►
From YouTube: Deploying IPFS Infrastructure - IPFS Camp Workshop
Description
Take your IPFS game to the next level by learning how to deploy and manage your very own IPFS node in the cloud.
https://github.com/ipfs/camp/tree/master/CORE_AND_ELECTIVE_COURSES/ELECTIVE_COURSE_E
A
All
ready,
hey,
everybody,
I,
think
we're
gonna
start
now,
thanks
for
being
patient,
our
talk
kind
of
requires
you
to
be
here
from
the
start,
so
we
wanted
to
make
sure
everybody
was
in
here
before
we
got
started
the
talk
it
today
is
going
over
how
to
deploy
your
own
IP,
FS
infrastructure
or
another
word
for
this
to
be
kind
of
how
to
set
up
your
own
node
in
the
cloud
or
how
to
be
our
competitor.
It's
kind
of
this
will
be
your
step
by
step
guide.
A
So
for
the
talk
today
we
have,
we
did
have
three.
We
got.
We
got
myself.
My
name
is
Matt
I'm
from
pinata,
and
we
have
Michael
here
who
is
from
protocol
labs
he'll
be
hopping
in
later,
on
kind
of
when
we
do
a
little
bit
of
Q&A
type
work.
He
works
on
the
infrastructural
side,
so
should
have
some
knowledge
that
you
guys
might
find
useful
at
the
end.
A
Before
we
start
have
you
guys
loaded
up
the
course
github
page
I'd
recommend
doing
that
right
now,
there's
going
to
be
a
lot
of
copy
and
pasting
to
make
this
a
easy
process
for
you
guys.
So
you
know
if
the
type
all
the
little
words
and
letters
that
we
have
on
screen
is
that
can
be
kind
of
annoying.
So
for
those
of
you
that
don't
know
how
to
get
there,
we
can
go
to
just
search
ipfs,
camp
and
Google.
A
You
should
see
the
github
link
there
and
the
Google
results
should
show
up
pretty
quickly
and
then
wait
a
little
bit
so
yeah
I
just
just
go
to
the
the
ipfs
camp
github
page
and
then
once
you're
there
check
out
core
and
elective
courses
and
then
you'll
notice
that
there
is
one,
for
course,
e
once
you've
clicked
on
that,
you
should
see
a
page
that
looks
like
this.
If
you're
there
doc,
congratulations,
you've
made
it
to
the
instructions
for
this
course
that
should
help
you.
How
quite
a
bit
I'll
give
it
a
little
bit
here.
B
B
A
Do
you
wanna
go
get
Cuba,
yeah,
okay,
I,
guess
to
start
off,
while
we're
still
here,
if
you
guys
all,
have
phones
right
now,
if
you
could
turn
the
Wi-Fi
off
on
those
that'll
help
us
a
lot
just
to
kind
of
minimize
the
bandwidth
that
we're
using
on
our
laptops.
The
majority
of
this
course
you
should
not
need
a
lot
of
bandwidth
as
we're
gonna
just
be
doing
the
most
of
this
through
SSH,
which
is
very,
very
minimal
and
bandwidth,
so
shouldn't
be
too
crazy.
A
A
Okay,
I'll
talk
a
little
bit
here.
Hopefully
we
can
get
that
loaded
anyway
through
so
in
ipfs.
There's
two
important
types
of
nodes
that
you
should
all
be
aware
of.
We
have
server
nodes
and
we
have
client
nodes
and
they
perform
a
little
bit
of
a
different
role
in
the
ecosystem
and
the
type
of
kind
of
settings
that
they're
going
to
need.
So
server
nodes
are
going
to
be
always
online
they're,
typically
going
to
be
hosted
in
data
centers
cloud
environments,
they're
focused
on
providing
content
to
consumers
that
want
to
retrieve
it.
A
So
there's
a
lot
of
cloud
providers
out
there.
We
have
digitalocean
Google
cloud,
Amazon
Azure,
all
of
them
pretty
much,
provide
the
same
functionality
and
kind
of
offerings.
Mostly
we're
gonna
provide
that
virtual
server
that
you
can
just
spin
up
and
every
one
of
them
should
allow
you
to
spin
up
in
a
bun,
2
or
Linux
instance,
which
is
what
you
should
be
hosting
your
cloud
infrastructure
on
most
of
the
time
just
runs
better.
A
So
for
this
we're
going
to
be
using
digitalocean,
that's
what
we
pinata
used
to
host
our
infrastructure.
We've
been
working
with
our
team
quite
a
bit
to
I
kind
of
get
nodes
spun
up
and
make
sure
that
everything's
working
right
I
had
everybody
shut
off
their
phones.
Have
you
guys
noticed
any
sort
of
better?
Ok,.
C
A
A
So
to
start
off
with
some
of
you
may
already
have
a
digitalocean
account
if
you
do
no
worries
just
kind
of
tie,
it
we'll
get
to
what
you
do
here
in
a
minute
for
signing
up.
We
do
have
a
code
that
you
guys
can
use
yet
I
think
it's
50
dollars
in
free
credits.
