►
From YouTube: 5. #everyonecancontribute cafe: HashiCorp Waypoint
Description
Visit https://everyonecancontribute.com
Agenda: https://gitlab.com/everyonecancontribute/general/-/issues/54
Blog: https://about.gitlab.com/blog/2020/10/15/use-waypoint-to-deploy-with-gitlab-cicd/
Getting started: https://www.waypointproject.io/docs/getting-started
Learn: https://learn.hashicorp.com/waypoint
Roadmap: https://www.waypointproject.io/docs/roadmap
A
We
had
to
hear
myself
on
youtube
now.
The
thing
is
we
kind
of
thought
about
what
we
wanted
to
discuss
next
and
one
of
the
things
was
yeah,
maybe
learning
rust
or
maybe
doing
something
else,
and
then
there
was
hashicorp
waypoint
last
week
and
for
me
this
is
totally
exciting
because
it
was
last
kind
of
like
trying
things
out
thursday
afternoon
or
thursday
evening,
and
I
was
wondering-
maybe
so
I
don't
have
a
plan
for
today,
but
I
was
asking
brandon
and
our
buckler
to
join
turn
on
in.
A
B
B
I
guess
now
they
have
a
much
larger
product
portfolio
right,
but
they
we've
known
them
to
have
a
very
tight
focus
on
you
know
terraform
and
vault,
and
of
course
they
had
consul
and
packer
and
vagrant,
but
then
they
added
to
last
week
so
one
was
boundary
which
is
not
waypoint.
So
I'll
talk
about
that
first
and
then
I'll
get
the
white
point.
Boundary
is
like
this.
B
I
don't
even
know
how
to
describe
boundary
a
proxy
like
the
idea
is
not
having
to
take
a
whole
bunch
of
steps
to
get
like
connected
to
your
production
instances
of
things.
So
that's
really
cool.
I
haven't
played
with
that
at
all.
Yet
that's
another
cool
thing
that
I
haven't
played
with,
but
waypoint
and
I'll
share
my
screen,
for
that
is
basically,
I
think,
did
I
say
terraform
for
your
applications.
Is
that
what
I
said
at
first,
michael
or
terraform
for
your
build,
I.
A
B
C
B
To
no
matter
where
you're
going,
if
it's
kubernetes
or
nomad
or
ec2
or
ecs,
theoretically
you're,
you
know
abstracting
away
that
from
from
the
developer,
which
may
or
may
not
make
sense,
depending
on
what
you're
doing
and
then
also
the
simplicity
of
vagrant
right.
So
the
thing
that
vagrant
gave
us
all
or
gave
many
of
us
in
our
development
environments
is
like
hey.
B
You
know
to
a
repository
of
some
sort
and
then
deploy
is
like
okay.
We
want
to
take
that
artifact
that
docker
imagery
created
and
deploy
it
somewhere.
So
that
could
be
like
a
kubernetes
cluster.
It
could
be
an
ecs
cluster.
It
could
just
be
to
docker
running
on
a
machine.
It
could
be
the
nomad
but
there's
a
number
of
different
deploy
targets
basically
and
you
target
the
platform
and
then
we
take
your
image
and
and
push
it
out
there,
and
this
thing
that
I
I
don't
fully
understand.
B
Yet
is
this
concept
of
a
deployment
url
where
okay?
If
I
deploy
a
no
matter
where
I
deploy
it,
I
get
this
like
waypoint.run
url.
That
is
a
direct
link
to
that
deployment.
So
it's
pretty
interesting
methodology
there,
although
I
don't
fully
understand
it
yet
and
then
releasing
so
like
it
also
has
this
concept
of
you
know,
taking
away
the
old
deployment
and
releasing
the
new
one.
B
So
that's
the
basics
of
the
idea,
but
you
know
at
its
core:
it's
really
just
kind
of
like
a
cli
tool
that
that
does
that
and
again
it's
defined
with
this
like
an
hcl
file.
