►
From YouTube: The Level Up Hour (E30): Podman v3 and the Cool Store
Description
This week we continue to explore Docker Compose support in the latest Podman release. Let's give Podman v3 a spin and see if we can launch our Cool Store using the Compose-like features. We'll also experiment with exporting the compose pods and run them in OpenShift.
Follow us:
Langdon White https://twitter.com/1angdon
Chris Short https://twitter.com/ChrisShort
What is the Level Up Hour?
Every Wednesday at 9am ET we stream live demos on 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
http://red.ht/3ayHlHI
How to participate in the The Level Up Hour:
Watch on YouTube every Wednesday at 9amET
Chat with us on Discord https://discord.gg/5VMVGJt
Check out our GitHub repo https://github.com/level-up-hour/
About OpenShift:
Red Hat OpenShift is an open source container application platform based on the Kubernetes container orchestrator for enterprise application development and deployment.
https://www.openshift.com/
https://github.com/openshift/
#RedHat #OpenShift #tluh
A
Good
morning,
good
afternoon,
good
evening,
wherever
you're
hailing
from
welcome
to
another
edition
of
the
level
up
hour
here
on
openshift
tv,
I
am
chris
short
host
with
the
most
and
executive
producer
of
openshift
tv.
I
feel
like
I
have
something:
that's
plain
that
shouldn't
be.
I
don't
know
a
phone
went
off
or
something
that
was
weird
anyways,
I'm
joined
by
the
one
and
only
the
illustrious
langdon
whites.
How
are
you
doing
today,
sir?.
B
B
A
A
B
B
B
Yeah,
oh,
that
was
disappointing,
so
maybe
they'll
get
snazzier.
I
wouldn't
hold
my
breath,
I'm
not,
but
you
know
they
might.
They
might
at
least
look
different,
so
that'll
be
fun
and
exciting
for
you.
So
this
is
a
level
up
hour
where
we
talk
about
kind
of
why
containers
are
cool
and
why
you
might
want
to
get
using
containers
yourself
as
kind
of
like
a
rel
admin,
or
you
know
a
system
administrator
who
hasn't
got
a
lot
of
experience
with
containers.
B
You
know
we
also
cover
some
development
activities
as
well.
So
you
know,
obviously,
if
you're
a
developer,
this
might
be
relevant
as
well,
but
then,
with
the
goal
of
kind
of
expecting
that,
as
you
kind
of
get
deeper
and
deeper
into
containers,
you
discover
that
the
need
for
orchestration
is
really
really
important,
and
so
you
tend
to
start
to
look
for
ways
to
do
orchestration.
B
So
we're
going
to
talk
about
one
of
those
today
with
docker
compose
and
using
podman,
and
we
may
be
doing
the
docker
compose
as
a
as
a
group
here
today
and
we'll
see
how
that
goes.
I
do
know
that
I
got
it
all
running
in
podman,
so
so
at
least
we
know
that
variable
is
taken
care
of,
but
the
rest
of
it.
B
So
I
did
want
to
mention
a
couple
of
the
upcoming
things.
So
next
episode
we
are
going
to
be
doing
a
ubi
deep
dive,
so
a
lot
of
people
are
really
interested
in
how
exactly
ubi
is
put
together.
So
we've
invited
scott
mccarty
to
come
on
the
show-
and
you
know
I
scheduled
it
months
ago,
so
I
need
to
confirm
that
he's
going
to
be
able
to
make
it.
B
But
in
theory
we
should
be
talking
to
him
next
week
about
the
ubi
and
how
that
works,
and
maybe
the
future
of
ubi
like
as
in
what
do
we
have
planned
for
it
coming
up
and
then
the
week
after
that,
yes
father
linux,
jp
date
on
on
the
twitters
and
then
the
week
after
that,
we're
actually
going
to
do
a
big
special
show,
with
our
cto
at
red
hat,
who
I
just
totally
blanked
on
this
name.
Chris
wright.
B
Yeah
also
known
as
cdubs
on
the
on
the
interwebs
and
or
colonel
cdub,
I
think,
is
his
twitter
handle
so
because
he
was
a
kernel
guy
for
a
long
time
and
also
primarily
in
the
kind
of
networking
space
and
then
our
other
guest
with
kel
with
chris
is.
I
cannot
talk
this
morning,
kelsey
hightower
of
currently
of
google,
but
well
known
in
the
kubernetes
community,
and
they
so
they
did
a
show
together.
B
That
is
going
to
be
appearing
on
one
of
our
channels
in
a
few
weeks
time,
basically
releasing
like
at
exactly
the
same
time
as
the
show,
but
it's
a
really
really
short
cut
down.
You
know
quick
grab
bag
of
content
around
kubernetes
and
both
of
them
expressed
an
interest
in
kind
of
expanding
about
that
concept,
so
we're
inviting
them
both
to
the
show
to
talk
about
that
and
also
to
talk
about
upcoming
summit,
which
will
be
the
week
after
nick.