That
should
be
more
than
enough
for
the
purpose
of
this
course
and
then,
when
you're
done
no
worries,
you
won't
get
charged
just
make
sure
to
take
down
the
node
that
you
set
up
otherwise,
then
yeah.
A
They
will
eventually
charge
you
because
you're
using
them
to
house
stuff,
but
your
credit
should
actually
you
don't
get
charged
for
awhile.
So
to
start
off
with,
and
once
we
have,
our
account
signed
up
we'll
go
to.
Basically,
the
confirmation
page
I'll
give
it
a
pause
here
for
I'll
go
back
to
the
sign
up
just
so
you
guys
have
that.
A
Let
me
know
when
you've
all
signed
up
and
then
we
can
go
to
kind
of
continue
on
from
there.
The
signup
process
is
probably
going
to
be
the
most
like
lengthy
process
of
this
whole
event.
So
I'll,
let
you
guys
kind
of
go
through
that
and
and
if
you
don't
feel
comfortable,
signing
up,
feel
free
to
just
kind
of
watch
a
partner,
and
hopefully
this
will
still
be
an
informative
course.
Most
of
what
we're
talking
about
here
today
can
be
applied
to
any
other
cloud
providers.
A
A
A
Yep
yep
you'll
just
need
an
Ubuntu
machine,
that'll,
be
it
but
yeah
we're
all
we're
just
going
through
the
signup
process
right
now,
I'm
trying
to
take
it
a
little
slow
just
because
this
will
be
the
most
annoying
part
of
the
course
so
I'm
assuming
most
of
you
have
gotten
there.
Now
that
are
interested,
we'll
go
to
the
billing
page
enter
in
your
credit
card
stuff.
You
won't
be
charged
and
you
can
destroy
the
note
at
the
end
of
the
course
and
basically
get
rid
of
your
account
and
all
will
be
well.
A
A
A
Okay,
perfect,
alright
I
will
move
on
then,
okay,
so
once
we've
gotten
that
entered
in
this
is
less
important,
but
just
entering
a
project
name
and
just
kind
of
pick,
some
random
stuff.
It
does
not
really
matter.
It's
I
think
it's
just
so.
I
have
an
idea
of
what
kind
of
stuff
they
should
be
prioritizing
for
their
platform.
A
Yeah,
don't
worry
about
it
too
much
just
kind
of
throw
whatever
there
and
then
just
click
start.
So
once
we're
there
we're
going
to
the
fun
part,
which
is
creating
our
own
virtual
server
on
digitalocean
they're
called
droplets
droplets
of
water.
You
know
this.
We
all
like
puns
in
this
space,
so
yeah
you
know
for
droplets
to
create
a
droplet,
we're
going
to
start
off,
and
basically
just
click
on
that
that
little
thing
that
says
new,
droplet
and
that'll.
A
Take
you
to
the
droplet
creation
page
and
yeah,
we'll
go
from
there
all
right,
so
don't
keep
pausing.
If
anybody
wants
me
to
slow
down
or
pause,
feel
free
to
raise
your
hand,
it
might
be
a
more
kind
of
easier
way
to
make
sure
everybody's
following
along
in
the
same
pace,
don't
want
people
to
be
left
behind
so
once
we're
there.
A
D
A
Are
you
it
does
this
look
where
you're
at
like
right
here?
Are
you
like
this.
A
A
So
yeah
once
we're
there,
we'll
click
create
new
droplet
and
then
we're
going
to
choose
Ubuntu
there's
a
variety
of
different
nodes
specifications.
You
can
choose
from
the
$15
a
month.
One
is
a
nice
kind
of
it
gives
you
good
specs
if
you're
just
running
some
hobbyist
type
work.
If
you
start
to
get
into
more
enterprise
level,
projects,
you'll
probably
want
to
increase
the
size
of
your
node,
but
yeah
for
right
now
eat
it.
A
You
don't
need
too
much,
and
since
we
have
some
free
credits
might
as
well
use
them
make
things
fast
for
this
course,
so
you
can
enable
backups
if
you
want
they're
nice
to
have
they
cost
about
20
percent
of
the
droplet
price.
It's
up
to
you.
Block
storage
is
a
good
option
for
those
of
you
that
are
looking
for
more
flexible
data
limits.
So
that's
it's
an
option
for
when
your
your
node
specifications
are
going
to
not
really
need
to
be
increased
after
a
certain
point,
you
don't
need
more
RAM,
you
don't
need
more
cores.
A
You
just
need
more
storage,
so
block
storage
allows
you
to
have
a
little
bit
more
of
a
flexible
data
store.
It
gives
you
up
to
16
terabytes
that
you
can
utilize
for
your
machine.
That's
a
little
bit
out
of
the
scope
of
this
tutorial,
but
I'd
be
glad
to
help
any
of
you
out
this
week.
Set
that
up,
if
you,
if
you'd
like
so
just
find
me
around
here
and
I'll,
gladly
help
you
out
with
that.
But
yeah
choose
whatever
data
center.
A
You
want
Frankfurt's,
probably
gonna
be
one
of
the
fastest
ones
for
all
of
you
here
is
it's
the
closest
one
to
a
Spain
and
then
once
we've
chosen
that
we
like
to
set
up
ipv6.
