be
the
12th
actually
upcoming,
that's
in
january,
and
then
we
do
one
in
india
as
well,
which
is
actually
usually
right
around
august,
but
virtual
has
made
everything
crazy
and
then
there's
also
the
one
I
happen
to
help
run,
which
is
dev
conf
us,
which
is
in
about
two
weeks.
B
It's
on
the
23rd,
there's
workshops
and
then
two
days
of
talks
and
it's
all
about
like
it
can
be
anything,
but
it
tends
to
be
because
it's
red
hat
sponsored
it
tends
to
be
the
upstream
projects
we're
involved
in
so,
for
example,
there's
talks
about
podman,
which
we've
been
talking
about
a
lot
here.
There's
talks
about
core
os
or
fedora
core
os
right.
There
will
be
talks
about
okd.
You
know
the
the.
B
Yeah
so
there's
talks
about
all
the
containery
things.
There's
talks
about
the
cloudy
things
and
the
way
I
actually
arrange
the
tracks
is
it's
about.
We
do
talks
about
like
things
that
run
a
cloud
versus
things
that
run
on
a
cloud
right,
and
so
the
things
that
run.
B
B
Thank
you.
We
also
have
just
devconf
dot
us
as
well,
because
yeah.
B
Because
we
manage
the
whole
site
as
one
thing
right,
so
yeah
so
check
it
out.
You
know
we
don't
usually
get
a
huge
number
of
people,
but
it's
a
great
way
to
meet
red
hat
engineers.
It's
a
great
way
to
meet.
You
know
kind
of
other
people
in
the
open
source
community,
because
it's
very
it's
basically
you
have
to
it-
has
to
be
open
source.
It
can't
be
product
facing
or
product-based
it
can't
be
product,
pitches
and
yeah.
B
I
think
we
have
a
lot
of
good
stuff
this
year,
but
again
virtual,
so
you
can
also
watch
it
on
demand.
If
you
like.
B
You
know
failing
you
know,
aside
from
failing
technology
or
whatever
you
know,
we
record
everything
we
run
and
then
make
playlists
on
youtube
of
it
all,
so
you
can
go
and
see
any
of
it
at
any
time
and
a
lot
of
it's
really
interesting.
So
those
are
my
those
are
my
pitches
for
the
day.
I
appreciate
those.
A
A
No,
so
the
the
the
only
pitch
I
have
is
that
they're,
you
know
from
the
cncf
side
of
things
being
a
cncf
ambassador
envoy
cons
coming
up.
It
is
to
a
virtual
event.
These
virtual
events
do
cost
money
to
throw
as
we
learned
with
kubecon.
They
spent
a
lot
of
money
on
that
yeah
and
it
yeah
so
envoy
con.
If
you're
not
familiar
with
envoy,
it
is
literally
the
thing
underneath
all
the
service
meshes
now
so
right
kind
of
a
useful
thing
to.
B
A
If
you're
not
familiar
with
envoy
envoy
con,
is
this
the
cheapest
you're
ever
gonna
get
into
it
right,
50
bucks,
and
if
you
need
help
with
it,
I'm
a
cncf
ambassador.
Let
me
know
I
can
probably
get
you
some
help
if
you're
definitely.
B
In
that
is
definitely
a
misnomer,
I
mean
you
know,
I'm
experiencing
the
same
thing
with
devconf
right,
the
yes,
it's
significantly
cheaper,
not
having
a
venue,
but
we
still
have
to
pay
for
all
the
software.
B
And
so
yeah
there's.
A
The
problem
with
with
the
whole
situation
right
now
is
that,
like
you
know
the
whole
reason
we
started
open
shift
tv,
we
need
to
reach
people
more
effectively
more
cheaply
more.
You
know,
like
there's
a
very.
B
A
B
Yeah
I
was
actually
going
to
talk
a
little
bit
about
today.
You
know
because
one
of
the
things
that
you
know
basically
some
of
the
topic
of
the
show
today
and
we'll
probably
maybe
we
should
start
there
and
maybe
we
should
kind
of
move
into
it,
but
it
is
basically
the
distribution
okay.
So
we
made
all
these.
You
know
helper
containers,
the
toolbox
containers,
the
whatever,
but
how.
A
B
Yeah,
exactly
nobody
ever
wants.
Aj.
B
I
just
realized
I
didn't
actually
log
into
my
dealio
hold
on
a
sec,
then
I
can
start.
A
Did
you
log
into
like
oh
zoom,
you
didn't
log
in.
B
No,
my
ssh,
like
I
just
didn't,
ssh
to
my
machine
yet
so,
but
let
me
go
ahead
and
share
my
screen.
A
And
if
no
one
is
familiar
with
my
t-shirt
set
and
force,
one
is
an
sc
linux
command
and
we
encourage
people
to
leave
sc
linux
on
and
the
next
time
you
will
see.
This
t-shirt
is
when
dan
walsh
is
joining
us
on
the
channel
on
the
24th
so
subscribe
to
the
calendar,
if
you're
not
already
and
dan
walsh
will
be
on
the
channel
soon,.
B
Yeah,
isn't
there
there's
a
brand.
C
A
Website,
yes,
so
there
is
a
website
for
the
t-shirt
where
you
can
buy
it.
The
new
show
that's
coming
up
is
called
in
the
clouds
with
red
hat
leadership,
and
it's
actually
starting
this
week.
Yes,
the
10th
and
we'll
have
joe
fernandez
and
that'll
be
an
interesting
show,
because
joe
fernandez
is
the
vp
of
oh
gosh.
I
forget
his
official
title.
Oh
boy
cloud.
A
Like
it's
like
vp
of
cloud
platforms
and
blah
blah
blah
right
like
so
like
super
vp
and
gm
of
core
cloud
platforms,
there
we
go
yeah,
so
that
is
kind
of
a
big
deal,
and
if
your
cios
are
interested
in
what
red
hat
might
be
doing
or
your
directors
or
your
boss
or
whoever
that
would
be
a
good
show
to
like
recommend
they
tune
into,
because
you
know
we're
helping
you
get,
the
tech
done,
help
them
get
the
culture
done
right
like
that.
A
I
think
is
going
to
be
the
next
piece
of
this
big
puzzle
of
transitioning
to
a
cloud
native
infrastructure
is
getting
everybody
on
the
boat
right
right,
not
just
captain
cube
and
phippy
and
z,
but
like
phippy's,
boss
and
z's
boss,
and
you
know,
captain
cube's
boss.
They
all
need
to
be
on
the
boat
as
well.
Exactly.
B
So
yeah,
so
this
is
the
website
I
was
thinking
of.
It
is
stop
disabling
sc
linux.
It
is
put
up
as
as
it
says,
down
at
the
very
bottom
in
small
letters.
A
public
service
from
major
hayden
who,
when
the
site
went
up,
was
not
a
redhead
employee,
but.
C
B
And
is
this
the
t-shirt
you're
wearing?
No?
This
is
a.
This
is
a
slightly
different
one,
but,
as
you
can
see,
it
does
make
dan
walsh.
A
A
A
B
Okay,
all
right
so
yeah,
so
major
hayden
as
spot
says
in
the
chat.
