►
Description
Building and Shipping Node.js Apps with Docker - Sophia Parafina, Docker, Inc.
Docker is a powerful set of open source tools built with a ‘batteries included but swappable’ philosophy. It is rapidly becoming the emerging standard for building, shipping, and running applications. For Node.js developers, Docker offers advantages that dramatically streamline the development process. In this session I'll show some of the advantages you get using Docker such as: * Getting your development environment setup on a new computer is a matter of minutes * Working with many different Node versions * Reducing dependency hell, allowing you to work with a polyglot language team without having to install. * In-container development, which allows for live debugging using your favorite IDE or editor. * Easy portability of apps to avoid lock-in
A
A
If
we
have
time
so,
first
containers
are
not
virtual
machines
and
basically
you
think
that
containers
that's
a
very
easy
connection
to
make
and
but
they're
really
fundamentally
different
architectures
and
have
very
different
benefits
and
I
guess.
The
best
analogy
that
we
have
is
that
virtual
machines
are
basically
like
single-family
houses
with
their
own
infrastructure,
one
family
inside
the
house.
That's
you
that
use
that,
as
opposed
to
containers,
which
tend
to
be
much
more
like
apartment
buildings,
where
you
have
separate
rooms
or
separate
apartments
with
a
shared
infrastructure.
A
So
what's
the
big
deal
behind
that
is
that
between
virtual
machines
and
containers,
they're
different,
but
they're,
not
mutually
exclusive.
So
on
the
let's
see
on
the
left,
hands
on
the
right
hand,
side
actually
is
at
one
type
of
application
where
you
have
an
operating
system,
a
docker
engine,
and
you
have
a
number
of
containers
running.
On
top
of
that.
However,
you
can
also
have
another
alternative
architecture.
A
We
have
physical
server
running
hypervisor
running
any
number
of
operating
systems
that
our
virtual
machines
with
docker
engine
in
that
with
docker
engine
running
on
them,
and
then
you
can
run
multiple
applications.
So
you
just
get
a
higher
density
for
the
amount
of
whether
your
server
for
applique,
so
this
is
actually
I've
use
this
before,
and
this
is
I
couldn't
find
the
actual
tweet
for
it.
But
I
think
somebody
showed
it
yesterday.
A
So
typically
people
say
you
know
now
what
when
they
come
to
docker
and
they
put
something
in
a
container
so
try
to
answer
that
and
what
what
we
like
to
call
a
docker
is
a
build
ship
and
run
philosophy.
So
our
idea
is
to
build
and
ship
and
run
that
app
anywhere,
so
you
can
basically
run
from
any
applicate
starting
at
the
very
bottom
layer.
A
This
actually
looks
a
lot
like
my
common
work
week
of
things
that
I
touch
in
every
given
day
and
starting
off
all
the
way
from
development
cycle
from
dev
and
also
pushing
all
the
way
the
ops,
so
we're
gonna
go
through
a
whole
example
and
demo
we're
going
to
do
this
so
to
start
off
we'll
do
a
little
bit
of
docker
vocabulary.
Apologies
for
people
who
already
familiar
but
docker
image
is
just
the
basis
of
container.
It's
represents
just
that
whole
application.
A
The
container
itself
is
the
image
that
is
running
inside
docker
engine,
which
is
the
next
thing,
which
is
what
runs
the
docker
containers
and
it's
deployed
on
a
physical,
virtual
or
local
host
and
finally,
there's
a
registry
service
where
we
actually
put
the
images
where
people
can
access
them,
can
access
them
over
the
cloud
or
internet
or
whatever
or
on
your
base
on
your
local
server.
So
to
put
it
all
together
and
developers,
we
have
different
build
environments.
So
if
you
have
different
applications
running
in
different
languages,
but
we
support
this
sort
of
polyglot
development.
A
We
support
shipping
through
docker
hub
and
also
the
docker
data
center,
where
we
create
and
store
images.
We
can
also
apply
security
there
and
scan
images
ahead
of
time
and
finally,
in
IT
operations
we
can
run
deploy
and
manage
and
scale
so
we're
just
going
to
run
through
this
here
on
this
demo.
So
you
can
try
this
yourself
running
docker
and
let's
go
had
put
this
down
and
you.
A
We're
going
to
go
take
a
hypothetical
application
that
was
given
to
us.
It
was
actually
originally
a
a
basic
Python
application
and
we've
converted
over
to
no
js'
application
and
running
their
Basic
Express
a
little
bit
of
formatting
using
pug
and
running
that.
So
we've
taken
that
and
ported
over
and
if
you
excuse
me,
you're
not
going
to
watch
me
type
a
lot
of
this
stuff.
So.
A
A
A
A
And
so
what
we've
done
here
in
terms
of
debugging
we've
sexually
set
the
run
and
we've
told
it
the
port's
it's
going
to
be
open,
and
then
we
also
set
a
volume
which
is
the
local
working
directory.