Well,
that's
it's
just
another
way
that
people
can
find
your
node
and
then
setting
up
monitoring
is
nice.
These
are
all
free
options.
They
don't
cost
you
anything
and
it's
nice
to
see.
If
anything
goes
wrong
on
your
node,
you
should
see
you'll
see
you
have
RAM
spikes,
CPU
spikes.
A
Okay,
most
of
you
we're
going
to
go
through
the
SSH
setup
process,
just
to
make
sure
that
we're
all
kind
of
just
make
sure
we're
going
to
cut
out
the
variables
SSH
is
a
nice
tool
for
any
developer
working
on
remote
systems.
It
allows
you
to
securely
restrict
access
to
your
machine,
so
only
you
can
access
it,
and
some
of
you
may
already
have
a
key,
but
for
those
of
you
that
don't
let's,
let's
learn
how
to
create
one.
A
So,
like
I
said,
if
you
are
to
have
a
key,
just
wait,
we'll
we'll
get
to
how
that
that
plays
into
things.
We'll
give
you
a
little
bit
of
a
breather
for
now,
so
we're
gonna
use
OpenSSH
for
those
of
you
that
don't
have
a
key
to
create
it.
It's
a
fairly
open
standard.
This
should
work
on
your
Mac's,
your
Windows
machines,
your
Linux
machines.
Basically
all
of
them
will
have
OpenSSH
compatibility
to
do
this.
This
is
where
we're
going
to
get
into
the
the
github
page.
This
will
come
in
handy
all
of
these.
A
All
of
these
texts.
Instructions
that
you're
going
to
see
there
verbatim
what
you're
going
to
github
page,
so
you
should
be
able
to
just
copy
and
paste
things
there,
but
if
you
run
into
issues,
feel
free
to
raise
your
hand
and
we
can
go
over
those
so
open
up
your
terminal
and
we're
going
to
type
in
SSH
desk
key
gem.
This
is
going
to
generate
a
private
and
public
to
you
that
you
can
use
to
basically
authenticate
with
your
remote
server.
Make
sure
that
that
server
knows
hey
it
actually.
A
Is
you
when
you're
connecting
it
helps,
keep
intruders
out
if
you
so
important
here,
do
not
choose
to
overwrite
if
the
key
already
exists,
if
you
guys
already
have
a
key
I
should
have
said
this
first,
but
if
you
already
have
a
key,
don't
override
it,
because
you
probably
are
using
that
for
some
other
system,
and
you
don't
want
to
forget
that
key
or
else
you'll
have
to
reset
some
things
up,
which
is
not
fun.
We
can
use
the
key
that
already
exists
for
for
digitalocean.
You
can
optionally
provide
a
password
each
time.
A
You
use
your
key
and
it's
recommended
for
security
reasons,
but
if
you're
lazy
just
hit
enter
twice-
and
you
won't
need
a
password
for
your
ssh
so
now
that
we
have
an
S
as
such
SSH
key
created,
let's
copy
that
and
that's
gonna
be
what
we're
gonna
put
into
digitalocean
so
to
copy
your
ssh
key.
You
can
do
any
of
these
things.
A
lot
of
you
are
probably
familiar
with
command
line
work
so
just
type
in
this
cat
SSH
yeah
copy
that
command.
You
can
get
it
from
your
terminal.
It's
likely
that's
where
it's
stored.
A
Some
of
you
may
have
your
key
saved
in
a
different
location.
I'm,
assuming
if
you
have
your
key
saved
in
a
different
location
that
you
know
how
to
grab
it
so
travel
to
wherever
that's
at
and
then
just
copy
the
contents
of
it
yeah
so
make
sure
that's
all
copied
on
to
your
clipboard
and
then
once
you
have
that
let's
go
back
to
the
digitalocean
GUI
and
basically
add
our
SSH
key.
There
should
be
a
button.
That's
basically
says
new
SSH
key
will
click
that
will
copy
the
contents
that
we
just
created
into
this.
A
We
can
name
it
whatever
we
like
and
we'll
click
add
SSH,
key
and
that'll
be
from
this
point
on
you
can
use
this.
You
can
just
check
it,
and
digitalocean
will
save
that
that
public
signing
kind
of
key
there
and
you
will
never
have
to
add
your
SSH
key
again
for
new
droplets.
It'll
always
be
there.
You
can
certainly
add
more
keys
if
you'd
like,
but
that's
up
to
you
so
once
we've
done
that
we're
going
to
name
our
node
whatever
we
like
I
like
to
name
it,
my
amazing
ipfs
node.
A
Once
we
have
that
just
click
create.
It
should
take
a
little
bit
of
spin
up
time,
probably
30
seconds
to
a
minute
and
then
once
it's
once
it's
good.
You
should
see
something
like
this
I'll.
Take
you
back
to
your
main
page.
That'll,
show
all
of
your
servers
that
you've
spun
up
and
you'll
get
a
little
IP
address
for
your
node
that'll
be
fun
that
we
can
connect
to
it.
A
How's
everybody
doing
so
far.
If
I'll
ask
the
hand
raised
again
or
you
guys,
do
you
guys
all
have
your
node
created
being
crated?
Ok!
