►
From YouTube: TGI Kubernetes 116: Deploying an app take 2
Description
Come hang out with Bryan Liles as he does a bit of hands on hacking of Kubernetes and related topics. Some of this will be Bryan talking about the things he knows. Some of this will be Bryan exploring something new with the audience. Come join the fun, ask questions, comment, and participate in the live chat!
A
A
Do-Do-Do-Do
there
we
go
hello
and
good
afternoon
good
evening
good
morning,
once
again,
if
you're
in
Hawaii,
this
is
TGI
K,
thank
goodness
it's
kubernetes
and
I'm
brian
miles
and
I
guess:
I
didn't
offend
enough
people
last
week
because
they
decided
to
bring
me
back
for
another
episode:
spoiler
alert,
I'm
gonna
talk
about
the
same
thing.
I
talked
about
last
time,
because
I
felt
in
that
hour
and
a
half
just
did
not
give
it
justice.
So
before
we
get
started,
let's
actually
see
who's
in
the
chat
all
right.
Let
me
pull
this
over
here.
A
So,
let's
see
what
we
have
here.
We
have
Lee
Maddie
and
Walid
and
Rodolfo
and
Martin
and
myself
hi,
Bryan
and
Christian,
and
let's
see
Dan
and
Steve
Maz
Wow
and
Vivian
hi
from
Munich
Jackie
and
Walid
and
I've
been
wow.
There's
many
people
here,
someone
from
Olaf
from
Copenhagen
alex,
is
not
here.
Sorry
Alex,
Pedro,
a
new
whom
Wow
so
many
people
in
here
so
we'll
do
like
we
do
all
the
time
and
let's
get
started
with
the
news.
So
let
me
find
the
news:
I
am
one
of
those
people.
A
I
have
two
screens:
they
are
both
32
to
34.
Inches
wide
and
I,
never
find
anything.
So
here
we
go
so
the
news.
Let's
actually
have
a
look
at
that.
Let
me
fire
up
my
screen
here,
so
you
all
can
see
what
I'm
seeing
all
right.
So
the
big
news
I
mean
the
big
news
here
is
Cooper
Denny's
1.19
alpha-2
is
out,
there's
more
updates.
Actually,
click
on
this
link.
I
have
not
clicked
on
this
link.
Yet,
let's
see
what
it
says,
I
was
just
a
changelog.
Oh
wow,
there's
lots
of
things
in
there.
A
I
do
know
that
there
was
a.
There
was
a
webinar
earlier
today
where
they
were
going
over.
Some
of
the
newer
features.
I
don't
know
if
they
were
focusing
on
118
or
119,
hopefully
119.
So
whoever
will
did
that
will
give
way
more
justice
than
I
am.
But
the
one
item
in
here
is:
let's
see
if
I
can
find
it.
Oh,
no.
It
was
a
118
fish
issue.
It
was
something
about
a
feature
that
I've
been
looking
for.
A
Is
that
whenever
you
launch
a
workload-
and
you
know
pods
and
it's
slow,
kubernetes
gets
a
little
bent
out
of
shape
if
things
don't
come
up
in
time
and
and
118,
they
actually
added
something
to
make
that
easier.
So,
let's
look.
What
else
is
in
the
news?
Ooh
there's
the
big
one
kind
0.8
has
been
released.
A
A
Let's
see
what
else
we
have
here:
wow,
it's
changing
in
real
time,
nice,
so
helm
is
a
graduate
at
CNCs
project,
shout
out
to
all
the
core
Deb's
on
helm
and
all
the
people
who
have
put
in
all
the
work
over
the
past
few
years
to
make
it
so
that
make
it
so
that
home
can
get
to
the
graduated
status.
It's
not
easy
to
get
there.
A
I
actually
do
work
a
little
bit
of
work
with
this
UCF
and
and
do
know
that
there's
actually
quite
a
bit
of
things
that
need
to
happen
for
this
to
happen.
So
shout
out
to
that
team
delish
changed
to
say
what
I
already
said.
Another
item
here
is
cube.
Ctx
big
update
its
right
lingo.
Well,
that's
kind
of
interesting
I
actually
do
use
cube
CTX,
and
for
those
of
you
who
don't
know
what
cube
CTX
is
actually
I'll
go
to
the
homepage.
A
Cube
CTX
allows
you
to
change
your
context
and
your
namespace
from
the
command
line
for
people
who
work
in
multiple
contexts,
whether
they
want
to
work
in
with
multiple
users
at
once,
or
they
want
a
mult
working.
Multiple
clusters,
cube
CTX
is
actually
is
actually
one
of
my
favorite
tools.
If
I
go
to
my
shell
right
now
and
this
piece
right
here,
this
mini
cube-
this
is
all
controlled.
Well,
it's
pulled
out
up
here,
but
I
can
definitely
miss
class
first
night
up,
but
I
can
definitely
run
cube.
A
Ctx
and
Cuban
s
to
get
around
I'll
do
a
little
bit
more
of
that
when
I
get
into
today's
topic,
so
shout
out
to
I'm
it
actually
a
big
shout
time.
It
is
this.
Guy
does
so
much
amazing
stuff,
whether
it
be
coming
from
writing
interesting
things
around
the
crew
project,
or
you
know
what
he's
Turkish
I'm
gonna
do
a
Turkish,
some
Turkish
content,
so
people
can
actually
get
content
and
the
language
next
weekend.
You
know
this
is
a
pretty
big
shot
out
to
I'm
a
big
fan
all
right.
The
next
time.
A
Next
item
on
this
list
is
domesticating
kubernetes
for
the
home
server.