That
I've
got
the
application
in
and
it's
mapped
directly
to
a
directory
inside
the
docker
image.
Okay,
so
and
then
here's
the
doctor
image,
that's
running
currently
so
go
ahead
and
change
that
out.
We've
change
it
out
to
mega
Corp.
A
A
A
A
A
A
What
we've
done
through
this
we're
going
to
skip
over
all
this
stuff
here
and
the
run
portion?
So
what
we're
going
to
do
is
set
up,
what's
known
as
docker
swarm,
which
is
a
set
of
containers
and
swarm
mode
and
we're
also
going
to
set
up
with
visualizer
here.
This
is
I've
run
this
earlier,
because
it
does
take
a
while
to
run.
A
My
god,
so
we
have
a
visualizer
that
we've
set
up
so
in
we
have
a
lab
repo,
where
we
have
a
lot
of
tutorials
and
stuff,
and
you
can
start
set
up
the
stock
or
swarm.
It's
basically
just
a
shell
script,
which
sets
up
which
uses
a
Oracle's
VirtualBox
and
sets
up
a
number
of
managers
and
workers
as
a
swarm
and
one
of
the
things
we've
run
in
there
is
a
visualizer
that
shows
you.
The
different
managers
that
are
running
at
any
given
time
see.
A
A
Visualizer
is,
you
can
see,
we've
been
said,
we've
setup
at
least
ten
instances
running
here
and
if
you
switch
it
over
to
application
itself
and
it
will
run
a
little
bit
faster,
a
little
bit
more
responsive
with
number
Singh.
So
this
is
our.
What
we
do
with
docker
is
part
of
our
application
of
a
build
ship
and
run
so
swarm
itself,
like
I
mentioned
earlier,
is
running
on
an
Oracle
VirtualBox
and
there's
a
number
of
managers
and
workers
running
in
the
background.
Visualizer
is
a
little
tool.
A
We
wrote
to
see
the
things
that
are
running
on
inside
a
swarm
itself,
so
we
ran
this
actually
on
the
application
that
we
created
and
we're,
and
it
and
scaled
it
to
ten
instances.
So
why
would
you
do
this
in
docker
and
the
sort
of
the
salesy
part
here
is
that
from
some
surveys
that
we
taking
like
most
of
that,
most
of
the
things
that
people
are
doing
are
making
you're.
A
Modernizing
your
app
and
docker
is
pretty
central
to
a
lot
of
the
cloud
strategies
and
melodic
organizations
are
looking
to
adopt.
Devops
is
a
method
of
running
their
business,
so
one
of
the
things
that
docker
does
is
develop
speed,
so
it
increases
the
speed
and
velocity
of
delivering
software
releases.
It's
extremely
portable.
You
can
use
it
at
grass
spawn
node,
any
polyglot
development
environment.
You
know
c-sharp
whatever
in
Windows
and
also
you
also
have
much
greater
control
and
cost
and
reduction
in
maintaining
existing
applications.
A
Since
all
these
images
are
basically
mutable,
they
don't
change,
and
so
you
get
a
very
controlled
environment
for
people
to
develop
in
and
also
ship
and
deliver
so
yeah
one
platform
and
delivers
everything
for
all
applications.
You
can
containerize
a
legacy
legacy
applications
itself.
We
actually
have
a
lift
and
shift
tool
for
windows
applications
that
you
want
to
dock
your
eyes.
We
can
transform
all
the
legacy
stuff
to
micro
services,
you
break
down
models
into
micro
services
and
we
can
accelerate
all
the
new
applications,
development,
new
applications,
basically
greenfield
type
stuff,
so
containers
in
production.
A
You
can
use
docker
datacenter,
which
is
our
commercial
product
of
docker
hub.
So
you
have
enterprise
container
management
orchestration.
It's
currently
available
for
Linux
and
should
be
available
for
Windows
environments.
Soon
in
beta,
so
thanks
a
lot
thanks
for
putting
up
with
all
these
glitches.
So
if
you
have
any
questions
happy
to
take
them,
we
have
one
minute.
A
A
Yes,
stalkers
swarm
is
very
so
it's
similar
to
both
kubernetes
and
maysa
in
terms
of
it
is
Dockers
orchestration
platform
for
containers,
but
you
can
run
it
any
in
different
ways.
There's
actually
there's
a
if
you
don't
want
to
install
docker
itself,
there's
a
website
that
was
put
out
called
play
with
docker,
just
google
that
up
and
you
can
run
docker
commands
and
do
a
lot
of
the
docker
exercise.
We
have
in
labs
there
without
actually
having
to
install
docker,
which
is
really
nice.
Actually,
we
have
about
four
more
minutes.