Well,
we'll
give
that
another
another
30
seconds
or
so
yeah.
So
digitalocean
is
nice.
You
can
I
do
a
lot
of
stuff
from
the
GUI.
It
kind
of
makes
your
life
a
little
bit
easier.
If
you
don't
want
to
live
in
the
command-line
world,
they
have
a
little
options
for
you
can
turn
your
note
on
and
off.
You
can
get
statistics
for
it
feel
free
to
play
around
with
that
afterwards.
A
That
IP
address
you
see
here
once
it's
created
is
gonna,
be
the
most
important
part,
that'll
be
how
we
connect
to
things
cool
I'm,
hoping
everybody's
note
is
created
so
far
you
know
keep
moving
so
now
that
we
have
our
node
setup.
We
have
that
IP
address
that
we
saw
let's
go
ahead
and
copy
that
you
can
do
that
pretty
easily
just
by
like
clicking
on
the
the
IP
address
that
should
just
automatically
copy
it
to
your
clipboard.
A
Once
we
have
that
we're
going
to
use
that
SSH
key
that
we
just
created
and
we're
going
to
SSH
into
the
root
of
our
droplet
kinda,
like
the
admin
user
here,
so
you
should
need
to
type
yes,
if
you're
a
cessation
into
your
node
for
the
first
time,
I'm
just
kinda
hit
enter,
and
once
that's
done.
If
you
you
should
be
connected
to
your
droplet,
you
may
need
to
enter
a
password.
If
you
did
do
that
for
your
SSH
key,
it
may
prompt
you
for
that.
A
If
you
guys
run
into
issues
here,
let
me
know
I
can
I
can
provide
a
spare
link.
Some
to
that
did
you
guys
fix
the
ipfs
kind
of
distribution,
repo,
the
binaries?
Okay,
if
you
guys
run
into
issues,
let
me
know
it
can
give
you
an
alternative
length.
We
host
our
own
versions
of
the
the
binaries
just
to
make
sure
it's.
We
have
access
so
in
your
terminal
type
out
apt
get
update
once
you've
done
that
type
apt
get
upgrade.
A
This
is
kind
of
upgrading
your
node
and
all
the
packages
on
it
make
sure
everything's
up
to
date.
If
you
get
a
pink
screen
asking
whether
or
not
to
automatically
kind
of
do
some
stuff
hit
the
left,
arrow
and
then
hit
enter,
this
is
basically
a
little
annoyance
that
you
got
to
get
through
I'm.
Sorry,
sometimes
you
hit
it.
Sometimes
you
don't
and
then
once
we're
done
with
that,
you
may
get
another
pink
screen
hit
enter
again,
we're
not
that
you
can
auto
remove
the
kind
of
installation
stuff.
A
That's
up
to
you,
you
don't
really
need
to,
but
it's
nice
to
clean
it
up,
and
then
once
we've
done
all
this,
we
can
shut
down
our
node.
Let's
restart
it
make
sure
all
of
our
changes
have
been
made.
Your
droplet
should
now
be
updated
and
rebooted
fun.
Okay,
so
we
don't
worry
about
kind
of
just
let
it
do.
Its
thing
should
just
should
kind
of
do
its
stuff
in
the
background,
so
we
have
an
optional
step
here,
which
is
created
an
on
route
user.
A
This
is
something
I'd
recommend
for
those
of
you
that
are
serious
about
proving
making
this
stuff
in
production.
It's
good
practice
to
have
an
on
where
user,
so
you
don't
accidentally
just
destroy
some
things
on
your
node
due
to
a
typo,
that's
kind
of
outside
the
scope
of
this
course.
If
you'll
notice
that,
at
the
end
of
the
the
github
page,
there
is
a
link
that
talks
a
little
bit
more
in
depth,
it's
a
written
guide
that
I
have,
which
basically
goes
over
a
lot
of
the
things
you're
seeing
in
this
course.
A
So
we've
created
our
droplet,
but
anybody
in
the
world
right
now
can
connect
to
it.
We
have
no
firewalls,
we
have
no,
we
have
no
real
security
around
it
outside
of
SSH.
Ssh
allows
us
to
kind
of
control
the
node,
but
anybody
can
hit
any
port
on
our
machine
right
now
and
we
don't
want
that.
We
want
to
make
sure
people
are
only
connected
on
the
port's
that
we
want
to.
So
let's
do
some
network
insecurity
on
the
main
digitalocean
page.
Let's
click
on
network
in
and
once
you
are
there.
A
Good
awesome:
okay,
okay,
so
these
are
going
to
be
the
ports
that
you're
going
to
want
to
enable
on
your
ipfs
node.
For
those
of
you,
they're
unaware,
ipfs
communicates
with
the
rest
of
the
ipfs
network
using
port
4001.
That's
a
default
setting!
You
can
change
this
if
you
guys,
for
whatever
reason,
have
odd
firewall
settings
or
requirements,
you
can
change
that
port.
It
is
something
that
people
do,
but
it
is
4001
by
default.
A
It
allows
users
to
go
and
type
in
the
little
URL
bar
at
the
top
of
their
browser
and
retrieve
content
from
the
network
without
needing
to
spin
up
their
node,
so
I'd
recommend
enabling
it
we'll
go
over
at
the
end
of
the
course
kind
of
how
to
enable
that
gateway.