B
B
A
B
In
in
reference
to
their
backstory,
so
you
know
watch
and
this
kind
of
is
a
nice
segue
into
you
know,
watch
us
on
twitter
watch
the
open
shift
handle
as
well,
because
we'll
be
kind
of
sending
out
the
the
announcements
to
kind
of
say,
hey,
you
know
we're
we're
doing
the
show,
and
this
is
what's
going
to
be-
to
get
a
little
bit
more
on
the
abstract.
But
what
I
really
want
to
talk
about
around
kubernetes
is
also
around
kind
of
like.
B
B
A
B
B
Let's
see,
I
have
no
recollection
as
to
what
the
next
slide
is,
because
it's
one
of
those
mornings,
oh
yeah,
it's
points
and
then
no,
that's
one
after
that's
what
I
always
think
the
next
one's
going
to
be
points,
but
it's
actually
this
one,
which
is
not
very
informative.
But
today's
episode
is
about
podman,
v3,
potentially
docker
compose
and
the
cool
store,
and
then
the
show
notes
from
last
time.
B
B
List-
and
it's
even
marked
as
not
done,
I
don't
know
why
I
had
it
stuck
in
my
head-
it's
this
kind
of
neverending
problem
of
I've.
Never
if
I
can
remember
what
episode
we
are
on
right,
then
I
remember
what
the
last
one
was
and
whether
I
did
it,
and
so
I
was
totally
thinking
we
were
on
episode
29,
which
means
I
did
do
the
episode.
28
show
notes.
A
B
That's
the
challenge,
so
I
will
take
care
of
that
and
we
will
have
the
show
notes
for
next
time.
One
other
quick
thing
I
had
for
an
idea
this
morning
is,
we
know
pretty
closely
right.
We
know
a
community
organization
that
runs
an
open
shift
instance
aka
fedora,
and
I'm
not
sure
if
it's
I
need
to
find
this
out.
I
can't
remember
if
fedora
runs
one
and
the
centos
community
runs
a
separate
one
or
if
they
run
one
together.
B
B
So
like
a
production
instance,
which
is
always
hard
when
it's
your
own
software
right
and
then
you
know,
we
obviously
have
a
large
number
of
customers,
and
I
was
wondering
and
think
about
this
during
the
show.
Obviously
we
have
the
time
lag.
So
it's
hard
to
ask
and
get
answers
back
immediately,
but
what
I
was
thinking
is:
wouldn't
it
be
cool.
B
I
think
it
might
be
cool
to
do
like
a
three
episode
series
where
we
interview
the
community
group
hosting
openshift
and
then
do
another
episode,
which
is
interviewing
the
red
hat
internal
people
hosting
openshift
and
then
the
last
one
being
you
know,
we
find
a
customer
who's
hosting
openshift
who's
willing
to
come
on
the
show
and
kind
of
talk
to
all
three
of
them
about
what
it
means
to
host
openshift,
and
is
there
differences
in
those
different
environments?
So
I
thought
that
sounded
like
fun.
I
don't
know.
A
B
Oh
sure,
speed
through
it
yeah
yeah,
so
well.
First,
if
you,
if
written,
is
better,
you
know
check
out
my
twitter
handle.
I
have
a
whole
series
of
tweets
about
how
to
do
it,
but
the
the
short-
and
you
know
the
short
version-
is
post
somewhere
on
social
media,
about
something
that
you
learned
on
the
level
up
hour
then
put
a
link
to
that
posting,
and
this
is
so
that
we
don't
have
to
limit
you
to
what
kind
of
social
media
you
like
to
use
on
our
discord.
B
A
B
B
B
Interesting
all
right!
Well
so
long
story
short,
you
have
until
the
show
next
week,
not
actually
the
show
next
week,
because
I
want
to
be
able
to
announce
it
on
the
show
next
week,
but
the
night
before
the
show
next
week
so
tuesday
night,
you
know
midnight
ish
eastern,
so
that
would
be
whatever
it
is:
five
hours
off
so
about
five
a.m.
A
B
Maybe
or
I
just
conflated
chris
short
and
scott
mccarty,
because
I
was
reading
two
things
at
the
same
time-
that's
entirely
possible
as
well
all
right
so
with
all
that
aside
and
feel
free
as
to
chris's
point
right.
We
kind
of
burned
through
that
so
feel
free
to
ask
questions.
If
we
didn't
say
anything,
you
know
that
made
a
lot
of
sense,
the
okay.
So
what
I
wanted
to
do
today
so
originally
the
plan
for
the
show
was
to
talk
about
docker
compose
with
podman
and
the
cool
star.
B
B
A
B
What
I
was
wondering
actually
now
that
you
say
it,
but
I
hadn't
really
thought
about
it.
I
don't
know
where
I
got
it
from,
because
I
was
never
a
big
mac
user.
Let's
see
oh
and
by
the
way,
if
anyone
wants
to
at
least
partially
follow
along
the
stuff.
I'm
about
to
show
is
in
the
episodes
repo
on
a
branch
called
e30
initial
I'll,
open
this,
and
then
I
can
share
a
link.