So
it'll
look
familiar
to
the
folks
that
may
be
familiar
with
terraform
right
and
this
kind
of
hcl
definition
of
what
my
app
is,
where
it
is,
how
I'm
going
to
build
report,
deploy
and
release
it,
and
then
it
has.
B
So
right
now
it's
it's
like
that's
the
the
way
that
one
runs
it
in
production
in
theory
is,
you
know,
running
it
inside
of
some
sort
of
other
orchestration
tool,
and
so
that's
kind
of
the
work
that
we
did.
Michael.
I
don't
know
if
you
want
to
where
you
want
to
go
from
here,
but
that's
that's
way.
Point
in
a
nutshell.
A
D
A
Yeah,
probably
I
should
have
mentioned
that
I'm
now
taking
the
shortcut
into
finding
how
we
can
improve
the
ci
cd
templates,
because
the
previous
example
was
a
like
many
steps
after
the
other,
and
this
can
certainly
be
improved
and
abstracted
away
in
this.
I
think
we
came
to
the
conclusion
when
we
had
all
the
devops
and
and
including,
for
example,
the
security
scanners
into
the
gitlab
ci
configuration
so
yeah.
I
think
this
is
where
we
want
to
look
at
now.
Yeah
and
you
have
it
opened
already.
D
Yes,
exactly
so,
the
steps
here
involves
okay
for
every
time
you
run
it.
You
have
to
like
get
the
release.
You
have
to
zip,
it
move
it
to
where
it
comes
executable.
Before
you
run
where
you
need
to
run
so,
it's
not
efficient
and
we
decided
to
make
it
more
easier
and
not
having
to
specify
a
fashion
of
waypoints
that
you
want
to
run
like
in
the
example
here.
If
you
notice,
a
specific
fashion
to
download
was
not
specified.
D
So
one
of
the
things
we
did
was
to
we
have
this
draft
camera
that
brenda
worked
on
and
should
be
part
of
kit
lab.
Soon
we
created
templates
pieced
of
what
was
shared
on
the
weapons
documentation,
which
is
this
waypoint
gitlab
ci
and
all
it
does,
is
to
write
from
within
your
ci
cd.
You
can
include
viewpoint
as
a
template
without
having
to
write
it
yourself
and
it
uses
images
that
were
already
prepared.
D
We
created
a
project,
we
point
images
where,
basically,
what
it
does
at
the
moment
is
just
to
download
the
waypoints
binary
package
into
images
and
removes
the
entry
points
that
the
default
locker
hope
image.
That
hashicorp
has
adds
so
the
entry
point
breaks
ci,
because
if
you
want
to
display
like
a
description
of
what
the
binary
does
to
your
output,
but
with
the
ones
we
created,
we
removed
the
entry
point
and
made
it
much
more
easier
for
us
to
be
used
within
our
cicd
template.
D
It
also
makes
deals
to
run
faster
because
it's
already
on
gitlab
and
you
just
using
it
as
this
image
for
your
for
your
ci-
will
make
the
job
run
much
more
faster
instead
of
having
to
download
move
things
around
and
so
on.
So,
basically
for
now,
and
one
of
the
reasons
why
we
also
decided
to
do
this
is
like
the
telephone
images
that
we
have
gitlab
depending
on
the
need
in
the
future,
might
want
to
do
some
bootstrapping
and
other
things
around
the
binaries.
D
D
More
scripting
or
more
code
can
be
added
to
it
to
probably
make
it
work
with
auto
devops,
or
do
some
other
things
that
are
necessary
to
want
for
your
bills
to
be
more
efficient
when
they
are
using
the
images.
So
this
is
basically
what
image
does
this
project
does?
It
creates
for
every
version
of
waypoints
that
is
out
there.
It
creates
it,
creates
images
for
them
and
also
marks
the
most
recent
one
as
they
test.
D
If
you
look
at
the
container
registry
here,
we
have
all
of
the
versions
I
think
between
when
it
was
released.
To
date,
there
has
been
like
four
versions
that
have
been
released
from
0.1.0
up
to
0.1.3,
so
it
makes
it
much
more
easier
to
not
have
to
worry
about.