Major
rocks
major
is
a
big
contributor.
I
guess
to
openstack
right,
yeah
and
very
involved
in
that
community
he's
also
very
involved
in
the
federal
community,
which
is
where
I
know
I'm
from
so
yeah.
So
every
time
you
set
four
zero.
You
make
dan
walsh
weep.
A
B
Joking
aside,
there's
a
lot
of
tools
in
in
kind
of
the
rel
flavors
of
linux.
Right
that
help
you
discover
why
sc
linux
is
angry
with
you.
B
And
you
know,
using
those
tools,
makes
it
much
easier
to
quickly
discover
like
why,
whatever
it
is,
isn't
working,
you
know.
So
whenever
I
run
into
a
problem
that
is,
you
know,
that
seems
to
be
kind
of
security
related.
You
know,
I
kind
of
you
know
in
some
sort
of
controlled
fashion.
Right
I
turn
off
the
firewall.
I
turn
off
set
and
force,
or
I
turn
off
sc
linux
and
then
I
kind
of
slowly
add
pieces
back
until
I
figure
out
what
the
actual
problem
is.
Then
I
can
fix
it.
A
C
A
Very
much
my
pardon
the
dog.
This
is
very
much
how
I
stood
up
services
in
the
military
back
in
the
olden
days.
You
know
the
early
2000s
was.
We
would
establish
that
unsecured
connection
and
put
nothing
over
it,
but
a
test
signal,
then
layer
on
the
security
and
layer
on
the
additional
circuits.
You
know
like
the
individual
services
and
everything
else
that
are
also
individually
encrypted.
A
A
You
know
back
to
where
they
should
be,
as
you
modify
files
right
like
there's,
a
single
command
that
you
can
just
apply
to
like
if
you
just
put
a
bunch
of
stuff
in
your
html
directory,
for
example,
under
your
web
server,
it's
just
this.
I
think
it's
set
recon
or
something
or
yeah
yeah,
and
you
just
apply
this
one
command
and
it's
done
and
like
I
love
the
fact
that
christian
hernandez
is
like
helper
node
like
has
that
in
there,
because
he
never
turns
off
sc
linux
and
he
builds
this
amazing
node.
A
That
stands
up
openshift
for
you
and
it's
all
done
with
sc.
Linux
turned
on
in.
A
B
Yeah
so
always
set
in
force
one
I
it's
I'm
actually
super
impressed
myself.
I,
like
literally
almost
never
like.
I
never
have
it
off
now
more
than
you
know
an
hour
or
two.
So
it's
it's
really.
It
has
gotten
so
much
easier
to
actually
do
the
right
thing,
and
so
we
strongly
encourage
you
to
do
the
right
thing
all
right.
So
should
we
start
talking
about
distribution.
B
Okay,
so
let's,
let's
first
talk
about
what
that
means
a
little
bit
first.
So
when
you
create
a
container
image,
what
you're
getting
is
in
largely,
in
fact
a
tarball
and
if
you
are
unfamiliar
with
tarball
tarball
is
a
very
ancient
term
that
represents
basically
a
set
of
files.
So
if
you're,
if
you're,
you
know
relatively
new
to
computing,
someone
might
think
of
like
a
zip
file
right
so
except
a
tarball
by
default,
doesn't
actually
have
any
compression.
So
a
zip
file
by
default
has
compression.
B
Tar
is
also
fully
open
source,
unlike
zip,
although
it
may
be
now
but
zip
used
to
be
a
actually
have
a
a
license
on
it
that
you,
you
actually
weren't
supposed
to
use
it
as
much
as
everyone
did
right.
A
B
But
it's
better
now
yeah.
I
think
it
might
have
gotten
fully
open
sourced
somewhere
along
the
way,
but
there's
actually
better
compression
algorithms
now
et
cetera.
So
you
know
if
you're
interested
in
that
sort
of
thing,
definitely
check
out
something
like
xz,
which
has
a
very
good
compression
algorithm
super.
A
B
B
All
of
our,
like
almost
I
would
say,
maybe
not
universally,
but
pretty
much
all
of
our
major
like
image,
distributions
of
any
kind
so
like
it
might
be
a
fedora
iso
or
it
might
be
a
you
know,
or
it
might
be
a
container
image
or
whatever
are
almost
all
incr
compressed
sorry
with
xz,
because
of
how
nice
a
job
it
does
on
binaries,
so
tarball,
basically
just
a
file
structure.
So
you
know
if
we
did
like.
B
Let's
see
if
we
have
new
blender
render
in
here,
we
just
say
tar
c,
which
is
create
f,
which
is
the
file
name
we
want
to
make
and
then
we're
going
to
just
call
it
cow,
because
I
named
everything
cow
and
we
just
say
I
want
everything
here.
B
Did
I
do
something
wrong?
Oh
sorry,
not
dot
star.
It
seems
to
just
like
my
commands,
because
it's
a
day.
B
Yeah
it
worked.
It
just
gave
me
a
weird
error.
Yeah,
I
think,
like
tara's
tar's
been
around
a
really
long
time.
There's.
A
Hacking
away
here,
but
yeah,
there's
there's
one
about
tar.
It's
actually
quite
hilarious.
Oh
my
god.
B
So
the
thing
about
tar
is
that
when
you
just
kind
of
do
everything
by
default
and
you
throw
like
just
a
star-
it
keeps
the
directory
structures
which
is
really
important
on
linux,
right
because
on
linux,
everything
is
a
file,
so
keeping
directories
and
all
that
jazz
is
really
important
to
how
it
works.
So
if
you
look
at
a
container
image,
it's
basically
just
a
tarball
that
I
don't
even
know
if
it's
compressed,
I
think
it
must
be.
But
so
what
happens
is
it's?
B
You
know
kind
of
the
basically
the
root
file
system,
of
whatever
the
container
is
that
you
wanted
to
create.
So
once
you
have
that
tarball
great.
What
do
you
do
with
it?
Well,
okay!
So
that's
where
the
problem
is
because
if
you
look
at
my
podman
images,
you
know
they're
they're,
decent
sized
and
there's
definitely
some
complaints
about
that.
Right
I
mean
you
can
do
things
like
alpine
or
whatever,
and
some
of
those
will
be
smaller.
There's
a
lot
of
effort.
B
I
know
within,
like
the
fedora
community,
to
make
smaller
libraries
there's
actually
one
called
fedora
minimal.
A
C
B
B
Actually,
I
don't
know
why
that
broke,
although
it
could
show
us
a
little
bit
about
how
the
containers
work
on
linux,
we
can
look
at
registries.conf
and
this
kind
of
goes
into
what
we're
talking
about.
A
B
Oh,
oh
interesting,
so
by
default.
B
C
B
That
works
yeah,
okay,.
A
We
oh
just
to
let
people
know
what
I
did.
I
did
a
curl
dash
I
on
registry.fedoraproject.org,
which
said:
oh,
it's
a
302
which
is
a
redirect
to
the
https
version.
I
then
did
a
curl
dash
capital
I
to
that
https
version,
and
then
it
did
gave
me
a
ton
of
information
about
what
is
running
at
that
it
gave
me
the
headers
of
whatever
was
responding
at
that
url.