Oh,
my,
let's
see
what
this
is
so
vladimir,
a
copyin
has
written
this
article
on
writing.
Kubernetes
are
running
kubernetes
on
bare
metal
in
150
minutes,
Wow
I.
Actually,
if
you
all
are
doing
this
I
have
definitely
given
up
on
this
whole
pursuit
and
either
run
it
and
VMs,
but
never
on
metal,
but
shout
out
to
Vladimir's
for
actually
creating
and
documenting
his
his
efforts.
A
A
I
read
it
backwards,
I,
don't
know
if
that
works
for
everyone,
but
that's
definitely
how
I
read
B.
So
I
am
going
to
add
this
article
to
my
list
and
I
use,
notion
and
things
that
I
want
to
read
I
just
put
in
here
and
then
I'll
not
leave
my
browser
and
whenever
I
go
review
notion
tomorrow
morning,
it'll
be
there
for
me
next.
My
item
is
kubernetes
and
rust,
so
who?
So?
What
is
this
about?
A
So
I
know
that
rust
is
becoming
way
more
popular
as
the
stability
is
getting
there
and
people
like
the
type
safety
and
they're
checking.
So
they
don't
have
anything
like
no
null
pointer
exceptions,
but
it
changes
the
way
that
you
think
about
code,
but
I
really
like
what
Russ
brings
to
the
table.
So
what
I
was
going
on
in
here
so,
like
I,
said
before
my
way
that
I
read
I'm
looking
for
code,
I
didn't
see
any
code.
I
wish
I
saw
code
in
here,
but
I
dunno.
Just
looking
at
this
last
line.
A
Microsoft
is
definitely
putting
a
lot
of
muscle
into
doing
kubernetes
work
with
rusted
wasm,
so
I'm
interested
in
seeing
what
comes
out
of
these
efforts.
I
don't
know
my
friend
Brian
kettle
s'en,
who
runs
gopher
con
is
really
in
the
rest
right
now,
so
I'm
sure
they
have
some
pretty
cool
things
that
they're
working
on
right
now,
oh
my
gosh
I
thought
I
could
get
away
without
talking
about
this
one.
So
earlier
so
about
a
week
ago,
last
weekend
someone
came-
and
they
said,
wow
look
at
this
neat
idea.
A
A
I
went
crazy
and
and
I
started
doing
all
these
things
and
it
distracted
me
from
I'm
thinking
about
today,
so
I
put
it
away,
but
I
want
to
raise
the
gauntlet.
This
is
cool.
I
want
to
see
someone
do
something
interesting
with
this.
Google
Spreadsheets
is
pretty
powerful.
So
let's
see
what
people
could
do
with
this,
and
if
you
do
something
neat,
let
me
know,
and
I
could
be
a
little
shoutout.
A
Let's
see
in
the
last
item
on
our
list,
I
only
read
this
first
is
a
shout-out
to
Chris
Nova,
a
an
sis
dig
for
twitch.tv,
slash
sadness.
Chris
is
doing
a
weekly
session.
They're
not
doing
it
right
now,
but
chris
is
actually
a
pretty
good
friend
of
mine
and
I
will
always
support
what
she
does
so
give
this
a
watch
or
go
back
and
see
some
of
the
previous
episodes
all
right.
A
A
But
what
I
want
to
do
was
actually
have
this
whole
long
run
up
about
all
the
things
that
have
to
move
to
our
be
in
place
to
actually
have
app
deploys
and,
of
course,
20
people
have
done
it,
but
there's
still
a
huge
gap
from
the
number
of
people
who
who
understand
it
to
the
num
of
people
who
actually
just
copied
something
off
the
internet,
and
what
I
wanted
to
do
is
fill
that
gap
with
some
words
and
lots
of
yamo
to
show
exactly
what
you
can
do
here.
So,
let's
go
to
my
screen.
A
I
know:
where
are
you
on
the
screen?
So,
let's
go
to
my
home
is
over
here.
Let's
go
to
a
terminal,
and
let's
see
first
thing:
I
want
to
do.
Is
I
want
to
litter
this
thing
with
my
kubernetes
stuff
and
then
like
how
Brian
work
stuff
so
before
I've,
pointing
out
with
the
Cuban
s
and
the
keep
CTX
is
how
I
used
how
I
actually
changed
my
context
and
my
namespace
my
default
maintenace
all
the
time.
A
So
what
we
want
to
do
here,
though,
someone
change,
my
directory,
so
I'm
gonna
go
to
development
and
I,
don't
go
to
demos
and
listen.
There
I'ma
go
to
TV
I,
K,
1,
1,
6
and
you'll
notice
that
my
path
are,
my
prompt
has
changed.
It
shows
my
directory.
It
shows
that
I'm
using
go
one
dot.
14.2
I
should
take
that
out.
It's
not
really
important,
but
it
also
shows
that
my
kubernetes
cluster
has
changed
and
it's
changed
to
a
cluster
called.
A
Just
to
show
that
you
know
this
is
stuff
like
he
actually
does
work
out
in
the
real
world,
of
course,
but
I
chose
digitalocean
and
why
not
choose
distillation
1
I
was
one
of
the
first
engineer
said:
digitalocean
lot
of
people
don't
know
that
I
was
the
first
remote
engineer,
resolution
3,
I
wrote
software
digitalocean,
that's
still
being
used
and
still
people
are
still
making
it
better
to
this
day.
So
you
know
what
is
my
appreciation
for
all
the
good
experiences
I
had
there.
You
know
a
little
shout
out
and
I'll
use
visualization
for
this.