If
you
want
to
use
any
of
the
latest
versions,
you
can
just
specify
latest
latest
which
at
the
moment,
uses
0.1.3
now
and
for
people
that
might
worry
about.
Okay.
What
other
things
goes
into
the
the
images?
D
F
A
A
B
B
B
Well,
yeah
yeah
I
wanted
to
well.
I
wanted
to
play
with
it
and
I
wanted
to
like
have
waypoints
so
like
the
nice
thing
it
with
git
pod
is
like
you
can
define
the
container
you're
running
inside
of
right
and
so
like
I
installed
waypoint
into
the.
So
if
you
go
to
the
project
and
open
it
and
get
pod,
you
like
also.
E
E
B
B
B
A
B
C
B
B
Well,
yeah,
oh
yeah.
Instead
of
like
the
the
user
interface,
I
like,
I
have
no
trust
of
it.
I
just
I
like
I
mean
it
works
most
of
the
time,
but
that's
almost
the
problem
is
that
it
works
most
of
the
time
and
so
when
it
messes
something
up-
and
I
don't
understand
what
it
did-
I
get
confused
and
angry
at
it.
I
have
a
buddy
who
just
like
exclusively
uses
that
in
vs
code
and
like
I
cannot
pair
a
program
with
him
because
I'm
like
I
just
something's
gonna
go
wrong.
B
C
B
H
B
B
C
B
That's
a
great
question:
it
does
all
that
for
you,
so
I
I
only
I
just
gave
it
an
access
key
and
a
secret
key
and
then
and
that's
all
I
ever
have
given
it,
and
it
actually
we'll
see
it
here
once
it
gets
there,
but
it
actually.
If
we
pull
up
my
aws
account,
it
created
the
like
waypoint
automatically
created
the
ucs
cluster
created
like
I.
B
B
B
I
think
I
guess
it
just
names
the
cluster.
Maybe
you
can
specify
the
name
somewhere,
but
it
finds
this
easy.
You
know
it.
If
this
ecs
cluster
was
not
there,
it
would
create
it
it
create.
So
these
things
that
it's
saying
it
found
existing.
It
originally
created
the
first
time
it
ran,
found
the
im
rule,
so
it
makes
a
new
task
definition
make
sure
the
alb
is
targeting
it
security
groups,
like
all
that,
just
kind
of
abstracted
away
from
me.
E
B
F
G
B
B
Right
right
here
it
says:
storing
the
history
enabling
functionality
such
as
logs
exact
yeah,
so
the
the
other
idea
of
waypoint
that
we
haven't
even
touched
on
is
like.
I
want
to
be
able
to
just
say,
waypoint
logs,
and
it
doesn't
matter
where
this
thing
is
running
I'll,
get
the
logs
of
my
app
back
like.
Even
if
there's
not
like
a
standard
way
to
do
that
or.
B
H
C
B
B
A
B
G
C
A
B
G
A
B
G
G
B
B
C
B
Yeah
this
is
the
the
idea.
Is
you
can
see
all
of
this
like
releases
and
deployments
and
the
logs
from
this
thing,
etc.
But
obviously
I
don't
fully
have
this
working.
If
that
wasn't
obvious,
and
then
there's
this
concept
of
like,
like,
I
don't
know
how
the
authorization
and
authentication
to
your
point
of
like
by
environment
is
going
to
be.
There's
like
an
invite
thing
where
I
can
be
like.
Oh
here's
like
a
thing,
if
you
want
to
start
deploying
to
waypoint,
but
obviously
that's
not
their
long-term
vision.
A
I
think
it
will
also
be
interesting
to
see
how
this
integrates
with
vault
and
also
like
existing
strategies
with
terraform
and
how
how
the
whole
ecosystem
of
hashicorp
will
evolve
based
on
the
two
new
product
analysis,
maybe
maybe
everyone
else
in
in
the
in
the
coffee
chat
round
now,
did
you
check
out
check
it
out
already?
What
what
did
you
think
about
the
announcement
last
week.