So
I
could
tell
this
is.
A
Capital,
I
capitalize.
Yes,
that's
that
gives
you
just
the
headers.
So
if
you
see
here
it
tells
you
the
proxy
server
the
request
id
so
there's
clearly
some
you
know
something.
A
Here
and
something's
listening
here
so
basically
to
me,
it
makes
sense
based
off
what
I
know
about
containers
and
the
fedora
project.
The
way
they
distribute
everything
registrydafadoraproject.org.
This
verifies
that
something
is
working
because
you
got
a
200,
so
something
should
respond
with
some
kind
of
error
code.
If
it's
not
the
right
thing
there
you
go.
B
Yeah
and
if
you
really
want
to
be
a
jerk,
you
could
always
un-map
them
yeah.
So
I.
A
Like
a
shortened
link
or
anything
like
that,
curl
dash
capital,
I
short
and
link,
will
tell
you
what
it
expands
out
to
and
then
you
can
just
go
to
that.
If
you're
worried
about
you
know
something
not
being
like
right
or
whatever
you
know
so
yeah
like
totally
use
that
to
your
advantage,
yeah
exactly
if
you
do
that
here,
it'll
tell
you:
oh
it's!
It's.
B
C
B
I
think
I
typed
the
url
wrong
where
I
now
I'm
sorry.
B
A
Yes,
so
I
said,
remember:
narendev,
don't
drink
the
kool-aid
when
you're
going
to
all
these
meetups
right
like
be
careful.
Kool-Aid
is
not
a
ubiquitous
term,
apparently
in
the
the
world.
So
now
we're
talking
about
colts,
which
is
one
of
my
least
favorite
discussions.
Awesome
yeah.
B
Oh
and
actually
tillmore
has
explained
the
the
origin,
although
I
will
say
it
is
cool
with
a
k,
because,
yes,
that
cool,
not
cool.
A
B
Really,
I
don't
think
I've
seen
that
that's
I
might
have
to
go
find
that
myself
so,
though,
that
slight
digression,
okay.
So
what
I
wanted
to
point
out,
though,
before
we
moved
on,
was
this
where
I
put
in
fedora,
if
you
notice,
I
kind
of
stuck
it
in
the
middle
and
the
reason
I
did,
that
is
because
it
searches
these
in
order
and
it
will
by
default.
B
If
you,
if
you
don't,
if
you
don't
like
fully
qualify
the
target
that
you're
looking
for
it's
gonna,
find
the
first
one
that
it
finds
and
use
that
first,
so
in
other.
C
B
There
is
a
fedora
in
the
docker
io
registry,
but
normally
almost
always
when
I
want
to
pull
fedora.
What
I
want
to
do
is
actually
pull
that
from
fedora
project
directly,
not
from
docker
so
and
that's
kind
of
legacy
of
when
the
fedora
project
didn't
have
a
docker
registry.
So
you
know
it
kind
of
made
sense
to
have
it
in
docker
io.
But
this
way,
if
I
just
type
fedora
like
a
podman
pole
fedora,
I
will
get
the
one
from
fedora
project
rather
than
one
from
docker
dot.
B
Io
excuse
me,
and
the
same
is
true
for
any
of
the
red
hat
ones
as
well.
B
So
if
I
pull
hgvd,
for
example,
on
adulterated,
you
know
unfettered
unfettered,
I
will
actually
get
the
one
from
red
hat
rather
than
the
one
from
fedora
or
the
one
from
docker
and
depending
on
the
scenario
I'm
in,
I
may
want
to
do
different
things
right,
but
in
the
case
of
this
rel
box-
and
this
show
I
want
to
pull
from
red
hat
first
and
then
and
all
other
second
right,
but
I
also
want
to
put
fedora
as
a
higher
priority
than
which
we'll
call
it.
B
The
other
thing
to
point
out
here,
though,
is
that
a
nice
thing
about
podman.
So
if
you
install
docker
on
rel
or
on
fedora,
although
the
package
naming
has
gone
back
and
forth
because
they
changed
the
name
to
moby
for
a
while
and
now
back
to
docker-
and
I
don't
know
but.
B
Remember
is
this
file
is
editable
and
immediately
consumable
or
consumed
by
podman
when
you
run
the
next
command,
but
in
docker
you
have
to
restart
the
the
daemon
to
get
it
to
pick
up.
The
registry
change
so
just
kind
of
keep
that
in
mind.
They
operate
slightly
differently.
That
way,
if
you're,
you
know,
if
you're
ever
using
docker,
and
you
don't
need
to
modify
your
registries
so
there's
a
bunch
of
other
stuff
in
that
containers
directory.
B
B
Bunch
of
different
ways
that
you
can
store
so
the
member
I
mentioned,
the
the
containers
are
basically
tar
balls.
Well,
those
tar
balls
can
be
stored
on
disc
in
different
ways.
You
know
so
the
cheapest
and
easiest,
but
probably
most
dangerous
way,
is
basically
as
a
loopback
device.
B
But
then
I
think
the
default
now
is
actually
to
use
this
thing
called
overlay
fs,
which
basically
allows
if
you
notice,
when
we
build
an
image,
let's
see
if
I've
got
one,
I
can
show
quickly.
A
Quick
shout
out
to
my
father-in-law:
that's
working
in
the
basement
right
now
here,
an
outlet
for
me.
None
of
this
show
would
be
possible
without
him,
because
he
is
the
person
that
was
able
to
manage
to
fish
a
ethernet
line
down
from
upstairs
to
the
office
here
so
that
I
could
get
hardwired
across
the
office
and
bring
this
show
to
you.
So
thank
you
father-in-law,
who
remain
nameless
for
privacy
reasons.
A
Yeah
he
is
a
certified
electrician,
most
of
columbus
ohio,
the
downtown
of
that
he
wired
so.
A
A
Here
in
detroit
and
we've
told
him
he
can't
travel
so
right,
yeah,
no.
B
Yeah
so
the
reason
we
use
a
file
system,
that
is
an
overlay
file
system
and
overlay
in
this
sense
means
basically
it
only
changes
things
when
they
are
changed.
So,
in
other
words,
you
do
a
layer
of
the
file
system,
and
then
you
put
something
on
top
of
it.
Hence
the
overlay,
and
then
you
get
differences
between
them
and
basically
that's
what
supports
why
I
thought
I
already
built
this
here.
Oh
no,
oh,
is
it
yeah?
B
It
should
be
okay,
but
the
point
is
each
of
these
steps
here
is
a
layer
in
the
container
image
and
that's
that's
supported
by
overlay
fs.
This
has
gone
back
and
forth
of
what
file
system
to
use
for
this.
I
can't
even
remember
what
the
original
one
was,
but
overlay
fs
is
now
in
the
kernel
and
is
used
by,
I
believe,
docker
and
podman.
Yes,
it
is.
B
Yeah,
I
think
that
might
be
right,
but
the
point
is
it
never
made
it
into
the
kernel
for
a
bunch
of
reasons,
I
don't
remember
the
details,
but
this
but
anyways
well,
this
other
one
overlay
fs,
which
actually
had
been
sitting
around
for
a
while
and
was
just
kind
of
like
hanging
out,
got
reinvigorated
and
then
made
it
into
the
kernel
and
so
it
got
adopted.