A
So
if
I
get
K,
node
I
think
this
is
like
a
three
node
cluster
around
one
16.6
before
we
were
running
mini
cube
with
1.18,
but
nothing
I'm
talking
about
actually
doesn't
doesn't
really
matter.
So
the
first
thing
I
want
to
do
here
is
do
a
little
recap,
because
you
know
we're
in
this
whole
coded
19
era,
and
every
week
feels
like
the
year.
I,
barely
remember
what
I
talked
about
last
week
and
so
I
actually
had
to
write
it
down.
So
what
do
we
do?
A
A
Just
let
me
go
back
and
this
1.14
is
my
go
version
like
I
said
before
that's
supposed
to
be
a
gopher
I
guess-
and
this
is
the
go
version
like
I
says
it's
not
very
useful
but
we're
doing
let's
say
and
I
first
started
by
building
docker
file.
So
we
can
build
image
for
and
pushing
it
off
to
docker
hub
and
then
I
showed
how
we
can
use
scene.
App
are
not
seen
ab
c
and
b
cloud
native
built
packs
to
actually
automate
the
building.
A
So
we
didn't
have
the
right
docker
files,
and
then
we
finished
up
with
tech
time
and
Tecton
tasks
to
actually
automate
that
piece
now
I
want
to
start
there,
but
what
I
want
to
do
now
if
I
want
to
automate
the
whole
thing?
I
want
to
automate
the
building
from
whenever
I
check
something
into
github,
and
then
I
also
want
to
make
sure
how
do
I
actually
automate
deploying
this
website
as
well.
A
So
that's
what
we're
going
to
do
today
and
there
was
a
couple
things
that
I
skipped
over
last
week,
but
I
think
I
should
have
explained
more
and
so
I'll
start
with
those.
The
first
thing.
The
first
thing
is:
I'm
I
used
this
tool:
tkn
it's
a
sea
alive
for
Tecton
and
last
week
I
was
just
sitting
here.
Typing
and
I
was
typing
like
Tecton
TR,
a
TR
f
stands
for
tasks
right
there
just
shortcuts.
A
So
this
week,
you'll
see
me,
do
Tecton
tasks
run,
maybe,
and
then
some
pipeline
run,
maybe
and
just
looking
there
that
that's
just
what
this
is,
but
there's
also
something
else
inside
of
Tecton.
That
I
did
not
show
you
last
week
either
so
first
thing,
I'll
do
is
I'll
load
up
octant
and,
let's
see
we'll
go
to
another
namespace,
we'll
go
to
the
Tecton
pipelines
namespace
because
its
installed
in
there
and
then
we'll
go
down
to
the
pods
for
Tecton,
dashboard
and
I
just
wanted
to
show
you
that
Tecton,
the
folks
over
there.
A
They
actually
wrote
this
dashboard
and
you
can
see
all
the
things
that
I
do,
because
I
wanted
to
make
sure
this
demo
worked
out.
Well
so
had
to
try
a
few
different
times.
So
the
neat
part
about
this
is
now
you
can
see
it
from
the
command
line
or
you
can
see
it
from
the
website.
It's
actually
quite
a
neat
tell
you
truth.
So
if
I
click
on
I
click
on
this
I
can
actually
go
through
these
and
if
I
click
on
one
of
these
I
can
see
the
build
block.
A
Pretty
neat
and
the
cool
part
about
this
is
that
these
logs
are
actually
stored
as
container
logs.
So
the
way
that
this
is
coming
out
is
that
it's
actually
hitting
the
API
and
saying
give
me
the
logs
for
this
container
and
then
it's
a
doing
all
the
nice
color
stuff
here.
So
misinformation
is
here:
it's
just
not
easily.
It's
not
easy
to
see
and
this
tool
does
make
it
easier
so
back
to
what
I
was
doing
last
week.
A
A
Actually,
this
is
the
command
line.
I
want
to
use
so
this
command
line.
Here,
let
me
make
this
just
a
little
bit
bigger.
There
we
go
so
this
command
line
is
Tecton
start
a
task
start
start
one
called
bill,
pecks
v3,
that's
from
this.
The
cloud
native,
build
packs,
use
a
service
account
called
buildbot
and
then
have
an
input
and
say
the
source.
Is
this
website
get
which
I'll
tell
you
what
that
was
again
and
it
output
is
an
image
and
then
we'll
show
the
log
so
the
source
and
the
source
website?
A
That's
a
description
of
a
github
repository
in
this
case.
It's
github.com,
slash,
Brian
L,
slash
website
and
the
image
is
just
an
image.
I
think
this
one
is
I'm
a
doctor
hub
engine
image
under
Brian
L,
slash
website,
:,
Tecton,
I,
think
so
so
we'll
go
ahead
and
we'll
build
this
and
we'll
see
what
happens.
A
A
Okay,
there's
nothing
here.
Let's
see,
let's
say,
guess,
there's
nothing
here,
but
it's
actually
still
running
so
we'll
just
look
at
the
command
line.
So
what
it
does
is
it
uses
the
cloud
me
to
build
packs
and
it
figures
out
that
you
have
a
go
app.
It
actually
goes
and
it
builds
it
and
now
what
it's
doing
is
it's
actually
pushing
it
off
to
docker
hub
and,
like
I,
said
before,
Brian
L,
slash
website,
:
Tecton,
so
I
go
to
docker
hub
just
to
make
sure
this
works.
A
If
I
go
there
go
over
here
and
I,
go
there
and
I
see
Tecton
alright,
so
that
task
does
work,
but
we
want
to
do
more
than
that.