A
A
A
A
B
B
B
All
right
so,
let's
see
this
yeah,
so
if
you
are
unfamiliar
with
meld,
that
is
what
I'm
sharing
right
now.
I
think
it
is
a
really
cool
piece
of
software.
It
is
open
source
and
allows
you
it's
wondering
I
don't
know.
If
I
can,
can
I
make
this
anything
bigger.
I
don't
normally
present
with
meld.
So
sorry,
if
it's
a
little
small.
B
B
A
B
B
Yeah,
I
don't
know
it
doesn't
seem
making
any
difference,
but
at
least
I
can
feel
better
about
clicking
the
button
okay.
So
what
I
wanted
to
show
was
okay,
so
originally
coolstore.
As
we
talked
about
a
few
episodes
ago,
we
talked
about
it
in
terms
of
open
shift
and
if
we
get
lucky,
maybe
I
can
show
you
a
nice
picture.
Let
me
make
sure
that
will
be
possible
in
a
few
minutes.
B
Oh
come
on,
but
the
suffice
to
say
kind
of
the
important.
Are
you
gonna
go?
The
important
part
is
it's
made
up
of
a
number
of
different.
Let
me
collapse
these
a
number
of
different
versions.
Oh
my
goodness,
I'm
so
distracted
today,
all
right,
so
that
should
be
cooking,
hopefully
soon.
Okay,
so
the
way
this
the
architecture
of
the
cool
store
works
is
it's
got
two
kind
of
data
stores,
and
one
of
them
is
your
inventory.
So,
basically
like,
oh
actually,
why
don't
I
share
yeah
share
the
repo.
B
B
Okay,
so
here's
the
cool
store,
it's
got
two
different
kind
of
data
sources.
Right
one
is
what
are
the
things
in
the
store
and
then
how
many
of
them
we
have
left?
Okay,
so
and
then
so
the
inventory
is
and
then
we
can
actually
see
because
it's
a
demo
app
right.
It's
kind
of
nicely
set
up
so
that
you
can
see
the
individual
pieces
so
in
corkus,
whoa.
B
So
the
way
I
have
it
set
up
is
I
set
up
the
back
end
in
a
podman
pod,
so
in
other
words
in
like
an
open
shift,
pod
or
kubernetes
pod.
So
now
the
back
end
is
not
addressable
directly,
which
is
exactly
what
you
want,
except
when
you're
trying
to
display
stuff.
That's
in
the
back
end,
so
let
me
find
my
terminal
window
and
we
are
going
to
okay.
So
if
you
see
I
have,
thank
you
a
whole
mess
of
containers.
B
B
B
B
B
Is
they
aren't
great
at
caching
because
they
go
and
get
a
lot
of
their
libraries
off
the
internet.
So
I
could
fix
that
by
going
and
hand
building
all
this
stuff
and
putting
it
all
local,
which
arguably
I
should
if
I
was
going
to
warn
this
in
production,
but
because,
for
the
sake
of
a
demo,
what
I
want
to
do
is
keep
it
up
to
date.
More
that
I
want
to
keep
it
running
in
a
sense
so,
but
it
does
take
a.
B
A
B
B
A
B
One
the
web
is
running
on
okay,
so
I
think
no
all
right.
Well,
let
me
show
you
what
I
was
going
to
show
you
and
then
we
can
worry
about
that
in
a
minute,
because
we're
gonna
have
to
mess
with
that
anyway.
If
we
want
to
talk
about
the
compose
stuff,
but
suffice
to
say
so,
if
we
look
at,
let's
see
catalog
is
on
80..
It
looks
like
oops
9080.
B
All
right-
and
so
here's
the
catalog
and
you
can
hit
this
nice
little
test
hit,
which
is
kind
of
cool
and
because
firefox
decides
to
render
json
nicely.
But
as
you
can
see,
it's
the
description
of
the
of
the
item
and
then
the
inventory
carries
the
the
amount
of
them
that
we
have
left.
So
I
was
going
to
kind
of
walk
through
the
individual
pieces.
So,
on
the
left
hand
side
we
have
kind
of
the
openshift
version
of
the
cool
store
and,
on
the
right
hand,
side.
B
We
have
a
slightly
hacked
like
as
in
it's,
not
a
great
version,
but
I
wanted
to
kind
of
show
the
differences
of
kind
of
running
it
in
podman.
And
what
I
want
to
kind
of
in
particular
point
out
is
that
we
could
make
it
so
that
it
would
run
in
both.
But
the
original
authors
of
the
pool
store
were
looking
to
run
in
an
open
shift.
So
that's
how
it's
it's
focused.
So
I
wanted
to
kind
of.
B
I
thought
it
would
be
useful
to
talk
about
a
few
of
the
things
that
are
why
it
breaks
down.
If
you
try
to
run
it
in
pod
man
and
so
and
you
might
be
able
to
want
to
run
it
in
both
right,
you
might
want
to
be
able
to
experiment
with
things
in
podman
and
then
post
it
to
openshift,
as
you
know,
kind
of
a
way
to
approach
it.
So,
first
up.