But
so
we
don't
have
to
go
back
and
set
up
the
firewalls.
Just
I
didn't
able
it
for
now
in
terms
of
outbound
rules,
just
enable
everything
it
doesn't
really
matter.
A
A
So
installing
ipfs
fairly
straightforward
just
copy.
These
commands
that
you
see
here.
First
we're
going
to
grab
the
ipfs
update
binary.
So
ipfs
update
is
a
really
nice
tool
that
the
protocol
ops
team
has
put
together.
That
allows
you
to
basically
continually
update
your
IP
FS
node
from
the
command
line
without
having
to
manually
pull
new
branches
every
time.
Ipfs
has
an
update
or
a
new
release.
A
It's
how
we
prefer
to
update
our
infrastructure.
I'd
recommend
using
it
yourself.
It
also
helps
with
migrations.
So
if
there's
ever
like
a
crazy
breaking
change
between
versions,
ipfs
update
has
automatically
handled
those
data
store,
migrations.
I
think
it
will
continue
to
do
that
in
the
future.
So
yeah
again
highly
recommend
using
IP
FS
update,
so
once
you've
downloaded,
it
will
unzip
it,
and
then
we
will
go
into
the
ipfs
folder
using
the
CD
command.
A
Once
we
have
ipfs
update,
downloaded,
there's
a
handy
little
install
script
that
they
provide
as
well.
You
made
it
pretty
pretty
straightforward
for
us,
so
just
hit
install
dot
our
type,
the
command
you
see
there
installed
on
Sh,
that'll
run
the
shell
script
and
yeah
ipfs
should
be
a
PFS
update
should
be
installed
on
your
computer.
A
A
A
A
A
Ok,
so
sorry
I'm
trying
to
try
to
poke
some
fun
in
the
command
line.
Stuff
I
know
it's
staring
at
text:
it's
not
the
most
exciting,
but
it's
stable.
Alright,
so
you'll
notice
there
on
your
on
your
command
line,
you'll
see
you
should
see
the
the
IP
addresses
that
your
swarm
is
listening
on.
You
should
see
a
local
IP
address.
You
should
see
your
public
IP
addresses
and
the
ports
they're
listing
on
they
transport
their
listing
on
this
is
all
kind
of
lid.
A
A
We
don't
want
I
P
FS
to
shut
down
when
we
close
our
our
SSH
instance.
So,
let's
learn
how
to
keep
that
running
and
make
sure
things
actually
do
stay
online,
so
it
differentiates
the
server's
from
the
laptops
they
stay
online
when
we
shut
them
down
so
to
keep
ipfs
running,
we're
going
to
create
a
system
service
system
services.
A
A
You
can
use
vim
if
you
like,
I
will
refrain
from
passing
my
judgment
on
you
I
like
Nano,
it's
I
won't
get
I
poke
fun,
but
yassuh
can
use
nano
vim
whatever
you
want,
just
create
ipfs
service
in
the
system,
D
slash
system
directory
again
this
is
all
on
github.
So
copy
and
pasting
is
your
friend
here
to
make
sure
that
this
is
all
going
to
work
stable.
It
once
we're
there
we're
gonna
copy
and
paste
the
following
instructions
into
our
system
service.
A
The
you
had
a
little
description,
the
main
thing
you're
going
to
notice.
There
is
that
this
exact
start
command.
This
is
basically
saying:
hey
here's
the
command
that
we're
going
to
run
to
start
the
daemon
every
time
we
start
this
service.
We're
gonna
run
this
command.
It's
the
same
thing
as
running
it
in
the
command
line,
we're
gonna.
What
it's
doing
here
is
pointing
to
where
I
PFS
that
binary
is
installed,
and
then
it's
just
running
the
daemon.
A
A
So
this
means
is
that,
like
painting,
operations
are
unable
to
be
used,
while
your
notice
garbage
collect
in.
So
if
you
get
to
production
type
scale,
that's
something
to
be
aware
of
from
what
I
understand
garbage
collection
reference
counting
is
something
that
ipfs
team
is
working
on.
So
maybe
by
the
time
you
at
that
point,
it
won't
be
a
problem,
but
for
now
ipfs
cluster
is
your
friend.
A
You
can
basically
maintain
multiple
nodes,
so
you
can
take
one
down
run
garbage
collection
on
it
while
pinning
to
the
other
one
and
then
they'll
synch
up
later
on.
That
gives
a
do
a
little
more
of
an
advanced
topic,
though
so
talk
to
Hector
or
check
out
ipfs
cluster.
If
you
guys
are
interested
in
doing
something
like
that.
Yes,
this
guy
so
know
for
that
red.
We
have
the
restart
variable
here.
This
is
what
we're
gonna
make
sure
you
guys
have
this.
A
This
basically
says
hey
if
ipfs
crashes,
for
whatever
reason,
let's
screw
started,
if
you
shut
off
your
node
for
whatever
reason,
if
you
want
to
take
down
your
droplet,
bring
it
back
online,
no
worries
you
don't
have
to
manually
restart
all
this.