What
we
really
want
to
do
is
I
really
want
to
go
here
to
my
website
and
let's
say
that
I
had
a
PR
or
or
PR,
or
you
know
what
we're
gonna
do
today.
I
just
edited
a
I
headed
it
right
here.
A
A
What
I
would
like
you
to
do
is
be
tagged
with
the
get
job
one
Oh,
whatever
head
is
in
master
branch
and
I
say
that
even
an
easier
way
I
want
to
actually
so
this
latest
commit
right
here.
I
want
it
to
be
tagged
with
this
latest
commit.
So
this
a
6-1
or
si
item
and
like
we
have
right
here.
So
how
do
we
do
that?
So
let
me
go
and
we're
going
to
do
a
lot
of
google
searching
today,
so
we'll
go
Tecton
pipeline
and.
A
Hopefully,
on
their
website,
because
I
actually
am
not
sure
about
this
nope
there's
no
fancy
images
here
on
their
website,
so
I'll
just
explain
it
to
you.
So
Tecton
has
quite
a
few
number
of
resources
we're
going
to
last
week.
We
use
tasks
and
tasks
run
and
pipeline
resource
this
week.
We're
gonna
use
a
few
more
and
this
first
one
that
we're
going
to
use
is
called
pipelines,
and
this
pipeline
is
what
the
pipeline
does
is.
It
takes
a
set
of
tasks
and
it
runs
them.
A
So
it's
a
waiting,
a
group
task
and
what
it
can
do
is
it
manages
the
resources
and
it
allows
you
to
do
that.
So
the
first
thing
that
we're
going
to
do
is
we're
just
gonna,
create
a
we're.
Gonna
create
a
pipeline
using
your
existing
tasks.
So
actually,
let's
see
how
we
can
do
that.
So,
let's
go
back
to
the
cluster
and.
A
K&Amp;S,
like
I,
said:
I
love
this
tool
and
I'll
tell
you
what
this
is
like
something
like
a
little
sidebar
here.
I
knew
that
you
could
do
up
and
down.
I
did
not
know
that
you
could
use
your
mouse
on
here
until
like
right
before
I
was
doing
this,
so
we're
going
to
change
to
our
live
namespace
and,
let's
see
what's
in
this
thing,
get
tasks.
A
A
So
the
first
thing
we
need
to
do
is
I'm
going
to
create
a
pipe
we're
on
a
critter
and
then
we're
going
to
test
our
pipeline
out
locally,
just
to
make
sure
that
it
works,
and
it's
going
to
be
pretty
close
to
what
the
task
was.
But
let's
get
this
all,
let's
get
this
all
running,
so
the
first
thing
I
need
to
do
is
figure
out.
Where
I
did
the
task
see
I?
Hopefully,
I
fit
all
this
up.
A
A
These
are
the
steps
that
are
required
to
build
with
cloud
native,
build
packs,
and
it's
actually
it's
a
lot.
I
think
there's
like
six
or
seven
steps
in
here,
so
I'm
not
gonna,
go
through
all
of
them,
but
just
know
that
we
tested
this
and
we
actually
just
tested
it
before.
We
do
know
it
works
so
now,
I
take
a
good
test.
I
should
have
my
build
packs
b3.
A
So
now
what
we
want
to
do
is
wrap
that
with
a
build
a
pipeline
and
the
way
that
we'll
do,
that
is,
let's
see
we're
gonna
I'm,
trying
to
just
I'm
debating
whether
I'm
gonna
type
it
in
or
I'm
gonna
cut
and
paste
from
a
website.
I
think
I'll
just
cut
and
paste
some
website
so
we'll
go
up
here
and
we'll
make
a
new
directory.
A
No,
there
we
go
so
now
we're
on
our
live
directory,
and
the
first
thing
we
want
to
do
is
create
this
I'll
just
call
it
build
pipeline
Yambol
and
what
I'll
do
is
I
type
this
in
because
I
am
an
engineer,
but
most
of
my
time
with
my
I
deal
with
kubernetes
is
just
dealt
with
wrangling
yeah
and
like
python,
it
kind
of
messes
with
me
sometimes
and
I
mess
up
the
indenting,
so
I
typed
them
all
in
just
to
make
sure
that
I
had
them.
So
let
me
explain
see
how
this
works.
A
So
this
is
a
pipeline
resource,
it's
a
tech
time,
dev
/v
1,
beta
1,
and
in
this
case
it's
called
build
pipeline
and
what
it
does
is
it
if
we
go
down
to
the
middle,
what
it
does
is
it
has.
It
has
one
task
and
then
whenever,
but
it
does
is
it
starts
from
beginning
and
it
calls
a
task
and
then
a
task
run
the
steps,
and
then
we
ought
to
get
the
information
into
it
and,
like
I,
was
telling
you
last
time
that
Tecton
works
with
this
concept
of
resources.
A
So
you
can
have
a
get
resource
or
an
image
resource
and
we're
actually
using
both
of
those
here.
So
what
we're
going
to
do
is
we're
going
to
supply
a
get
resource
and
supply
and
image
resource
to
this,
and
then
the
pipeline
will
tell
the
task
what
the
image
resource
and
the
and
the
source
resources
are
and
it'll
run.
So
this
is
just
a
wrapper
around
what
we
did
before.
A
So,
if
I
go
like
this
and
then
I
do
ok,
apply
eff,
build
pipeline.
It's
in
my
cluster.
Let's
actually
you'll
see
me,
go
to
this
tech
time
dashboard
quite
a
bit
today
and
go
to
pipelines
and
live.