So,
oh
cockpit,
oh
my
one
of
my
other
favorite
things.
B
Yeah,
anyway,
so,
okay,
so
now
we
have
this
tarball.
So
what
we
want
to
do,
though,
is
we
want
to
distribute
it
around
right,
so
we
want
to
get
it
to
so
then
much
like
a
linux,
repo
or
repository
in
containerland.
We
talk
about
essentially
the
same
idea
except
it's
called
a
registry,
and
you
will
actually
often
hear
me
say,
like
the
word
repository
when
I
mean
registry,
because
they
essentially
serve
the
same
function,
the
biggest
difference
being
that
with
a
registry
you
can
push
to
it.
B
B
Right
and
then
then
you
have
to
get
repository
which,
just
you
know,
completely
screws
everything
up
which
isn't
even
really
kind
of
core
central
to
get.
In
a
sense,
it's
I
mean
that's
what
the
git
forge
has
brought
but
yeah.
So
when
we
talk
about
registries,
there
are
several
available
on
the
interwebs
and
the
one
we
were
talking
about
the
beginning,
which
was
easily
the
most
popular
for
a
long
time
was
docker
hub,
which
has
gotten
increasingly
hard
to
actually
find
the
container
images
in
yeah.
B
It's
not
that
I'm
bitter
or
annoyed
by
that
at
all,
but
a
couple
things
that
I
would
point
out
here,
as
you
mentioned,
if
you're
doing
personal
stuff,
if
you're
pushing
things
to
this,
it
now
has
a
lifespan
as
far
as
how
long
it
will
stay
there
on
its
own.
B
You
know
unless
you
pay
for
an
account,
so
do
keep
that
in
mind
if
you're
looking
for
something,
however,
what
I
did
want
to
point
out
is
that
you
know
in
some
ways
you
know
people
consider
you
know
a
docker
a
reputable
source,
so
they
have-
and
this
isn't
showing
very
well
because
the
colors
I
chose,
but
they
have
a
number
of
images
that
that
docker
creates
versus
random
people
on
the
internet,
create
those
are
indicated
by
an
underscore
here
in
the
url
yeah.
So
you
can.
C
B
Pretty
confident
that
things
that
are
official
from
docker
are
likely
safe
and
things
you
can
kind
of
run
from
the
internet.
I
still
don't
necessarily
recommend
it
unless
it's
like
a
base
os
like
fedora
itself,
but
in
httpd
for
example,
I
often
want
what
I
will
often
do
is
actually
just
gonna
guess
here
and
hope
that
that's
there.
So
what
I
will
often
do
is
actually
go
and
take
a
look.
B
B
Okay,
well
so
what
I
will
often
do
is
actually
go,
and
so,
if
it's
not
maybe
if
it's
the
unofficial
ones,
so
let's
go
back
to
our
stupid
telnet
example
stupid
telnet.
Here
it
is
so.
What
I
will
often
do
is
actually
go
and
pull
this
docker
file.
B
You
know
and
then
build
that
thing
locally.
So
I
know
that
what
it
says
it
is
is
actually
what
it
is
right,
but
I
can
at
least
leverage
this
content
to
to
make
my
life
easier,
so
docker
hub,
you
can
push
stuff
to
it
and,
like
all
other
registries,
you
have
to
log
into
them
first,
but
the
other
one
we
wanted
to
show.
You
was
not
tech
together,
quay
dot
io,
which
is
so
okay.
B
B
Yeah
decided
that
q-u-a-y
for
their
thing
would
be
pronounced
quay.
So
that's
why
I
say,
even
though
it
is
technically
supposed
to
be
pronounced
key,
which
I
only
learned
a
couple
of
years
ago,
because
all
of
my
vocabulary,
I
got
from
reading
books
and
the
downside
to
reading
books
is
that
there's
no
pronunciation.
A
Key
pronunciation,
yeah
funny
story.
When
I
was
in
high
school,
what
is
it
things
fall
apart,
the
classic
african
novel
right
about,
like
I
never
had
seen
the
word
machete
before
so
I
completely
mispronounced
that
every
time
there
was
like
a
group
reading
in
class
yeah,
I
was
like
every
time,
but
it
was
something
different.
It
was
like
cube
cuddle
right
like
every
time.
It
was
something
different.
A
B
So
narendev
in
the
chat
says:
if
we
click
on
tags,
it
will
take
us
to
the
github
repo.
Somehow.
A
B
B
B
Which
we
have
definitely
talked
about
in
the
past,
but
in
order
to
save
myself
some
time
I
will
often
you
know,
use
the
docker
file.
That's
there
as
a
starting
point.
Technically
speaking,
you
should
probably
go
and
review
the
licensing
on
the
docker
file
if
you
are
planning
on
using
it
directly,
but
using
it
as
a
as
a
hint
to
what
you
want.
I
mean
half
the
time
like
you
know,
or
some
portion
of
the
time
I
keep
losing
the
koi
window.
B
I
don't
know
why
yeah
it's
really
like
losing
it
on
me,
but
a
good
chunk
of
the
time.
It's
not
going
to
show
me
the
or
it's
not
going
to
use
the
base
os
that
I
want
you
know
so
you
know
so.
I've
switched
that
around
you
know
and
then
different
different
linuxes,
of
course,
occasionally
not
consistently,
but
occasionally
use
different
names
for
the
different
packages.
So
yes,.
A
B
Example,
apache
versus
htpd
yeah.
B
B
I
have
not
very
much
used
quay
or
I
like
I
swear
to
god
I
did
and
then
but
it
seems
to
be
m.I.a
or
maybe
a
different
account
or
something.
So,
let's
call.
A
A
B
So
we're
just
going
to
do
an
empty
repository
and
I
don't
know
why
the
fonts
are
so
off.
I
feel
like
I
have
this
zoomed
in
you
need
to.
B
Yeah,
but
it
seems
like
I
was
noticing
this
on
a
different
page
too.
What.
B
You
go
that
looks
like
yeah,
it's
fixing,
oh,
I
might
have
this
browser
set
to
automatically
zoom
to
maybe.
B
And
then
so
what
I
can
do
is
push
to
it
by
using
something
really
obvious.
So
the
thing
to
get
that
gets
tricky
about
this
is
that
you
need
to
tag
things
for
the
target,
and
so,
let's
explain
tags
just
a
little
bit
so
here
you'll
see
that
this
guy
has
a
tag
of
latest
and
but
this
is
actually
also
referred
to
as
a
tag
a
lot
of
the
time,
even
though
it's
actually
a
name.
B
So
did
you
have
a
comment?
Chris?
Oh
no,
sorry!
So,
even
though
it
says
repository,
it's
also
referred
to
as
a
tag.
So
what
I
can
do
is
say:
okay,
so
I
got
this
localhost.
How
do
I
and
if
you
don't
have
a
name,
so
you
just
have
a
nun
here.
You
can
use
this
guy
here,
so
the
image
id.
So
I'm
going
to
use
the
name
of
this
one
here,
assuming
if
I
could
actually
click
all
right
and
then
I'm
going
to
tag
that
guy.