Oh
this
has
got
the
nice
big
font.
B
No,
I
think
it's
the
the
file
comparison
is
in
the
big
font,
but
the
file
list
is
not,
or
maybe
that's
in
a
different
setting
somewhere,
but
okay,
okay.
So
the
first
and
foremost
thing
that
we
had
to
do
was
be
more
explicit
about
where
we
were
getting
our
base.
Container
images
from
so
openshift
provides
a
whole
mess
of
container
images
to
put
your
stuff
on
top
of,
and
so
we
had
to
be
more
explicit
about.
You
know
where
that
was
happening.
B
A
B
B
You
know
bring
the
files
in
for
the
go
app
and
then
do
the
go
get
and
then
I
go
do
the
go
build,
because
if
I
don't
do
it
as
root,
I
don't
have
permissions
to
access
some
of
the
files
because
they
were
brought
in
this
route.
So
I
think
that's
kind
of
annoying.
You
know,
but
it
is
what
it
is.
So
that's
why
I
switched
to
use
a
root.
B
The
hard
part
here
is
when
you,
if
you
want
to
switch
back
to
from
the
root
user,
is
knowing
what
that
user
is.
So
you
have
to
go
investigate
your
base,
container
image
to
find
out
what
user
it's
creating
and
what
it's
using.
We
tend
to
like
red
hat
based
container
images.
The
user
is
almost
always
default
d-e-f-a-u-l-d-e-f-a-u-l-t,
but
your
mileage
may
vary.
It
may
be
different
for
different
environments.
Yes,
jp
dade
from
scratch.
Okay,.
A
A
B
That's
exposed
here,
because
one
of
the
differences
when
I'm
when
I
don't
have
like
a
service
interface
like
you
do
in
openshift,
is
I
have
to
put
everything
on
different
ports
right
because
the
service
handles
that
for
me,
but
in
like
podman,
I
have
to
put
everything
on
different
ports
because
they
can't
all
listen
on
the
same
port.
So
those
are
some
of
the
differences
here.
There
was
no
particularly
good
reason
that
I
switched
from
alpine
to
ubi,
except
that
alpine
wasn't
working
for
me.
So
I'm
not
sure.
B
Maybe
that
version
is
a
little
old
or
something,
and
I
just
and
I
couldn't
be
bothered
trying
to
find
what
the
right
one
was.
So
I
just
used
uv
i8
minimal
and
you
know,
and
it
works.
So
I
just
kind
of
walked
away.
Okay,
so
that
was
the
differences
in
the
docker
file,
but
the
thing
that
really
particularly
bothered
me
is
that
I
had
to
change
the
actual
code
base.
So
if
you
look
here,
I
change
the
the
port
number
here
as
well.
B
B
Like
only
store
the
same
data
once
right
and
it
comes
from
ruby
land
and
I'm
totally
blanking
out
what
it
stands
for.
So
if
somebody
in
the
audience
remembers
what
the
acronym
expands
to
I
gotta,
we
would
love
to
hear
it
I'll
figure
it
out,
don't
worry
there.
We
go
all
right
so
that
8090
or
8080
in
the
old
version,
don't.
B
There
you
go,
don't
repeat
yourself,
that's
it,
so
that
is
both
in
the
docker
file
and
in
this
go
code
file.
So
it
really
should
be
in
some
sort
of
configuration
so
that
you
don't
have
to
maintain
it
in
multiple
places.
And
then
we
can
ignore
yeah.
So
we'll
move
on
to
the
next
component,
because
I
think
that's
the
only
differences
yeah
and
so
that's
our
go
application
and
exactly
christian
bad
coding,
so
hugh.
B
I
didn't
write
it
so
and
it's
funny
because
you
can
tell
at
least
I
can.
I
think
I
can
tell
different
developers,
wrote
the
different
components,
because
the
style
is
different,
so
it's
kind
of
funny,
but
so
moving
on
to
our
dot.
Actually,
let's
move
to
the
inventory,
just
so
for
the
sake
of
the
architecture,
and
let
me
see
if
this
thing
is
back
up
yet.
B
Just
to
give
you
a
picture
of
what
we're
talking
about
so
just
for
the
sake
of
you
know,
visualization.
I
wanted
to
show
you
this,
but
in
podman
I'm
actually
not
using
mariadb
or
postgres,
I'm
just
using
this
in
memory
database
provided
by
java
called
h2,
so
just
kind
of
ignore
those
in
the
picture
here
I
just
kind
of
I
had
them
set
up
in
a
script,
and
I
just
let
the
script
run
instead
of
like
trying
to
clean
it
up
too
much.
B
How
to
make
this
thing
go
away,
and
then
we
have
the
inventory
for
the
cool
store,
and
then
we
have
a
gateway
in
front
of
it,
and
then
the
web
goes
to
the
gateway,
which
then
goes
to
the
catalog,
which
then
or
the
inventory.
Actually
what
it
does
is
one
api
call
here
makes
two
you
know
one
to
here
and
one
to
here,
and
then
it
combines
the
data
set
into
what
the
web
front
end
is
actually
looking
for.