C
So
I
need
to
be
really
confessed
that
I
don't
see
all
the
new
features
out
of
that
literally,
I
think
it's
a
good
integration
in
all
of
the
products
because
of
the
strategy
that
hashtag
tries
to
force
so
that
they
also
use
zero
trust
model
for
all
the
networking,
all
the
components
so
that
we
have
everything.
If
you
want
to
ensure
trust,
you
need
to
control
all
the
stuff.
So
one
of
the
missing
pieces
was
the
deployment
tool
right
of
the
box.
C
So
in
the
past
there
was
also
auto
but
autobots
too
early,
and
mostly,
I
think
it's
a
more
rewrite
of
the
whole
application
stack
yeah,
and
I
think
it's
a
really
interesting
part
because
I
said
earlier
mostly
because
I
think
it's
a
more
idea
of
having
the
cloud
native
application
bundles
so
that
you
stick
everything
together.
So
your
application,
your
deployment
of
the
infrastructure
part
so
that
you
can
every
deploy
every
stuff
with
only
one
command
in
the
end.
C
So
that
you
reduce
the
complicated
stuff
a
lot
of
a
lot
of,
but
the
complexity
can
be
like
it
can
be
huge
in
the
end
so,
and
the
opportunity
of
that
is
that
you
can
have
a
continuous
validation
of
audio
changes.
Mostly
so
mostly
the
continuous
ct
would
be
the
most
important
part
about
that,
so
that
you
can
easily
also
continuous
deploy
all
the
stuff.
C
I
found
video
really
interesting
that
they
have
a
client
server
model
and
I
will
see
in
future
how
it
will
fit,
and
probably
I
will
implement
it
in
a
prototype
in
our
own
company
for
some
stuff,
because
I'm
trying
to
work
on
some
similar
problems
and
we'll
see
how
it
goes
in
terms
also
on
security,
mostly
because
I'm
interested
in
how
you
can
save
securely
all
the
secrets
and
so
on,
yeah.
So
it
would
be
for
my
side.
G
G
It's
somehow
like
waypoint,
but
I'm
not
sure
if
I
really
get
all
the
points
at
the
moment.
So
this
whole
server
component
part
is
not
really
key
to
me
at
the
moment,
but
basically
it's
the
first
version
and
I
also
shared
the
in
the
the
chat,
the
road
map.
I
think
there
are
some
nice
parts
on
the
right
mirror
so
yeah.
I
will
have
to
dig
into
it.
Maybe
on
the
weekend
or
the
weekend
after
or
tonight,.
A
H
So
I
have
no
time
to
dig
into
it
also
most
the
same
kind
that
I
want
to
to
have
the
experience
in
the
in
this
meeting
in
this
session.
For
me,
it's
it's
coming
from
a
non-cloud
native
site.
It's
really
interesting
because
it
looks
like
that.
It's
now
much
easier
to
get
into
cloud
native
because
it
do
the
most
magic
for
me,
and
so
it
could
be
easier
to
get
in
and
also
what
you
like
is
that
it
supports
nomad
out
of
the
box,
because
on-prem
nomad
is
better
than
on-prem
kubernetes.
H
I
can
guess
that
the
server
is
there
to
make
a
punch
hole
for
the
for
the
api
so
that
the
users
to
host
their
own
api
service
and
don't
use
the
api
servers
from
the
cloud
providers
themselves,
but
yeah.
I
definitely
have
a
plan
to
to
try
this
out,
and
hopefully
I
can
get
the
service
up
and
running
on
the
weekend.
So
I
will
see.
A
E
E
This
or
the
biggest
cicd
part
that
the
company
is
still
jenkins,
unfortunately,
but
yeah
go
to
jenkins
and
beside
that.
Most
of
the
stuff
is
just
python
circle
ci
and
we're
doing
a
lot
with
terraform,
and
I
would
say
python
and
terraform
is
like
this
saying
that
I
spend
most
of
the
time
of
the
day
with.
A
F
F
Our
team
is
currently
working
on
a
deploy
ruby
in
docker
to
cloudrun,
using
google
cloud,
build
and
terraform
for
a
couple
of
adjacent
things
and,
and
so
waypoint
is,
is
really
highly
interesting
to
me.