B
As
I
don't
know,
new
and
what's
going
to
happen,
is
now
I
have
a
new.
How
do
I,
except
if
you
notice
it,
has
the
exact
same
image
id
and
if
I
do.
B
Rmi,
okay,
so
you
know
remove
image
of
the
new.
How
do
I
all
it
does
is
untag
it.
It
calls
it
even
though
weirdly
it's
under
the
repository
flag
and
not
the
tag
flag
but
or
column
rather,
but
if
you
notice
it's
gone,
but
my
how
do
I
one
is
still
there
so
in
order
to
make
this
work
with
something
like?
Actually,
let's
do
sudo
pod
man,
images.
B
We
should
have
to
do
quay
dot,
io,
slash,
langdon,
slash,
perfect
tools,
so
I'm
going
to
guess
that
we
can
always
change
it.
We
can
have
as
many
tags
as
we
want
right
because
they're
just
they're
just
labels
on
the
same
tarball
and
what
did
I
call
perfect
tools.
A
B
As
a
read-only
registry,
yes,
yeah,
it's
slightly
more
difficult
to
do
a.
A
B
Go
right
exactly
and
there
there
are
lots
of
them,
and
you
know
so
running
your
own
is
actually
not
too
big
on
rel
seven,
you
can
actually
run
a
supported
instance
of
quay
as
well.
I
believe,
just
included
in
a
normal
subscription
on
rel
eight.
It
is
currently
not
supported,
but
there's
like
I'm
I'm
hopeful
that
is
going
to
change,
but
we're
also
going
to
talk
about
the
openshift
registry
as
another
way
to
do
this.
B
But
we
probably
won't
cover
that
today
because
you
have
to
set
some
flags
on
the
on
the
openshift
registry,
but
so
the
way
openshift
works,
as
does
kubernetes.
It
actually
has
an
internal
registry
where
it
it
holds
all
of
the
containers
you're
deploying
into
your
ecosystem.
B
But
you
don't
have
to
just
use
that
registry
from
inside
openshift.
You
can
also
use
it
from
outside
openshift,
because
it's
just
you
know
to
asdf
monkey's
point
right:
it's
just
a
web
server
that
happens
to
be
sending
and
pulling
tarballs.
So
that's
how
I
often
will
set
up
a
local
network
one
because,
like
many
straightforward,
looking
simple
software
to
write,
it
is
actually
more
or
in
practice.
If
you
want
something
that
is
performant
and
scalable,
it
is
not
quite
as
simple
to
write.
B
So,
for
example,
you
know
a
package
manager
actually
had
a
student
at
boston
university
who
decided
they
wanted
to
write
a
new
package
manager.
A
No
exactly
like,
like
dependency
management
package
management
like
those
are
like
now
like
it
is
things
right.
B
It
is
like
way
harder
than
it
looks
it
is,
you
know,
and,
and
the
worst
part
about
it
is
much
so
I
actually
wrote
a
database
years
ago
to
discover
the
same
problem,
which
is
it's
very,
very
easy
to
write
a
very
bad
one.
C
A
B
Well,
the
thing
that's
hard
for,
I
think,
an
experienced
developer
with
certain
classes
of
software
package
managers,
databases
and
when
I
say
package
manager,
I
kind
of
include
like
things
like
registries
and
repositories
and
all
that
they're
they're,
all
of
the
same
ilk,
databases,
source
control,
there's
one
more
too
that
I
use
as
an
example
is
that
if
you're
an
experienced
developer,
it
looks
like
an
easy
problem
and
you
go
and
build
one.
And
then,
oh,
my
god,
you
realize
how
terrible
a
job
how
hard
this
could
be.
A
B
So
yeah
well
so
back
in
the
day,
I
needed
an
in-ramp
database
in
java
and
they
didn't
exist,
and
so
I
ended
up
writing
one
for
doing
natural
language
correction
like
like
voice
recognition,
correction
yeah.
So
it
was
not
pretty,
but
it
worked.
So
you
know
good
enough
all
right
so
back
to
our
point,
and
I
don't
want
to
delete
my
own
permissions.
B
What
I'm
looking
for
is
my
login,
which,
let's
just
try
logging
in
so
I'm
going
to
sudo
okay,
so
this
one's
a
little
weird
too,
oh
boy,
so
for
all
of
these
registries
to
push
to
them
intuitively
enough.
You
have
to
log
in
so
I'm
going
to
do
some
guessing
and
hopefully
I
will
be
correct.
A
I
feel
like
we've
worked
enough
together,
that
I
could
understand
your
sql
instrumentation
right
right.
B
B
B
Do
right
right,
but
it
was
in
ram.
Actually
what
was
really
interesting
about
that
project
was
trying
to
do
because
the
problem
database
is
right.
Is
you
want
to
cache
what
the
next
thing
is
going
to
be
as
much
as
possible,
speed,
yeah
and
when,
when
you're
talking
about
language
and
what
we
were
literally
trying
to
do
was
take,
you
know
like
a
sentence
structure
and
then
make
corrections
based
on
the
sentence
structure
about
what
the
next
word
should
be
when
voice
recognition,
got
it
wrong
right,
but.
B
Yeah
yeah,
and
it
was
funny
because,
like
many
software
that
I've
gotten
into
voice
recognition
until
very
recently
always
hated
me,
I
was
like
the
worst
like.
I
didn't
fit
any
of
their
models
but
yeah,
so
we
used
to
use
this
was
for
mostly
medical
transcription
and
we
got
about
85
accuracy
out
of
the
voice
recognition
itself,
and
then
we
would
take
the
correction
to
try
to
get
up
to
like
95
99.
B
Basically
using
english.
Like
you
know,
looking
at
the
language
structure
problem
is
if
you're
trying
to
cache
knowing
what
the
next
word's
going
to
be
is
as
much
a
natural
language
correction
problem
in
and
of
itself,
so
it
was,
it
was
really.
It
was
not
as
fast
as
I
wanted
to
be,
because
I
could
never
really
crack
that
caching
problem
mind
you.
I
worked
for
ge,
research
and
development
at
the
time,
so
that
kind
of
made
more
sense.
B
Yeah,
so
I
left
because
the
only
way
you
could
get
an
office
was
to
have
a
phd
and
I
wasn't
in
the
mood
to
go,
get
a
phd
because
I
was
tired
of
school
at
the
time
and
then
I
just
kind
of
never
got
around
to
it.
So
going
back
to
our
regularly
scheduled
programming,
so
I
logged
into
quay.io,
I
typed
the
pass.
B
I
had
the
wrong
password
initially,
so
I
used
the
correct
password
and
that
worked
and
if
you
notice
I'm
using
sudo,
even
though
I
don't
have
to
but
the
reason
I'm
using
sudo
is
because
if
you'll
recall,
our
perfect
tools
container
needs
pseudo
to
run
so
the
only
place
it's
built
is
in
roots,
pod
man,
image
repository,
not
in
my
main
image.
B
C
B
There
we
go,
but
here
is
my
image
and
I
think
the
shaw
should
be
the
same.