It
starts
up
when
your
machine
starts
up,
so
your
node
will
always
be
online
and
then
the
environment
is
your.
I
PFS
path.
So
what
I've
put
here
this
is
the
default
for
what
we
have
installed
right
now
specific
to
this
course.
A
Multi-User
target,
that's
less
important
to
you
guys
but
yeah,
once
we're
done
with
all
this
hit,
control,
X
and
then
Y
for
yes
to
confirm
things
and
then
hit
enter
your
system
should
be
created.
So,
let's
enable
this
service,
we
just
created
we'll
reload
our
system
with
using
a
system
CTL.
This
is
kind
of
what
you
use
to
control.
All
these
system,
services,
you've
just
created.
First
we'll
enable
our
service
by
daemon
reload.
Then
we
will
enable
the
ipfs
service
that
we
specifically
just
created.
A
A
A
A
A
It
may
still
work
without
them
all
being
the
same
version,
but
for
your
sanity,
it's
highly
recommend
to
upgrade
them
all.
At
the
same
time,
yeah,
that's
yeah,
you
cannot
forget,
you
can
upgrade
them
independently,
but
just
be
aware
that
it
helps
to
upgrade
them
all
at
the
same
time,
if
you're
running
something
like
cluster
anything
else,
that
anybody
has
any
questions
on
cool,
let's
try
pinning
something
I'm,
assuming
many
of
you
have
done
something
like
this
before.
A
So,
let's
try
it
out.
I'm
assuming
you,
our
are
all
still
connected
to
your
your
droplet
in
the
cloud,
if
you're
still
seeing
that
kind
of
like
status
that
you've
had
from
earlier.
Just
hit
control-c
and
that'll
get
out
of
that
little
view
that
you
have
there
I,
allow
you
to
type
stuff
into
the
terminal
again.
So
in
your
terminal,
let's
add
a
hash.
A
This
is
just
a
hash
that
has
fairly
good
availability
on
that
ipfs
network.
We've
put
in
a
lot
of
blog
posts
and
have
it
pinned
in
a
lot
of
nodes.
So
hopefully
your
nodes
should
be
able
to
find
it.
If
not
I
blame
them.
Yes,
but
yeah
it's
it
should
be,
it
should
be
relatively
available
and
then
the
prot,
the
progress
command
or
the
are
there.
What
you're
actually
paying
is
a
website
that
was
uploaded
ipfs
in
directory
format.
A
So
what
we're
recursively
pinning
all
the
sub
sub
files
there
and
sub
sub
hashes
and
then
the
test
progress
is
really
nice.
You
can
kind
of
see
and
make
sure
that
things
are
actually
happening.
Hopefully,
things
have
succeeded,
if
not
I'm
very
sorry,
just
can
I
get
a
show
of
hands.
Have
you
guys
had
much
success
of
trying
to
pin
that
hash,
perfect,
so
yeah,
that's
nice!
Dht
is
working
exactly
exactly
so
now
that
we've
pinned
this.
Let's,
let's
try
that
gateway
that
we
talked
about
earlier.
A
We
enabled
port
8080
on
our
machine,
so
this
means
that
the
outside
world
can
hit
port
8080
on
our
machine
and
view
our
gateway,
but
the
default
IP
FS
config
does
not
make
that
port
available
to
the
world.
It's
assuming
that
you're
gonna
set
something
like
nginx
up
in
front
of
your
gateway,
which
recommended
most
people
don't
want
to
go
to
your
your
droplets.
Ip
address
slash
ipfs,
let's
loss
that
hash.
A
Most
people
want
to
go
to
something
like
my
node
comm,
slash,
ipfs,
slash
that
hash
just
a
little
more
user
friendly,
so
yeah,
let's
set
up
the
the
config
to
allow
for
this.
What
we're
gonna
do
here
is
we're
going
to
run
from
the
command
line,
I,
PFS,
config
and
then
we're
going
to
set
up
the
gateway
address.
A
There's
there's
nice
little
kind
of
CLI
tools.
If
you
know
the,
if
you
know
the
setting
you're
looking
to
change,
you
can
do
it
from
the
command
line
or
you
can
edit
the
configuration
file
itself.
That
gets
a
little
bit
more
manual
process.
So
what
we're
setting
it
here
to
is
just
the
localhost
version,
that's
available
to
the
world
and
make
port
8080
available,
and
once
we've
done
that
I
pee
fast,
when
you
make
changes
to
it,
you're
going
to
need
to
reset
your
daemon
or
else
those
changes
won't
take
effect.
A
So
we
need
to
restart
the
daemon,
make
sure
everything
boots
up
again.
Then
things
should
work
so
once
we've
done
that,
go
to
your
browser
and,
let's
type
in
your
droplets,
IP
address
colon
8080
/fs,
slash
that
hash
that
your
droplet
IP
should
be
the
only
thing
that
you
really
need
to
change
there.
You
should
just
be
able
to
copy
and
paste
it
from
digitalocean.
If
you
see
a
picture
of
a
little
spinning
pinata
you,
you
succeeded.
Congratulations,
you
have
a
working
gateway
on
your
machine
yeah.