We
have
this
thing
called
build
pipeline
and
neat
part
about
this
tool.
Is
that
I
don't
have
type
it
from
I?
Don't
the
type
all
the
time
what
I
can
actually
do
is
I
can
use
this
web
GUI
to
create
this
stuff
here.
A
So
what
it's
saying
is
that
there's
no
pipeline
resources
of
type
get
in
the
live,
namespace
yeah,
that
is
right
and
no
pipeline
resources
found
of
type
image
in
a
live
namespace.
So
we
need
to
come
up
with
those
two,
so
let
me
actually
come
up
and
let
me
read
the
comments
real
quick
before
I
go,
so
this
is
one
step
with
Co.
Yes,
I
know
all
you,
ex-googler
is
in
your
love
of
code.
I
will
actually
talk
about
codes
where's
the
end.
A
It
is
a
neat
tool
for
building
images
from
go
projects
does
Tecton
have
persistent
build
histories?
Yes,
it
can,
because
in
Tecton
what
you
can
do
is
you
can
actually
mount.
You
can
mount
an
image
along
with
the
steps,
and
you
can
actually
pass
that
that
volume
our
mounts
a
volume
along
with
the
steps,
and
you
can
pass
that
bong
around
and
inside
of
there.
What
you
can
do
is
you
can
actually
keep
cache
data
inside
of
there,
so
you
could
use
it
to
spits
to
speed
up
your
build.
A
I
can't
do
that
with
cloud
native
built
packs,
but
I
wanted
to
keep
it
simple.
So
I
did
not.
So
what
we're
doing
right
now
is
I'm
trying
to
remember
where
my
my
resources
are.
So
if
I
go
up
a
directory,
ooh
Tecton
resources,
so
okay
apply
f00,
Tecton
resources.
So
now
we
create
this
get
source
and
then
we
created
an
image
as
well,
so
whenever
I
reload
this
I'm
going
to
create.
Oh,
this
is
what
I
was
scared
of.
A
Let
me
try
this
one
more
time.
This
tool
is
been
a
little
finicky
for
me
today.
Hope
they
worked
this
time
so
now,
whenever
I
can
select
my
pipeline
resource
and
it's
website
good
and
the
out
and
the
output
one.
Is
this
website
I
like
this,
because
notice
I
didn't
configure
it
to
do
this
things
and
it
actually
knew
that
the
get
would
have
been
upfront
and
then
the
website
image
have
been
on
the
tail
end.
A
So
a
neat
thing
about
this
is
that
it's
way
better
than
the
command
line.
If
you
don't
know
what
you're
doing
I
mean,
the
command
line
is
definitely
faster,
but
if
you
don't
know
what
you're
doing-
and
you
want
to
see
this
thing
all
in
context,
this
tool
puts
it
all
in
context
and
actually,
by
the
end
of
this,
we'll
actually
touch
almost
all
these
types.
So
it's
a
lot
to
keep
in
your
head,
so
what
we
can
watch
it
run.
A
So
it's
it's
building
now
with
C
and
B,
it's
running
go
install
and
as
soon
as
it
runs
go
install
what
it'll
do
is
it'll.
Do
this
export
and
it'll
upload
it
to
get
up
and
looks
like
I'm?
What
I'm
trying
to
do
is
build
instead
of
jumping
all
the
way
to
the
end
of
trying
to
build
it
up
step
by
step
by
step,
so
you
can
actually
see
someone
go
through
it,
because
one
thing
that
I
noticed
and-
and
this
is
I
spent
a
lot
of
time
doing.
A
This
is,
if
I
go
to
text
on
pipelines
or
really
any
other
a
lot
of
other
software
out
there.
Oh
wow
wow,
there's
a
lot
of
docks
here
and
and
usually
they'll,
give
you
an
example,
but
no
one
answers
the
so
I
want
to
do
this
and
it's
like.
Why
would
I
want
to
use
the
workspace?
How
are
labels
helpful
here
and
so
what
I'm
trying
to
do
is
give
you
a
little
context
that
goes
around.
A
So
this
thing
is
all
done:
it
uploaded
it
to
Brian,
Allen,
slash
web,
say
and
if
we
go
look
at
this
image,
if
you
look
at
this
tech
time
image
it
shouldn't,
it
should
say
just
a
couple
of
seconds
ago:
yes,
so
we
are
there
so
now,
we've
created
this
pipeline
and
I
want
to
talk
about
something
else.
It's
kind
of
a
little
bit
of
a
sidebar,
but
it's
very
important
for
the
rest
of
this.
A
This
demo
uses
a
cluster
on
the
Internet.
This
is
NYC
3
out
in
somewhere
in
Jersey,
and
the
issue
here
is:
is
that
now
we're
gonna
have
communication
school
across
the
internet?
I
could
actually
go
the
easy
route
and
say
you
know
what
we'll
just
do
plain
text
and
TLS
is,
is
not
needed,
but
I
wanted
to
do
it
the
right
way.
A
So
what
I'm
going
to
do
is
change
my
name
space
in
an
octant
to
go
to
my
life,
except
because
we're
live
right
now
and
I'm
gonna
go
and
look
at
like
ingress
--is.
So
I
created
this
ingress
right
here
and
it's
called
live
and
the
reason
so
I
had
this
idea
when
I
thought
about
this.
The
other
day,
as
I
said
I
would
use
this
domain.
A
I
would
use
my
my
personal
domain,
Brian
L,
dev
and
then
I
put
a
TGI
K
116
in
front
of
that,
and
then
I
said
what
I
would
do
is
and
I
would
just
dynamically
generate
TLS
certs
with
with
let's
encrypt,
because
whenever
I
write,
whenever
I
first
started
using
less
encrypt
many
years
ago,
you
would
generate
a
cert
10
seconds
later.