The
thing
that
makes
me
a
little
bit
hesitant
is:
we
are
like
maybe
two
months
into
the
project,
just
onboarding
a
completely
new
team
on
on
on
that
infrastructure,
because
none
of
us
has
done
any
cloud
work
before
and
already
our
project
is
more
complicated
than
what
I
saw
now
can
be
addressed
with
waypoint
right.
F
We
are
running
unit
tests
during
the
build.
We
are
adding
resources
that
are
not
just
cloud-run
containers
like
a
firewall
and
a
network
for
our
environments,
and
we
have
to
interact
with
outside
systems
for
the
acceptance
testing
than
when
we're
actually
triggering
interactions
with
with
the
services
that
we
deploy.
So
I
I
I
could
see
how,
if
you-
and
it
was
already
a
huge
load-
to
learn
all
the
concepts
to
get
that
working
together
and
and
you're
now.
F
But
I
I
I
I
can't
see
how
waypoint
would
have
made
any
of
that
onboarding
at
the
start
easier
or
simpler,
because
I
think
we
would
still
have
to
understand
how
are
we
using
the
services
of
the
underlying
cloud
servers?
What
are
the
adjacent
services
that
we
have
to
integrate
et
cetera,
et
cetera,
and
that
also
means
I
don't
see
that
we're
that
mobile
to
another
cloud
platform
because
we're
using
firestore
the
json
thing
and
and
we're
using
the
cloud
apis
throughout
our
code.
So
I
I
don't.
F
B
B
F
That
that
did
not
reassure
me.
Sorry,
because,
like
I,
I
don't
actually
gcp
has
been
great
as
a
platform
for
us
because,
like
yes,
the
road
to
getting
all
the
code
bits
lined
up,
etc,
etc
and
figure
out
that
actually
we
just
need
to
use
the
ruby
libraries
they
give
us
and
use
the
defaults
and
they
actually
work
out.
And
if
you
have
the
I
am
configured
correctly,
it
just
connects
to
the
database
and
it
works
right.
So
it
was
hard
getting
there.
F
But
then
the
solution
was
simple,
and-
and
so
I
I
don't
see
how
waypoint
could
have
made
it
any
simpler
other
than
having
better
documentation
around
hey.
This
is
how
you
do
it
in
ruby
and
and
it
it
just
works
right
and
there
I
can
see
it
right
like
there
could
be
a
a
waypoint
demo
project
for
gcp
application
and-
and
I
just
start
from
there
like
waypoint,
init,
ruby
and-
and
I
get
a
field.template
that
already
has
all
those
integrations
kitted
out.
F
B
For
sure
yeah,
no,
I
mean
that's
exactly
like
I
mean
it's
this,
it's
a
similar
problem
space
as
get
lab,
auto
devops.
In
that
same
way,
right
like
hey,
we
want
to
make
it
really
easy
that
it's
kitted
out
and
it
works
out
of
the
box,
but
does
that
make
it
harder
to
than
actually
you
know
engineer
it
once
it
exists
like
heroku?
B
I
don't
think
people
have
really
solved
for
the
other
side
like
I
I
work
for
gitlab,
I
don't
think
autodevops
solves
for
that
other
side
completely,
I
mean
I
like
some
of
the
decisions
we've
made
it's
plug
and
play
like
you
can
pull
different
pieces
out
of
it.
That's
great!
You
can
get
security
from
gitlab,
but
then
like
go,
do
your
own
deploy
stuff.
When
you
know
you're
gonna
have
to
do
like
your
own
thing.
B
I
would
say
the
thing
that
got
me
so
interested
in
this
as
an
idea
is
that
the
only
people
that
have
ever
done
that
successfully
for
anything
is
hashicorp
with
terraform
right,
like
people
do
actually
use
terraform
instead
of
the
aws
cli
a
lot,
and
so,
if
anyone
has
a
chance
at
doing
it,
it
might
be
a
waypoint,
but
then
again
in
a
year
you
know
waypoint
could
be
nowhere.