A
B
Put
it
in
it
actually
doesn't
have
to
do
anything.
It
doesn't
have
to
be
there
at
all,
it's
purely
as
guidance
for
the
consumer,
so
I
can
do
exactly
so.
I
can
give
it
any
arbitrary
port.
The
the
reason
I
don't
is
the
guidance
for
the
user
problem
is
so.
Yes,
I
could
on
the
command
line,
change
the
port,
and
in
fact
I
do,
but
that's,
but
basically
that's
the
problem.
B
We're
we're
talking
about
is
that
you
know
it's
more,
the
the
guidance
to
the
user,
not
so
much
the
actual
code
that
is
required
to
you
know
be
that
way
in
a
sense,
all
right,
so
going
back
here.
Moving
on
to
the
inventory,
so
one
thing
that
happens
with
the
java
build
out
is
just
that
this
target
directory
gets
created
and
all
the
all
the
bits
when
it's
compiled
when
the
java
stuff
is
compiled,
it
lands
in
this
directory
called
target.
B
A
B
The
other
thing
I
want
to
point
out
here
is
the
differences
between
why?
Oh
sorry,
the
other
thing
is
another
annoying
factor
which
I
probably
could
fix,
but
I
wasn't
sure
how
to
originally-
and
this
is
kind
of
like
another.
One
of
those
examples
is
when
I
run
a
quarkus
app.
I
want
to
also
have
tests.
B
The
thing
is
when
I'm
running
the
corkus
app
in
podman,
I
had
a
chicken
and
egg
problem.
Basically,
I
needed
to
build
it
and
then
build
the
container
image,
and
then
I
needed
to
run
the
container
image
before
I
could
run
the
tests
because
there
was
no
other
way
to
run
it
because
I
don't
have
java
installed
or
for
the
sake
of
argument.
B
So
maybe
maybe
somebody
smarter
than
me
knows
a
better
way,
but
that
was
the
problem
I
ran
into,
and
so
that's
why
I
removed
the
tests.
You
know
generally
a
bad
idea,
but
you
know
this
is
a
consideration
to
take
into
account
when
you
want
to
make
an
application
that
is
kind
of
multi-deployable
right.
B
B
They
may
have
perfectly
good
reasons
for
not
doing
it,
but
an
h2
server
can't
be
compiled
native
it.
It
even
warned
me
it
said
you
can
do
this
with
the
client,
but
you
can't
do
it
with
the
server.
So
so
that's
why
it's
not
compiled
native
instead,
I'm
using
this
other
corkicism,
which
I
think
is
actually
a
javaism
now
as
well
or
maybe
was
first
but
called
an
uber
jar.
So
if
you
do
any
java
and
for
me
it's
been
a
really
long
time,
but
I
know
it
pretty
well,
you
know.
B
B
B
B
So
one
of
the
things
that
sysadmins
in
particular,
really
like
about
linux,
is
that
all
the
linux
is
at
least
within
their
own
distro,
follow
very
strict
rules
about
where
files
land
on
the
on
the
system
and
my
running
joke
with
an
old
friend
of
mine
who
success.
Admin
is
like
somewhat
because
at
3am,
when
you
need
to
replace
openssl,
you.
A
B
Very
blindly
be
able
to
type
you
know,
move
openssl
to
new
directory
and
know
with
your
fingers
where
that
goes
without
actually
having
to
be
fully
awake,
so
that
consistency
has
the
problem
of
what,
if
you
want
to
have
multiple
versions
of
things?
Okay,
so
two
languages
in
particular
do
a
good
job
of
solving
that
problem
in
the
language
itself,
one
is
python
and
one
is
java.
So,
however,
a
few
years
ago,
like
ten,
now,
no,
not
ten,
let's
say
seven,
eight,
seven.
B
Okay,
I'm
trying
to
remember,
because
it's
it's
slightly
less
than
I've
been
at
red
hat,
so
I
have
to
do
it,
so
I
can
do
the
math
that
way.
But
there
was
this
concept
introduced,
called
a
software
collection
which,
through
some
magic
trickery,
allowed
you
to
have
multiple
versions
of
something
not
java,
not
python,
installed
on
the
same
system
and
then
use
it
or
choose
it
by
application.
So
it's
super
nice
and
advantageous
and
is
actually
superseded
now
by
app
stream.
B
B
B
So
long
story
short
in
the
container
file.
I
could
go
try
to
set
up
the
software
collection.
You
know
basically
through
a
few
different
like
double
and
commands,
or
I
can
shove
it
off
into
a
shell
script
and
then
actually
call
the
shell
script,
which
has
bash
the
bash
commands,
are
being
run
first
then,
the
shell
script,
then
off
you
go.
B
So
that
was
a
long
story,
as
I
explained
to
chris
before
the
show
this
had
some
interesting
challenges
and
trying
to
make
this
work
with
podman.
But
I
think,
like
I
said,
hopefully,
y'all-
are
finding
this
a
good
learning
exercise
of
like
these
are
the
kinds
of
differences
in
the
real
world
that
you
might
see
between.