Acting
would
show
up
a
couple
years
ago.
A
minute
or
Sally
would
show
up
this
week.
It
was
between
five
minutes.
A
If
I
did
it
early
in
the
morning
or
if
I
did
in
the
middle
of
the
day,
it
might
be
thirty
minutes
or
45
minutes
and
I
realized
that
that
was
not
make
for
a
good
demo.
So
I
paired
back
and
I
thought
of
another
idea.
So
what
I'm
going
to
do
is
everything
that
I'm
going
to
do
demo
today
is
going
to
use
this
live
to
gij
Brian,
L,
dev
and
I'm,
just
gonna
use
on
paths
that
way
I
don't
to
generate
any
more
certs.
So,
let's
actually
so.
A
What
else
do
we
need
to
do
now?
So
we
have
our
pipeline
running
and
what
we
need
to
do
is
figure
out
a
way
to
go
from
github
to
our
pipelines.
Our
friends
at
at
the
tech
time
project
have
actually
thought
about
that,
and
what
we
have
is
this
project
called
Tecton
tick
triggers
Tecton
triggers
not
Tiggers.
A
The
wonderful
thing
about
Tecton
triggers
is
triggers
are
wonderful
things.
I
guess,
y'all
have
no
kids,
but
what
this
is
about
is
it's
a
way
to
trigger
pipelines,
and
this
is
why
we
went
with
the
pipeline
first.
So
what
this
allows
us
to
do
is
have
something
external
and
it
can
send
a
request
and
then
the
Tecton
triggers
can't
answer
that
request
and
then
they
can
call
the
pipeline
and
the
pipeline
can
do
what
it
needs
to
do.
So
how
does
this
work
Wow?
Well,
let
me
explain
it
to
you.
A
It's
it
seems
like
like
once
again
it
seems
very
overwhelming,
but
it's
actually
pretty
simple.
There
are
three
new
terms
that
we're
going
to
talk
about.
One
is
called
the
event
listener
and
what
an
event
listener
does
is
surprise,
surprise
it
listens
for
events,
then
there's
this
thing
called
a
trigger
binding
and
what
a
trigger
binding
does
is
it
takes
the
payload
out
of
the
event
listener
and
it
binds
it
to
local
variables
or
parameters.
A
This
idea
called
a
trigger
template
and
well
and
we'll
go
through
all
those
eaten
warm.
Whenever
we
see
them
is
just
trigger
template
that
actually
estates
your
pipeline
so
that
it
can
run
so
it
goes
event.
Listener
bind
with
the
trigger
binding
and
then
call
the
tip
the
trigger
template
and
we
are
going
to
be
so
fed
up
with
the
ammo
by
the
time.
We're
done
with
this
we're
gonna
go
into
the
weekend,
pretty
happy.
So,
let's
get
started.
A
A
Automate
the
POI,
what
we
have
in
here
trigger
to
poi
cool,
that's
what
we
want
we'll
get.
This
trigger,
deploy
and
I'll
launch
this
up
and
I
will
walk
through
it.
To
show
you
what
we
have
so
like
I
said:
we
have
an
event
listener,
we
have
a
trigger
binding
and
we
have
a
trigger
template
and
let's
go
through
these
real,
quick
and
see
what
they
do.
So.
What
happens
here
is
the
event.
Listener
has
a
service
account
name
and
it
needs
a
service
account
because
it
has
to
create
use.
A
A
So
let's
see
how
this
works,
oh
wow,
and
then
we
have
this
trigger
template
and
what
the
trigger
template
does
is
it
takes
the
parameters
that
we
had
before
and
and
then
what
we
do
is
we
have
this
thing
called
our
resource
template
one
or
more
or
actually,
zero
more
and
we're
gonna
create
a
pipeline
run
so
before
whenever
we
went
when
we
were
in
text
on
dashboard
and
we
were
at
pipelines
and
I
was
a
build
pipeline
that
created
create.
We
have
this,
create
our
Tecton.
Our
trigger
template
is
actually
just
gonna.
A
So
let's
go
back
to
them,
and
so
let's
look
at
this,
so
what
it
does
is
match
the
params,
so
we're
gonna
get
a
get
revision,
a
URL
and
a
namespace,
and
what
we're
going
to
do
with
that
is
we're
going
to
create
this
pipeline
run.
We're
gonna,
give
it
a
name
of
deployed,
run
and
you're
gonna
see
a
lot
of
this
dollar
sign
semicolon
our
semi
circle.
Whenever
we
call
this
parenthesis
parenthesis
so
dollar
sign
left
paren,
you
would
write
paren.
What
happens
is
whenever
these
are
evaluated.
A
Tecton
triggers
knows
how
to
evaluate
those
two
something
so
in
this
case,
it'll
be
like
like
a
six
digit
alphanumeric
screen
or
or
in
this
case
its
params
that
we
had
before
or-
and
this
one
is
another
print
we
had
before.
So
this
useful
so
I'm
going
to
change
this
just
a
little
bit,
we'll
just
put
live
there,
because
we're
I
think
we're
in
our
live
main
space
and
I'll
go
ahead
and
I
will
create
this
thing.
A
A
What
I'm
waiting
for
is
my
my
there's
gonna
be
a
pod
and
the
service
created
that
I
can
actually
forward
my
ingress
to
so
I'm
waiting
for
that
to
show
up
so,
while
that
the
size
of
show
up,
let's
go
see
what
else
there
is
what
else
there
is
going
on
here.