B
So
let's
just
take
a
look
dc9c,
I
don't
know,
I
guess
they
must
reshot
okay,
so
it
is
now
there.
Well,
it
says
it's
it
looks
like
it
might
still
be
cooking
or
no
it's
getting
it's,
it's
being.
It's
gonna
be
scanned,
so
whatever
so
it's
theoretically
there,
and
so
what
I
wanted
to
do
was
so.
B
The
reason
we
were
talking
about
this
right
is
that,
even
though
it's
a
public
repository,
I
can
do
a
couple
of
things
one
I
can
make
that
repository
itself
private,
but
I
think
that's
only
with
paid
accounts,
so
you
know
we're
just.
B
A
I
think
it's
like
five
or
ten
bucks
a
month.
It's
not
much.
B
Or
so
yeah,
exactly
you
know,
ssh
keys,
I
would
say,
would
be
a
really
common
problem
right
if
you
can
find
a
way
to
not
put
that
stuff
in
in
there
the
better.
But
if
you
can't,
for
whatever
reason
you
know
you
can
just
make
them
private.
B
So
what
we
can
do
now,
though,
is
because,
because
it's
now
sitting
out
in
this
registry
land
right,
we
can
jump,
we
can
jump.
Let's
see
if
I
have
a
terminal,
open
nope.
I
apparently
have
a
lot
of
terminals,
though
so,
even
though
this
one's
called
rally
base
we're
going
to
oh
shoot,
I
think
I
don't
know
if
my
7
is
running.
B
That
guy,
it
should
take
like
two
seconds.
Okay,
oh
you
have.
C
B
A
C
A
B
B
We
can
now
pull
this
guy
over
and
this
is
very
close
to
the
bone.
As
a
container
goes.
You
know
very
close
to
the
kernel.
As
we've
talked
about
a
bit
last
week,
it's
the
closer
you
get
to
the
kernel.
The
more
containers
start
getting
a
little
wonky,
so
this
may
not
work,
but
let's
give
it
a
try.
B
Container
run
label
run
now,
oh
actually
before
I
do
that.
So
now,
if
I
say
podman
images,
we
have
the
quayio
one,
so
we
can
relabel
that
and
make
it
like
a
little
easier
to
use
right.
So
we
can
just
say
tag
the
long
name
and
just
call
it
perfect
tools
and.
B
Exactly
exactly
and.
B
Right
and
actually
what
I
should
have
showed
before
is
that
if
I
delete
this
image,
I
will
actually
still
have
this
one
too.
Yes,
even
though,
like
I
showed
the
other
way
like,
I
created
a
new
tag
and
deleted
the
new
tag,
but
you
can
also
create
a
new
tag
and
delete
the
old
tag.
If
you
don't
want
it
around
anymore,.
A
B
Oh
so
this
I
I
thought
I'd
explained
this
a
little
bit
before,
but
I'm
only
using
sudo
for
this
particular
container,
because
if
you
remember
our
perfect
tools,
container
needs
a
lot
of
priv
for
its
access
from
the
container
into
the
host.
All
of
this
would
work
without
sudo.
If
I
was
using
a
different
container
or
a
different
image,
one
that
didn't
need.
You
know
if
I
was
just
doing
hdpd,
for
example
right.
I
wouldn't
need
pseudo.
B
Right
exactly
the
other
thing
I
wanted
to
point
out
too,
is
that
now
that
I
have
used
a
different
tag,
let's
say
the
one
in
quay
dot
io
gets
updated.
If
I
do
a
pull
on
this
guy,
this
one
will
remain
the
same,
so
do
keep
in
mind
that
they
will
branch.
If
you,
if
you.
B
B
Here
we
are
inside
of
our
perfect
tools
container,
we
shipped
it
from
a
place
where
we
built
it
up
through
quay,
and
then
we
downloaded
it
onto
a
different
operating
system.
You
know
because
it's
roll
seven,
not
rail,
eight
right
and
ran
it
and
it
seems
to
be
working
just
fine.
B
Should
have
less,
I
don't
know,
I'm
trying
to
think
of
a
file
that
would
be
in
here
by
default.
Oh
how
about
password,
so
we
have
less
right.
Less
is
not
there
by
default,
so
we
know
that
we're
using
our
real
one,
yeah
and.
B
Yeah
did
route
to
what's.
A
A
B
A
B
A
B
Yes,
so
yes,
routing
akamai
is
your
friend
so
yeah,
so
traceroutes
there
everything
you
know,
kind
of
all
that
jazz
is
is
doing
its
thing.
So
that's
kind
of
what
we
want
to
talk
about
is
like
how
do
we
distribute
this
from
one
place
to
another
inside
the
data
center?
B
Again
you
know
this
is
public,
so
it
may
not
work
for
all
use
cases,
but
I'm
not
sure
if
we'll
do
it
next
week,
but
soon
we
will
do
we'll
do
the
same
thing
except
using
the
openshift
registry,
that
with
any
luck
you
already
have
in
your
data
center
and
you
can
just
use.
Let's
see
anything
else
I
should
point
out.
Did
I
miss
anything.
A
I
don't
think
so.
There's
a
un
unrelated
issue
being
discussed
in
chat,
I'm
just
kidding.
I
can't
say
your
name,
so
please
forgive
me
medesa.
Just
send
me
an
email
see
short
at
redhead.com
I'll,
get
you
routed
to
the
right
place.
We'll
get
your
question
answered
seems
like
a
like
a
simple
checkbox
or
something
line
is
not
getting
written
out
or
something
in
a
config
file
for
this
problem.
So
yeah
your
perfect
tools,
containers
built
you're
distributing
it.
We
talked
about
scopio.
Do
we
wanted
to
mention
that
at
all.
B
Oh
so
yeah,
so
what
we
do
is
when
we
say
podman
search
and
let's
say
we
do
hcbd
now
we
should
get
a
ton
of
results
because
everybody
likes
apache
yeah.
B
B
Soccer
has
a
bunch
and
fedora,
I
know,
has
a
few
too,
but
I'm
not
seeing
them.
Are
we
on
rail?
Eight?
Are
you
on
railing?
Yes,
yeah,
I'm
not
seeing
the
fedora
ones,
but
they
should
be
coming
back
to
you.
I'm
not
sure
why
they're
not,
but
so
scopedio.
B
Container
it
should,
it
should
be,
let's
do
less
etsy
containers.
B
But
yeah,
so
I
think
if
they
are
a
secured
registry,
so
like
basically
https
versus
http,
they
only
have
to
be
in
the
search
list
and
they
should
just
show
up
it
could
be.
I
don't
know
I
could
be
doing
something
stupid
with
fedora,
but
going
back
to,
I
can't
remember
whose
point
it
was,
but
if
you
want
to
make
your
own
one
and
you
want
to
front
end
it
with
just
hcp,
you
can
do
that
by
using
an
insecure
register
here.
B
In
fact
you
can
you
know
one
of
the
things
that
you
can
certainly.
A
B
So
you'd
probably
have
to
investigate
a
bit.
What
registry
means
here,
like
the
google
foo
on
finding
a
basically
a
container
registry,
I
am
99
certain.