B
A
A
B
B
Right
so
it
does
not
tarball
only
the
things
that
you
say
add
or
copy
it
does
everything.
This
is
actually
a
big
difference.
One
of
the
things
I
really
like
about
podman
and
builda
over
docker,
so
everyone
always
focuses
on
running.
You
know
the
whatever
you
call
it
that
the
difference
of
being
podman
lets
you
do
rootless
as
well
as
not
having
a
damon
running,
but
one
of
the
things
I
really
like
is
that
I
don't
have
to
go
through
this
kind
of
machination
to
not
have
monster
build
times.
B
When
I
have
a
really
large,
you
know
website
or
something
like
that,
so
this
docker
ignore
file
conveniently
has
this
star
right
at
the
beginning,
and
then
it
says
not
on
these
few
directories
right.
So
I
spent
a
whole
bunch
of
time
and
look
and
also
looked
like
an
idiot
on
the
internet,
asking
questions
not
able
to
figure
out
why
I
could
not
get
this
run.
It
sh
to
show
up
in
my
container
turns
out,
because
this
is
the
original
and
star
covers
run
it
pretty
well.
B
B
B
Path
wrong:
you
know
something
stupid
or
whatever
you
know,
I've
run
into
problems
with
having
simulink
directories
around
and
things
like
that.
So
I
check
this
path.
The
path
looks
good
blah,
blah
blah,
you
know
and
then
I'm
like
you
literally,
are
telling
me
it's
it's
there,
but
you
decided
not
to
use
it.
So
it
took
me
a
really
long
time
to
figure
out
that.
That's
because
I
told
podman
not
to
use
it
right
to
filter
it
out.
B
A
B
One
of
the
things
that
actually
let
me
let
me
start
here-
yeah,
okay,
I
did
take
this
out.
I
thought
I
deleted
it.
Okay,
so
one
thing
that
was
both
you
know
expected,
but
also
annoying
was
just
that
the
original
author
of
this
doctor
file
relied
on
defaults
to
get
to
the
url
or
to
the
port
numbers
that
it
was
looking
for
so
by
default.
B
If
you
don't
set
this
environment
variable,
it
will
use
8080.
again
going
back
to
that,
I
want
to
run
it
on
different
ports,
oh
actually,
to
whoever's
earlier
question.
The
reason
I
do
need
to
specify
different
ports
in
here
and
not
in
the
docker
file
but
in
different
ports
is
because
to
run
it
in
a
pod.
B
That
pod
is
all
one
network
namespace
as
all
one
network
namespace.
I
cannot
change
what
port
it's
running
on
and
what
I
wanted
to
do
is
at
least
even
if
I
didn't
have
the
exact
same
files
between
openshift
and
podman.
I
could
have
the
exact
same
files
between
running
four
containers
and
running
one
pod
with
three
containers
and
then
another
container,
so
in.
If
I
want
to
run
all
those
things
in
one
pod,
they
all
have
to
be
on
different
ports.
B
B
B
B
But
if
you
notice,
even
though
I'm
root,
I'm
not
doing
a
global
npm
install
I'm
doing
a
local
one
so
that
everything
will
be
under
whatever
it
is
dot
npm,
but
it's
all
in
the
same
directory
as
the
code
base,
so
that
it'll
just
work
so
moving
on
from
that.
So
but
then
I
run
into
the
same
problem
again
as
the
with
the
software
collection,
because
node.js
is
actually
delivering
as
a
software
collection,
which
then
I
need
to
use
my
fancy.
A
B
B
A
B
Okay,
what
I
wanted
to
point
out
here
is
there's
a
configuration
here
that
it's
using
to
get
the
basically
where
that
gateway
service
is
being
hosted.
But
then
it
does
some
cheating
by
because
this
is
all
client-side.
So
it's
actually
using
the
javascript
location.host
to
find
out
what
the
server
is
and
then
it's
inserting
the
generated
by
openshift
service
name.
So
if
you
look
here.
B
B
And
because
that's
location.host
right
and
then
it's
injecting
the
name
of
the
service,
so
the
gateway
service
in
front
of
it
and
then
it's
adding
api
products
to
the
end
of
it
to
get
to
the
api
and
the
api.
If
you
see
here
is
here
and
if
you
do
the
test
it
here,
so
this
is
actually
running
in
vertex,
but
whatever
it's
the
same
code,
I
mean
it's
the
same
result.
B
So
if
you
notice
this
url
right,
so
it's
the
you
know,
location.host
with
gateway
and
stuff
injected
the
front
and
then
api
products.
So
speaking
of
not
dry,
I
was
I
was
not.
I
was
not
thrilled
with
that
so,
but
this
is
the
kind
of
thing
you
have
to
go
and
dig
through.
Is
that
now,
with
a
locally
running
set
of
containers,
you
don't
get
any
of
these
advantages
of
generated
routes,
generated
services
or
whatever
they
give
you
pointers
at
things.
All
you
get.