So
if
we
go
to
event
listeners
and
we
have
our
deploy
listener
and
we
go
to
Yambol.
A
A
A
A
A
I,
have
this
service
account
name
and
then
I
have
a
trigger
binding
I'm
like
this
line,
and
then
we
have
our
trigger
template,
so
we
should
have
a
deploy
trigger
template
and
the
names
are
matched
up
here
and
we're
gonna
create
this
pipeline
run
namespace
and
that
namespace
would
be
alright
looks
right
now,
so,
okay
apply
dash
F
trigger,
deploy,
okay,
get
pod;
okay,
that's
and
recap:
okay,
Oh
ans
recap:
okay,
get
pod,
know
right!
So
ans
live
okay,
get
pod
yeah
we're
still
waiting
for
these
the
two
to
actually
show
up.
A
No,
it's
a
baktun.
No,
it
does
not
show
up
I,
don't
think
it
shows
up
in
the
Tecton
namespace
I'm.
What
I
need
to
do
is
I'm
going
to
to
detect
on
namespace
and
I'm
gonna
see
if
it's
spitting
out
errors,
so
we
go
Tecton
pipelines
and
what
I'll
do
here
is
I'll,
go
to
the
Tecton
triggers
controller
and,
let's
see
if
these
logs
say
oh
gosh,
oh
gosh,
there
we
go
so
it
is
for
40-something
local
time.
I,
don't
see
any
errors.
Let
me
get
rid
of
timestamps
and
only
show
this
one.
A
It
does
not
look
like
I
have
any
errors
here:
I'm,
just
eyeballing
I'm,
looking
at
levels
and
I,
don't
see
any
errors.
So
what
in
doubt
this
is
what
I
do
in
kubernetes
I'll
go
here
to
my
triggers
controller
and
I'll
just
delete
it
rather
than
typing
in
that
long
thing
in
the
command
line
and
if
we're
any
bit
lucky
it'll
be
recreated
and
it
has
been
recreated.
So
since
that
has
now
been
recreated,
maybe
it
got
stuck
I,
don't
know,
I'll
go
back
to
my
live
namespace
and
we'll
look
at
our
pods
and
nope.
A
So
here
it
is.
This
is
what
happened.
There
needs
to
be
a
service
account
called
Tecton
triggers
admen
I
did
not
create
the
service
account,
so
let's
go
ahead
and
create
that
service
account.
This
is
what
happens
when
you
do
things
in
front
of
people,
so
I
will
go
up
to
here
and
figure
out
what
that
looks
like.
So
what
I'll
do
is
I've,
okay,
in
recap,
good
service
account,
tucked
on
sugars,
admin
and
I'll.
Do
oh
yeah
mole
see
what
that
looks
like
all
right,
not
much
going
on
there.
A
A
We'll
get
this
one
so
we'll
go
okay,
bash
and
recap:
get
roll
bindings
and
then
we'll
spit
that
thing
out
of
Shamel
see
it
looks
like
yeah
and
once
put
it
out
to
our
back
now:
yeah
Mille
and
in
vim,
our
back
yeah,
Moe
and
and
we'll
just
get
rid
of
all
the
stuff
that
doesn't
need
to
be
here
now.
If
I
go,
teh
apply,
F,
R
back
and
I
go
now.
I'm
gonna
go
back
to
when
I
go
back
here.
What
I'll
do
is
I'll
go.
A
Look
at
my
my
replica
sets
and
how
does
it
like
this
one
I
think
it's
a
little
confused
so
now
what
should
happen
is
actually
get
a
new
replica
set
because
there
it
is
my
listen
all
right,
so
my
deployments
there
and
now
I
should
have
an
L
deployed
listener.
Yes,
we're
now
here!
So
yes,
I'm
debugging!
This
live
what
you
what
you're
seeing
is
that-
and
this
is
actually
a
problem,
that
a
lot
of
people
run
into
a
kubernetes
I
still
run
into
it
and
I
use
communities.
A
All
the
time
is
that
the
graph
of
objects
that
you
need
to
have
to
make
something
work,
sometimes
complex
and
it's
opaque.
So
a
good
example
would
be
you're
missing
a
service
or
you're
missing
a
service
account
or
a
service
account
doesn't
have
a
role.
What's
the
canonical
way
that
you
can
actually
figure
that
out,
you
know
I,
don't
know.
So,
let's
see
what's
going
on
now.
It
says
that
it's
it's
initialized,
see
the
log
say
so
now.
A
Oh
you
know
what
I'm
going
to
do
here,
I'm
going
to
delete
my
thing
and
restart.
Ok,
the
deletes
f
trigger,
deployed,
ok,
apply,
f,
trigger,
deploy,
so
I
can't
say
your
name.
Is
the
email
work,
sorry,
which
tools
made
this
dependency
graph
easier
to
see
on
the
UI?
Yes,
I
agree,
and
so
this
is
actually
why
I
created
octant.
To
tell
you
the
truth,
one
of
the
things
that
I
use
in
octant
and
I
and
I
just
need
to
spend
some
time
making
this
piece
better.
A
Is
that
what
does
object.create
and
what
does
it
refer
to?
This
is
not
a
complete
list,
but
often
actually
will
show
me
so
sometimes,
whenever
I'm
looking
around
here,
oh
there's,
my
pod
actually
have
all
this
stuff,
working
and
and
what's
broken
right
now.
So,
like
oxidants
telling
me
right
ii
that
no
replicas
exist
and
this
replica
set
says
expected
one
but
zero
exists.