You
can
find
a
container
image
that
already
does,
let's
encrypt
plus
container
registry
and
done
you
may
have
to
build
it
yourself.
B
You
know,
but
largely
just
done,
and
in
fact
this
looks
like
here's.
A
B
Yes,
yeah
and
so,
but
as
we
were
talking
about
a
while
back,
podman
is
kind
of
a
rapper
around
man.
Is
a
rapper
run
dmc,
but
it's
also
a
rapper
around
vanilla
ice
I
had
to.
I
had
to
called
scopio.
B
Let
me
make
this,
hopefully
a
little
easier
to
read
and
basically
but
scopio
I
mean
build.
It
does
too,
but
scopio
has
a
bunch
of
other
kind
of
commands,
available,
commands
related
to
kind
of
interrogating
registries,
and
it's
actually
much
more
sophisticated
than
like
docker
search
was
or
at
least
it
used
to
be,
or
at
least
dockers
used
to
be
because
one
of
the
things
that
you
want
to
often
be
able
to
do
is
list
tags.
B
Let's
do
fedora
minimal
just
minimize
typing
yeah,
okay,
so
I
think
you
have
to
do
this.
B
Yeah,
okay,
so
the
reason
I
wanna
okay,
so
first
of
all,
you
have
to
give
it
a
protocol.
The
registry
v2,
I'm
not
really
sure.
Maybe
I
don't-
I
actually
haven't
tried
this
in
a
long
time.
Is
there
an
oci.
C
B
B
Right
and
there
should
be
a
v1
accessible
as
well,
but
I
think
it
will
just
work
with
the
word
docker
like.
I
don't
think
you
need
to
say
docker,
like
v2
v1
or
whatever.
B
Yeah,
but
what
I
want
to
show
you
is,
we
have
been
almost
entirely
during
the
show,
only
been
using
the
default
tag
latest,
but
so
fedora
current
release,
version
of
fedora
is
32
a
32.
I
can't
talk.
B
And,
however,
and
so
right
now
today-
and
this
is
the
only
thing
that
I
kind
of
just
like
is-
it
doesn't
show
synonyms,
which
would
be
nice,
but
if
I
call
latest
what
I
get
is
actually
32
but
labeled
as
the
latest
tag.
B
B
Yeah
here
it
is
okay,
so
fedora
is
actually
in
a
weird
transition
state
that
it
does
every
six
months
for
like
a
month
or
two
or
a
month
or
so.
Basically,
the
beta
release
of
33
is
out,
but
32
is
still
ga,
so
the
bleeding
edge
version
of
fedora
right
now
is
actually
34,
so
the
the
version
after
the
version
we're
about
to
release.
So
as
a
result,
we
have
three
different
available
versions
at
the
moment.
B
Actually
we
actually
have
more
than
that
we
have
five,
which
is
the
highest
we
ever
have
so
the
oldest
one
that
is
still
supported,
which
is
30.
B
the
next
one
back,
which
is
also
still
supported,
which
is
31
current,
which
is
32
33,
which
is
one
will
be
released
in
october
and
then
34,
which
is
the
one
that
will
be
released
basically
they're
referred
to
as
the
halloween
and
easter
releases
so
around
easter
time
next
year.
But
I
can't
remember
what
the
the
actual
dates
are
like
what
month
it
actually
is,
and
then
you
also
have
another
special
keyword
called
rawhide,
which
is
another
fedora
thing,
which
is
basically
whatever
the
most
recent
version
is
so
34,
but
also
rolling.
B
So
it
will
also
get
immediate
updates.
But
the
point
is
one
of
the
things
that
you
often
want
to
do
in
your
container,
so
in
fact,
in
our
empty
squat
here
it
would
be
much
actually
because
I
don't
know
the
actual
answer
here.
A
Oh
by
the
way
I
shared
a
screenshot
of
your
umpysquat.txt
file
last
week
and
all
my
military
buddies
that
worked
with
me
and
heard
me
say:
empty
squat,
whatever
a
bajillion
times
really
appreciated
it.
B
Well,
I
I
not
only
included
a
video
highlight
of
it
from
the
last
show.
I
also
found
where
in
builda
you
can
make
a
pr
that
would
fix
it
to
include
it,
so
I
may
actually
get
to
it.
Who
knows,
but
I
think
I
found
the
spot
in
the
code
and.
B
All
right
so
find
us
on
twitter
find
us
on
discord.
I
yesterday,
so
you
do
receive
sweet,
sweet
internet
points
for
joining
our
discord.
Specifically,
you
need
to
join
the
level
up
our
channel.
If
I
found
you
in
there,
I
sent
you
a
private
message
with
the
code
to
go
and
get
the
points
for
that.
B
If
you
didn't
get
a
private
message
from
me,
message
me
and
tell
me,
and
I
will
send
you
the
the
points
code
for
that
anybody
who
has
filled
that
out
did
not
get
those
points
yet
so
just
fair
warning,
because
I
did
it
last
night,
so
we
mostly
ended
up
talking
about
distribution.
B
The
show
notes
are
in
this
place,
and
do
you
want
me
to
grab
the
link.
A
B
Are
we?
We
are
restreaming
chat
again
right.
A
Yes,
that
should
be
fixed
youtube.
We've
we've
had
conversation
with
youtubers
today,
already
so
yeah,
oh
good,
okay,
cool.
B
B
Yeah
so
I
I
misallocated
an
extra
100
points
to
narendev.
He
is
still
in
the
lead,
but
I
was
noticing
that
he
would
have.
He
would
be
getting
a
big
bump
from
points
if
he
had
actually
registered
for
points
for
five
shows
and
currently
he
has
not.
A
B
B
Just
screwed
up
so
that's
fixed.
Now,
however,
narendev
is
on
the
bubble
for
having
seen
five
shows,
but.
B
B
I
pretty
much
do
them
all
by
hand
unless
you
find
a
code
on
the
internet
and
submit
that
or
you
attend
a
specific
episode
and
then
give
out
you
know
and
follow
the
code
that
we
we
showed.
B
Right
now,
so
let
me
drop
the
discord
link
because
there
are
points.
B
B
There's
the
level
up
point
form
and
then
the
code
for
today's
episode
or
the
you
know
the
easier
one
that
you
can
also
submit.
If
you
want
to
get
points
from
a
prior
show,
you
need
to
just
go
back
and
watch
those
shows
and
during
that
stream.
Obviously
I
go
through
this
screen
and
you
will
see
that,
as
for
the
five
show
submission
that's
what
I
was
saying,
I
do
those
myself.
B
I
think
so
yeah
cause,
I
think,
he's.
B
Well,
not
it's
not
entirely,
I
mean
I
do
have
a
script
that
actually
does
the
point
calculation,
but
basically
I
have
to
go
in
to.
C
B
One
and
submit
like
this
code
or
the
the
attend
attending
any
show
code
to
join
the
discord.
When
you
do
yourself
highlights.
A
B
Yeah
so
yeah,
so
so
I
will
occasionally
make
mistakes.
I
actually
was
looking
at
the
mistake
I
made
and
really
don't
understand
how
I
made
it
because
it's
like
it
was
a
really
obvious
one,
but
I
I
did
I
did
make
this.