B
B
B
B
B
B
Believe
it
okay
so,
and
then
I
created
a
couple
of
shell
scripts
to
to
launch
it
just
to
make
my
life
a
little
easier.
Those
aren't
super
interesting
yeah.
We
have
two
different
versions,
but
so,
but
they
just
kind
of
launched
the
container.
I
thought
about
using
like
even
a
function
in
there
to
try
to
clean
that
up.
But
let's
talk
about
points
real,
quick
before
we
are
officially
out
of
time.
B
B
B
B
B
I
don't
know
if
there's
a
different
better
term
for
that,
but
basically,
if
you
watch
a
certain
number
of
episodes,
you
get
a
big
bump
in
points,
and
so,
but
you
don't
always
know
where
that
is
you
know,
if
you
do
certain
kinds
of
activities
you
get,
you
know
more
points
and
less
points.
So
if
you
go
and
enter
that
form,
then
you
will
get.
You
know.
You'll
get
your
your
initial
points
and
one
of
the
things
I
wanted
to
call
out
for
this
episode.
B
A
B
B
So
but
long
story
short
the
you
know,
I'm
really
glad
to
see
that
somebody
we
featured
as
newcomers,
you
know
are
still
with
the
show
and
are
steadily
collecting
more
points.
So
rapscallion.
We
look
forward
to
you
collecting
your
points
for
the
next
episode
and
yeah
and
right
now
they
only
have
intrinsic
value
our
goal.
A
B
For
now,
our
goal
is
to
have
them
have
extrinsic
value,
and
you
know-
and
we
are
assured
that
it
will
be
any
day
now.
We
are
mostly
just
waiting
for
t's
and
c's,
but
then
we
will
also
be
able
to
give
out-
maybe
some
whatever
you
call
it
like.
You
know,
access
to
the
red
hat
cool
store
for
some
swag
around
the
show
you
know,
etc.
So
we
we
really
look
forward
to
you
participating.
B
So
we
didn't
cover
what
we
were
supposed
to
today.
I
thought
it
was
a
really
valuable
exercise.
So
I'm
sorry
if
people
were
just
coming
here
for
compose,
but
I
will
reschedule
compose
and
we
will
do
it
in
a
future
episode.
I
promise
you
know
it.
Unfortunately,
it
won't
be
for
a
few
weeks
because,
as
I
said
before,
we
already
have
a
scheduled
guest
talking
about
ubi
scott
mccarty.
B
Right
right
just
have
everybody
watch
it
for
the
whole
team,
and
you
know
we
will.
We
will
happily
entertain.
If
you
want
we
can,
we
can
have
more.
We
can
have
just
a
set
of
slides
that
just
continuously
run
on
the
side
to
make
it
feel
like
it's
a
it's
a
meeting
with
boring,
slides,
so
narendev
was
expecting
some
live
debugging
yeah.
I
didn't.
I
didn't
quite
get
that
far.
B
That's
what
I
was
going
to
do
with
the
pose
part,
the
you
know,
but
it
does
you
know
like
I
said
I
posted
it
in
the
get
repo.
You
know
this
part
of
it
and
I
want
to
get
to
the
next
part
of
it
and
we
will
be
doing
live
debugging
for
that.
Definitely,
with
the
trying
to
make
this
work
and
compose,
I
don't
know
it's
kind
of
where
I'm
at
I
know
chris.
How
do
you
feel
what
do
you?
What
do
you
think
we
should.
B
In
compose
which
so
I
did
not
oh,
I
forgot-
I
made
a
nice
little
directory
over
here,
so
this
is
not
in
the
repo,
because
I'll
literally
show
you
this
vagrant
file
so
didn't
seem
valuable,
but
what
I
wanted
to
show
was
if
you
do-
and
I
need
to
fix
this
one
of
these
days
but
yeah
all
right.
So
if
you
do,
let
me
just
just
oh,
I
can
just
do
it
this
way.
B
B
Excuse
me
so,
in
other
words,
although
I
guess
this
has
kind
of
been
around
for
a
while,
but
docker
compose
is
going
to
look
for
a
container
engine
when
it
tries
to
install
in
order
to
ensure
that
the
container
engine
it
looks
for
is
podman.
Without
knowing
the
ins
and
outs
of
how
that
rpm
works,
I
installed
podman
and
then,
as
a
software
operation,
I
installed
docker
compose
and
if
I
had
been
smarter,
I
would
have
shown
the
basically
the
rpm
install
pod
man,
the
rapper.
Yes,.
A
B
B
B
A
B
B
B
B
That
I
was
a
little
confused
about
when
I
was
looking
into
this.
The
other
day
was
or
yesterday
or
whatever
was
one
of
the
veins
of
my
existence
or
anybody
else.
Who's
used
docker
on
at
least
fedora
and
rel
and
kind
of
its
derivatives
is
that
there's
a
there's,
a
socket
which,
because
it's
linux
right
is
this-
is
a
file
called
docker.sock
that
is
owned
incorrectly,
if
okay,
so
so
backing
up
a
little
bit
so
docker.
When
you
run
it
is
always
running
as
root.