That
was
a
fun
little
demo
website.
A
It's
a
react
application.
If
you
guys
are
interested
in
hosting
your
own
websites.
We
have
a
few
blogs
out
on
that
as
well.
It
should
kind
of
take
you
through
that
process,
so
the
important
parts,
how
do
not
get
your
credit
card
billed
we've
done
some
fun
stuff
in
this
course.
If
you
want
to
keep
your
droplet
online
for
testing
purposes,
you're
more
than
welcome
to,
but
I
would
feel
bad.
If
I
didn't
mention
this,
and
then
you
guys
forgot
about
it
and
your
credit
card
got
charged
so.
C
A
So
learn
from
Michaels
mistakes,
unless
you
want
to
keep
your
note
up
in
which
you're
awesome
because
you're
helping
the
ipfs
network
yeah.
So
let's
go
through
that
real
quick.
What
next
a
little
bit
of
a
choose-your-own-adventure
at
this
point,
we're
going
to
show
of
hands
missing
kind
of
what
you
guys
want.
We
have
two
options
here:
we
can
go
over
some
tips
and
tricks
for
node
settings.
So,
based
on
your
guys's
specific
use
case,
we
can
talk
about
some
of
the
configuration
settings
that
may
be
very
helpful
for
you
in
ipfs.
A
The
other
option
is
a
little
bit
of
a
Q&A
session.
Both
Michael
and
I
are
fairly
well
versed
in
maintaining
ipfs
infrastructure,
so
we'd
be
happy
to
pass
on
any
of
our
knowledge
to
you
yeah.
So
let's
get
a
little
bit
of
a
show
of
hands
at
this
point.
For
those
of
you
interested
in
tips
and
tricks
or
like
configuration
settings
raise
your
hand.
A
Okay,
it's
about
even
okay,
all
I'll
quickly
burn
through
the
tips
and
tricks.
Then
we
can
ask
some
questions
so
for
those
of
you
that
are
running
on
local
networks
or
like
home,
Wi-Fi
settings.
There's
this
discovery
settings
there's
this
thing
called
mdns,
which
is
basically
how
you
find
local
peers
on
your
network.
It's
pretty
useful!
It's
from
one
of
that,
like
offline
first
scenario,
that
a
lot
of
people
might
be
passionate
about
for
like
IOT
type
stuff
or
you
know,
private
networks.
C
A
A
This
is
again
a
lot
of
your
like
applications
will
be
running,
DHT
nodes
or
full
DHT
nodes,
and
then
we
talked
about
clients
earlier
this
DHT
client
setting
means
that
your
client
will
only
consume
content
from
the
network.
It
won't
serve
it.
There's
a
lot
of
trial
and
error
that
we've
we've
encountered,
or
not.
Sometimes,
we've
forgotten.
When
we've
done
some
testing
stuff
like.
Why
isn't
it
loading
it's
because
one
of
the
server
nodes
is
set
to
DHT
clients?
A
This
is
really
good
for,
like
desktops,
most
people
are
not
going
to
be
providing
on
their
local
desktops
they're,
mainly
going
to
be
consuming
content
so
for
desktop
type
applications
or
phone
applications
unless
you're,
specifically
sending
content
to
other
servers,
you're,
probably
going
to
want
to
set
the
DHT
client
setting
other
tips
and
tricks.
There's
this
connection
manager
setting-
and
there
are
these
two
options
that
you
can
configure
called:
high
water
and
low
water.
A
These
are
going
to
be
the
maximum
number
of
nodes
your
server
will
connect
to
and
then
once
it
hits
that
max
setting
it's
going
to
garbage
collect
a
couple
of
those
connections
able
garbage
collect
down
to
what
the
low
water
setting
is.
So
there's
a
lot
of
different
options
you
can
use
here.
It's
really
just
kind
of
pick,
whatever
works
best
for
your
infrastructure.
A
For
us,
we
have
a
lot
of
connections,
because
the
DHT
doesn't
really
we've
kind
of
outgrown
the
DHT
as
a
provider.
So
we
need
to
hop
over
top
of
that
and
connect
to
as
many
nodes
as
possible
and
that's
how
our
content
gets
found
quickly
if
we've
relied
purely
on
the
DHT,
where
we
have
so
many
route
provider
records
that
our
node
can't
send,
it
can't
announce
it
to
the
network
fast
enough,
so
our
nodes
are
set
to
announce
things
every
12
hours.
A
A
Would
not
be
able
to
be
found,
so
if
you
guys
are
running
high
high-volume
processes
where
you're
doing
a
lot
of
a
lot
of
root
nodes
or
a
lot
of
individual
files
that
you've
pinned
and
not
just
a
bunch
of
a
few
big
files,
you're
probably
going
to
want
to
have
a
very
high
high
water
set
in
and
your
low
water
setting
I
would
set
it
to
a
couple
thousand
below
that
because
you
don't
want
to
garbage,
collect
all
the
way
down
to
like
a
thousand
cuz.
Then
you
just
gotta
have
accumulated
connections
again.
A
We
connect
to
around
30,000
per
node.