I
don't
know
yeah.
C
I
need
only
to
know
how
the
aws
packing
is
rightly
working
so
for
providers
what
you
know
I
can
create
for
resource,
but
at
least
you
need
to
understand
the
concept
of
aws,
but
this
is
also
the
same
for
waypoint.
So
you
can.
You
can
extend
your
own
pipeline
with
your
own
plugins
if
you
need
to,
but
you
can
rely
on
many
plugins.
C
This
mostly
like,
like
with
terraform,
I
think
the
interform
you
have
at
least
250
custom
patterns
right
now,
custom
providers
that
are
officially
or
really
mainstream
supported,
and
then
we
will
see
also
what
is
interesting
about
the
project
concept
that
you
have
different
phases
where
parkins
come
in,
so
you
have
a
builder
phase,
you
have
a
deployment
phase
or
you
have
a
release
phase.
So
they
can.
C
Your
protein
can
play
your
strengths,
so
you
don't
need
to
build
a
plugin
that
is
omnipotential
that
can
do
everything
you
have
a
little
more
finer
strain
so
that
you
can
say.
Okay,
my
protein
is
really
good
in
the
builder
phase
or
it
can
be
good
in
the
release
phase
or
it
can
be
rooted
in
the
deployment
phase
yeah,
and
then
we
will
see
mostly.
C
It
depends
on
the
adoption
rate,
mostly
because
it's
like
with
every
new
tool
that
comes
out
it's
hard
really
to
bring
this
into
companies,
for
example,
to
change
all
this
stuff,
because
they
are
relying
typically
on
tools
that
are
working
in
the
past
and
whenever
adoption
is
not
easy
enough,
or
it's
can't
use
all
the
constraints
that
you
currently
have
in
your
company.
I
use
all
the
constraints,
then
it
won't
be
adopted
mostly
oh,
this
is
what
I
know
and
yep
we
will
see
it
in
future.
C
F
And
and
to
repeat
my
very
cynical
corporate
viewpoint
from
last
week,
I
I
don't
see
any
way
to
monetization
for
that
right,
like
yeah
sure
they
have
that
server,
and
you
can
have
a
centralized
point
and
you
can
integrate
it
with
your
authentication,
but
is,
is
that
is
that
their
monetization
play?
Is
that
what
will
give
them?
The
10x
return
on
the
nvc
that
they're
looking
for.
B
A
B
Trying
to
make
the
sas
their
play
right
because,
like
most
of
their
money,
comes
from
vault
today,
which
is
great,
but
if
they
want
to
monetize
the
rest
of
the
stack
it's
like
going
to
have
to
be
a
sas
play,
which
I
don't
know
if
they
win
that
or
not
that.
That
is
very
debatable,
but
you
know
the
terraform
because
they
had
a
number
of
like
cloud
things.
You
know
so
they've
got
terraform
cloud
which
is
kind
of
like
the
toe
in
the
water
thing
and
they're
all
excited
about
that.
B
C
C
F
That
sounds
a
lot
more
appealing
to
me
as
a
hobbyist,
right
and
and
then
you're
just
on
on
that
line
of
coke
to
to
commercialization
for
in
in
your
company,
because
it
works
so
well
at
home,
but
that's
also
a
very
long
game
to
play,
because
first,
you
need
to
build
it.
Then
you
need
to
find
people
who
are
using
it
to
deploy
their
applications
at
home
and
and
to
get
into
the
companies.
F
A
I
think
it
was.
There
was
a
good
last
good
famous
last
word
and
we
will
just
look
out
what's
coming
next,
I
would
say
thanks
for
the
for
the
interesting
insights,
I
have
no
idea
what
we
will
be
talking
next
week,
but
let's
just
see
if
something
new
pops
up,
otherwise
we
might
be
looking
at
the
kit
board
rust
or
anything
else
with
that.
I
wish
you
a
pleasant
evening
afternoon
morning,
depending
on
the
time
zone
yeah
and
see.