B
You
don't
feel
like
it's
running
as
a
route,
because
normally,
when
it
tells
you
how
to
set
it
up,
it
says:
okay,
put
yourself
in
the
docker
group,
the
docker
group
has
the
ability
to
run
containers
using
the
docker
daemon,
which
is
in
fact
running
as
root.
The
thing
is
is
that
if
you're
in
the
docker
group,
the
ownership
on
var
run
docker
sock,
which
is
where
the
socket
lives
that
docker
uses
for
jpi
communication
is
owned
by
the
wrong
user.
B
B
So
the
only
way
to
do
it
is
to
run
docker
as
like
with
sudo,
which
I
don't
like
doing
partially,
because
I
didn't
realize
when
I
first
started
using
docker
that
was
always
running
as
rude
anyway,
but
now
it's
kind
of
like
habit,
but
so
what
you
had
to
do
was,
if
you
tone
the
docker
sock
to
be
owned
by
root
colon
docker
and
then
you're
in
the
docker
group.
It
works.
B
So
what
I'm
a
little
confused
about
is
pod
man,
as
far
as
I
can
tell,
is
not
putting
a
docker.sock
in
the
directory,
and
I
would
assume
that
dockercompose
is
accessing
it
through
that,
and
this
is
why
I
wonder
if
that's
why
we're
getting
this
error
here,
because
it's
a
no
such
file
right.
So
I
suspect
that
this
this
is
the
biggest
hurdle,
is
figuring
out
how
to
make
compose
realize
that
it's
using
podman
in
a
sense.
B
So
that's
the
that's
the
the
first
looking
point
my
suspicion
is:
is
that
once
we
get
over
that
hurdle,
everything
will
just
work.
It's
just
this
kind
of,
and
this
is
a
dumb
setup
thing,
not
a
not
a
important
thing
in
a
sense,
so
I
was
hoping
or
what
I
was
expecting
is
that
we
could
either
figure
that
out
during
the
show
today
or
we'll
figure
it
out
when
we
reschedule
to
do
docker.
Composes
podman
for
some
future
show.
B
A
Psps
are
going
away,
go
to
the
kubernetes
blog
to
find
out,
but
yeah.
The
the
the
concerns
that
I
had
are
about
like
folks
that
are
starting
to
use.
Psps
right
now
are
addressed
in
that
article
right
like
stay
close
to
the
core
set
of
psps
and
try
not
to
be
v8
with
too
much
wildness
and
you'll
be
okay
in
the
future,
as
the
community
develops
a
new
solution
for
psp
and
there's
also
many
alternatives
as
well,
including
opa.
A
Policy,
a
pod
security
policy
will
basically
say
this.
Software
can
only
do
this
and
this
is
connect
to
a
database
connect
to
an
outside
resource,
some
s3
bucket,
whatever
talk
to
other
stuff
inside
the
cluster,
but
basically
it's
to
prevent
people
from
designing
bad
software,
because
when
bad
software
breaks
it
breaks
in
unexpected
ways.
So
it's
targeted
at
me
is
that
what
you're
saying.
A
A
B
What
podman
v3
yeah,
so
it
does
do
a
fork
exec,
but
podman
v3,
what
it
one
of
the
things
that
it
does
so
the
way
docker
compose
and
a
lot
of
other
things
that
wrapped
up
around
the
docker
ecosystem
work.
Is
that
there's
a
rest
api
that
docker
exposes
the
docker
daemon
exposes.
That
is
then
called
by.
I
don't
know
if
the
actual
docker
client
uses
it
or
not,
but
all
these
other
things
like
docker
compose,
for
example,
bit
right,
because
docker
compose
wasn't
originally
developed
by
docker.
B
It
was
developed
by
a
third
party
who
I
can't
remember
the
name
of,
but
I
think
it
was
a
guy
who
just
built
it
for
fun,
and
so
it
calls
the
rest
api
podman
v3,
its
big
bring
to
the
table
is
support
for
that
docker
rest
api
in
the
vast
majority.
Apparently
there
are
a
few
things
in
the
in
the
api
that
it
does
not
support,
because
it's
kind
of
antithetical
to
how
podman
works.
B
A
A
B
B
A
Yeah
so
coming
up
next
on
the
channel
folks,
it's
the
one
and
only
andrew
sullivan
and
myself,
and
we
have
a
special
guest
today,
we'll
be
going
over
the
various
open
shift,
installation
methods.
You
know
we
kind
of
went
through
them
all
last
year,
but
there's
more
but
wait,
there's
more
so
stick
around
for
that.
That's
at
11
a.m:
eastern
1500
utc
and
then.
A
A
B
A
A
A
Definitely
look
into
that
sign
up,
get
your
team
to
sign
up
whatever
they
need
to
do
check
it
out
and
see
what
you
can
learn
without
any
further
ado.
I
will
sign
off
for
the
day.
Thank
you
all
for
watching.
Well,
for
the
moment,
you'll
be
for
the
moment,
not
for
the
day
for
the
for
the
for
the
illustrious
one.