That's
just
we're
continually
increasing
that
just
because
we
don't
want
to
go
full
these
guys
yet
and
connector
like
a
hundred
thousand
peers,
but
because
we
haven't
needed
to
yet.
But
we
are
continual
increase
in
this
kind
of
testing
our
limits.
We
haven't
had
any
issues
with
it,
but
yeah
it
depends.
We
some
of
our
machines,
are
running
like
32
cores,
190
gigs
of
ram,
but
keep
keep
in
mind
there.
That's
we're
not
using
or
even
remotely
like
close
to
the
maximum
out
there.
A
Our
machines
are
probably
at
like
5%
usage
of
our
resources
at
any
given
time.
The
big,
the
big
reason
why
we
do
that
is
because
the
more
cores
you
have,
the
better
that's
more
blocks.
You
can
be
fetching
in
parallel
from
the
network
and
that's
kind
of
a
it's
a
dirty
trick
that
we're
using
right
now
until
like
single
core
parallelization
becomes
more
of
a
factor
in
ipfs
yeah
we'd
love
to
be
able
to
kind
of
max
out
our
cores.
But
we
just
can't
do
that
right
now,
so
yeah.
A
We
just
have
a
lot
of
course:
oh
yeah,
okay
yeah.
Lastly,
NAT
that
horrible
thing
that
everybody
hates
it's
what
kind
of
prevents
nodes
from
being
found
a
lot
of
the
time
it's
really
annoying
to
deal
with,
but
it's
not
impossible
so
to
basically
get
your
your
fire
past
the
firewall
situation.
Ipfs
has
some
relay
type
stuff.
Basically,
this
means
you
can
enable
if
you
have
Auto
relay
enabled
on
your
node.
A
E
A
Okay
yeah,
so
the
relay
setting
basically
means
people
can
use
your
notice
of
relay
to
transfer.
Yes,
fair,
warned
in
there.
You
are
basically
paying
for
everybody's
bandwidth
that
uses
you
as
it
realized
so
enable
that
with
caution,
and
then
the
enable
nat
service
allows
people
to
check
your
note
and
say:
am
I
behind
a
terrible
in
that
situation?
Those
are
often
enabled
together.
Yes,
we're
at
the
Q&A
session
now
so
you're
up.
A
D
D
A
That
is
open
any
question
yeah,
so
the
very
short
answer
is:
there's
a
lot
of
possibilities.
The
kind
of
more
where
I'd
see
stuff
going
is
ipfs
has
a
lot
of
very
powerful
potential
to
act
as
a
CDN.
It
kind
of
like
a
build-your-own
CDN
type
situation.
The
caveat
there
is
that
the
DHT
is
something
that
is
not
friendly.
To
that
my
nature.
A
E
A
Yeah
cluster
is
a
great
solution
for
that.
It's
interesting
right
now,
because
ipfs
dhts
are
kind
of
provide
they're
designed
for
very
popular
files
to
be
found
easily.
So
you
know
the
newest
version
of
Game
of
Thrones
comes
out.
There's
a
million
bad
people
out
there
in
the
world
that
are
all
torrent
in
it.
I'm.
A
The
their
access
to
very
infrequently
most
people
don't
care
about
that
data,
but
the
companies
themselves
still
want
that
data
to
be
findable.
So
it's
kind
of
one
of
those
things
we're
like
that's
where
actually
modern
web
kind
of
shines
with
DNS
with
ipfs.
It
gets
tricky,
so
it'sit's
kind
of
that
like
next.
E
A
Yep
yeah,
it's
a
it's
a
it's
kind
of
a
tragedy
of
the
Commons
situation,
where,
if
people
rely
on
gateways
too
much,
the
free
ones
are
gonna
get
abused.
If
you
want
your
your
gateway
to
be
popular
for
whatever
reason,
maybe
your
company
relies
on
that
gateway
as
it
gets
more
popular,
you
need
to
increase
the
specs
of
it
as
you
increase
the
specs
of
it.
More
people
want
to
use
it,
which
inherently
makes
it
run
slower
again
and
eventually
you
need
to
increase
the
specs
of
it.
C
A
And
that's
and
that's
the
problem
with
gateway
providers
right
now
is
this
is
this
is
why
CloudFlare
had
to
shut
down
their
their
video
streaming
is
because
how
gateways
work
is
they
go
and
when
you
fetch
content
from
the
Gateway,
it
caches
that
content,
and
so
many
people
were
basically
just
using
cloud
flairs
gateway
to
retrieve
video
content
and
I'm
sure
you
guys
get
this
a
lot
and
effectively.
It
turns
that
gateway
provider
into
a
makeshift
YouTube.
When
and
YouTube
gets
a
lot
of
traffic.
A
Video
uses
a
lot
of
bandwidth
so
and
then
the
sad
thing
is
they're.
The
original,
like
sources
of
that
data,
the
nodes
that
are
seating
that
data
initially
they
only
pay
the
bandwidth
for
giving
that
content
to
the
Gateway,
but
then
the
Gateway
basically
foots
the
cost
of
providing
that
content
once
it
has
it,
because
everybody
is
using
the
Gateway
to
retrieve
the
content,
so
it
it
becomes
a
very
big
problem,
very
fast
if
you're
not
careful
about
it.