C
B
Perhaps
perhaps
so
yeah
so
like,
I
said
if
you
join
the
discord
and
I
don't
send
you
a
dm
that
says,
here's
the
code
to
go
submit.
Please
just
ping
me
on
discord
and
I
will
send
you
the
code.
B
Excuse
me,
I
was
gonna
look
into
discord
to
see
if
there
was
a
way
that
if
you
join
a
channel,
I
could
auto
message
you,
but
I
haven't
done
that
yet
so
so.
B
A
B
I
figured
I
just
haven't
figured
it
out
and
I
wasn't
a
lot
of
the
times.
Those
are
notifications
that
are
public,
whereas
a
dm
like
in
a
lot
of
those
tools,
a
dm
I
haven't
seen
a
dm
before
so
narendev
is,
is
flexing
on
the
fact
that
he
is
going
to
have
a
lot
of
points
next
time.
B
Next
time
yeah,
I
will
point
out
there
is
the
option
on
the
level
up
form
to
to
just
say
private,
and
if
you
say
that,
for
your
name
or
like
your
nickname,
I
won't
be.
B
B
The
way
I
do
the
sorting
is
actually
by
email
address.
Okay,
so
I
would
actually
notice
that
you
know
hey
all
of
a
sudden.
It's
not
private
anymore.
It's
now
a
nick,
and
so
I
could
I
could
in
theory
I
I
would
in
theory
notice
if
you
wanted
to
be
absolutely
sure
that
it
was
true.
You
should
dm
me
on
discord.
B
Yeah
all
right,
yeah
christian
points
out
my
we.
A
A
People
educated
on
registries
and
how
they
work
you
can
use
scopio,
we
didn't
demonstrate
it,
but
you
can
use
scopio
to
copy
from
docker
hub
to
quay
potentially
right
like
as
like
a
pivot
point,
you
know
just
write
a
script.
Basically.
A
Like
since
podman
is
a
rapper,
you
can
totally
do
that
with
either
tool,
and
you
can
make
that
pivot
to
your
new
registry
of
choice,
if
you're
moving
off
of
docker
hub,
potentially
with
this
tool
very
simply
yeah.
Yes,
not
very
simply.
I
say
that
as
easy.
A
Right
like
could
this
be
easier?
Yes,
everything
could
be
easier,
but
as
easy
as
possibly
today,
you
can
do
a
quick
registry
pivot
with
this
tool.
If
you
wanted.
B
Yes,
and
and
really
writing
any
of
those
scripts
you
can
actually
wrap.
Podman
is
how
I
actually
normally
do
it.
B
I
actually
you've
seen
me
do
it
before
in
the
show.
I
haven't
really
commented
on
it
too
much,
but
I
usually
have
a
make
file
associated
with
any
given
container
one
that
has
a
build
command,
one
that
has
a
run
command,
but
I've
also
have
often
in
the
make
file
a
build
and
then
tag
and
then
deploy
to
registry,
and
I
just
do
it
all
in
one
thing,
and
it's
really
no
big
deal.
B
I
apparently
am
failing
today,
but
as
just
quickly
to
say,
was
in
the
empty
squat.txt
we're
using
ubi.e
or
ubi8.
We
should
in
fact,
instead
of
being
using
latest,
we
should
probably
be
picking
a
tag
so
that
we
know
what
version
we're
getting
going
forward.
Chris,
do
you
want
to
pitch
anything
about
what
else
is
happening
this
week,
or
should
we
wrap
it
because
we're
well
over
time
today
and
now
he's
throwing
a
basketball.
A
No,
I
threw
some
velcro
over
at
my
father-in-law:
did
it
bounce?
Did
it
hit
max's
box
and
then
bounced
over
here?
I'm
sorry
dad
I'll
I'll
get
it
in
just
a
second,
so
yeah,
coming
up
later
today,
we've
got
openshift
commons
briefing
here
at
noon:
eastern
1600
utc
transactions
and
kubernetes
and
openshift
with
spencer,
kimball
from
cockroach
labs
and
then
the
very
first
episode
of
eric
jacobs,
new
show
around
gaming
and
video
games
and
scaling
and
everything
else,
scalable
multiplayer
game
design
with
openshift
premieres.
Today
eric
is
doing
that
at
2
p.m.
A
Eastern,
if
I
can
read
my
calendar
correctly,
1800
utc,
yeah,
four
yeah,
okay,
so
yeah
1800
utc,
so
please
tune
in
for
that.
It'll
be
really
interesting.
Show
around
like
building
game
design,
designing
game
services
like
to
scale
on
openshift,
so
yeah
we
feel
like
there's
a
lot
of
mmos
out
there
that
are,
potentially
you
know,
could
gain
a
lot
from
the
openshift
platform,
as
we
saw
on
our
agonist
stream
a
few
weeks
ago.
A
So
if
you
go
back
in
the
archive
search
for
agonist
ago
nes,
it
was
surprisingly
kind
of
just
worked
out
of
the
box
with
just
a
couple
quick,
tweaks
and
boom.
We
were
scaling
with
a
real
live.
You
know
kubernetes
native
game,
dev
server,
so
it
was
pretty
cool.
B
Also
one
quick
thing,
if
you're
into
gaming
and
are
interested
in
some
of
the
history
of
gaming,
particularly
how
gaming
used
to
be
a
very,
not
technically
open
source,
but
an
open
source
endeavor
back
in
the
day
that
podcast
had
a
whole
season
where
they
talked
about
their.
B
Basically,
they
talked
about
the
the
origin
of
multiplayer
online
games
when
they
were
text
based
and
it's
command
line
heroes,
so
command
line
heroes
season.
Two,
if
I
recall
correctly,
is
basically
all
about
the
advent
kind
of
the
video
game
world
and
how
it
was
a
very
like
community
oriented
thing.
If
you
got
good
enough
at
a
particular
online
game.
For
example,
you
became
what
was
often
referred
to
as
a
wizard,
whether
it
was
a
whatever.
B
Based
game
or
not,
and
you
actually
contributed
to
the
code
about
how
the
game
worked
and
it
was,
it
was
really
very
cool,
so
yeah
command
line
heroes,
and
I
think
it
was
season
two.
If
you.
A
So
one
last
thing-
and
it's
because
I'm
doing
the
recording
for
it
today
there's
a
cool
coo
innovation
on
kubernetes
event,
coming
up
that
the
sales
team
wanted
to
use
the
openshift.tv
format
for
so
I'm
gonna
drop
a
link
to
that.
You
have
to
register
for
this,
but
christian
hernandez,
jason
dobies
and
I
did
the
keynote
so
that
is
definitely
worth
tuning
in
for-
and
I
will
be
interviewing
clayton
coleman
this
afternoon
for
this
very
event.
So
yes,
please
tune
in
for
that.
A
We're
working
very
hard
on
it
to
bring
it
up
a
level
right,
not
to
use
the
show
name.
A
C
A
A
A
Today
and
yes
thank
you,
langdon
for
the
experience
and
the
learning-
and
I
did
learn
something
new
today,
so
your
streak
continues
and
we'll
see
you
next
time
here
on
openshifttv.
Thank
you,
everybody
all
right.
Thanks.