So,
let's
actually
so
now
I
can
click
on
this
and
I
can
go
here
and
we
can
go
back
to
the
things
yep.
It
says
it's
happy.
A
A
A
It's
saying
that
there's
an
issue
between
octant
is
erroring,
but
it's
saying
that
there's
an
issue
between
the
role
and
the
service,
so
what
I'm
gonna
do
now
is
go.
Look
at
my
role
bindings!
Oh
I
know
the
problem
is
I
know
what
the
problem
is.
We're
referring
to
a
role
doesn't
exist.
That
is
it
so
oxen
should
have
been
on
will
tell
me
that,
but
did
not
code
that
in,
but
that
is
what
the
issue
is,
so
we
can
fix
that
too.
So
what
we'll
do
here
is
we'll
go.
A
Okay,
that's
in
recap,
get
roll
there.
We
go
and
oh
yeah
mole
into
our
back
yeah
mo
and
then
we'll
edit,
our
our
back
yeah
mo
and
we'll
get
rid
of
this
I
put
this
here.
We'll
get
rid
of
all
this
and
all
this,
and
all
this,
and
all
this
and
all
this.
So
what
the
role
allows
you
to
do
is
it
allows
us
to
create
it.
A
It
allows
us
to
look
at
event
listeners
and
then
it
allows
us
to
to
create
pipeline
runs
and
pipeline
resources,
which
is
what
we
need
to
actually
get
this
thing
started
and
then
it
allows
us
to
look
at
Big
Mac,
some
secrets.
So
now,
if
I
do
this
it'll
work,
nope
I
have
a
typo
in
my
gosh
there
we
go
so
now.
Oh.
A
A
So
the
million-dollar
question
is
that
I
plan
this
or
did
I
just
just
mess
up,
I,
don't
know
and
I'm,
not
gonna
tell
you
but
I
think
we
got
past
it.
So
it's
not
important
so
now,
I
go.
Look
at
my
pods
well
see
that
my
L
deploy
listener
is
now
created
yay.
That
was
a
lot
of
work.
I
wish!
You
know
really.
This
is
the
kind
of
tool
that
I
would
like
to
build.
I,
when
things
don't
work,
I
just
want
to
say:
hey
cluster.
A
What's
up
and
I
want
the
cluster
to
say:
hey
Brian
I
see
you're,
feeling
bad
you're
missing
a
role.
You
should
fix
that
and
then
Pat
me
on
the
head
and
send
me
on
the
way.
So
that
was
a
lot
of
work.
So
now
what
we
have
here
is
we
have
our
listener.
Listening
if
I
go
back
and
I
do
look
at
my
services,
we
have
an
L
deploy
listener.
Listening
but
now
we
need
to
do
what
we
need
to
do.
A
Is
we
mean
to
point
to
it
with
our
ingress
and
let
me
fix
that
and
make
that
work.
So,
member
before
I
was
talking
about
ingress,
--is
and
and
making
those
things
work
so
I
go
okay,
get
ingress.
I
do
have
my
live
once
ok,
get
ingress,
live,
that's
a
gamble!
That's
a
lot!
What's
actually
going
on
there!
A
I
don't
know,
but
if
I'm
looking
inside
of
octant,
if
I
go
look
at
my
ingress,
it's
a
little
bit
easier
to
see
in
octant,
but
I
can
see
that
I
already
have
one
for
this
live,
and
this
website
deploy
we'll
go
over
that
later.
So
what
I'll
do
is
I'll
just
edit
it
from
the
command
line.
Okay,
okay,
edit
and
one
thing
you
probably
know
this-
is
that
every
time
I
launch
me
of
them,
it
compresses
all
this
stuff,
and
this
is
actually
like
a
pro
tip.
A
You
never
know
how
big
the
Gamo
file
that
you
are
going
to
see
is
going
to
be
so
what
I
do
is
I
actually
automatically
fold
the
whole
entire
file,
because
I
don't
want
to
see
if,
if
I'm,
not
ready
for
that,
I
don't
want
to
see
it.
So,
in
this
case,
I
only
want
to
see
what's
going
on
in
here,
so
we'll
go
into
pack
ins
and
what
we'll
do
is
we'll
create
another
one
and
we'll
call
it
back
in
and
we'll
call
it
service
name.
Oh
gosh.
What
was
that
sugar?
A
And
we'll
go
one
more
and
the
service
port
is
80
80
and,
and
we
look
bad
half
and
we'll
call
it
I'll
call
this
one.
What
was
called
the
ploy
one?
No,
no
can't
call
the
play
one,
because
it's
not
deploy
one
will
call
build
one
so
now
inside
of
my
ingress.
So
if
I
go
back
to
octet
now
and
I
go
look
at
this.
No,
if
I
go
look
at
the
end
grass.
A
Where
is
my
ingress
there?
It
is
I,
go
look
at
this
ingress
and
we
look
at
live.
There's
gonna
be
two
rows
in
here
and
we're
gonna
have
slash
build
one.
Whenever
you
hit
build,
one
will
go
to
this
El
deploy
listener
and
whenever
you
get
to
this,
and
you
go
to
web,
say
that
deploy
it
goes
there,
so
we're
almost
there.
The
next
thing
we
need
to
do
is
we
need
to
generate.
We
need
to
create
a
github
webhook.
Now
there
are
two
ways
we
can
do
this.
A
A
Alright,
maybe
won't
come
back,
we
can
add
it
here.
The
problem
is:
is
that
what
are
you
gonna
add,
but
what
we
need
to
add
is
we
need
to
have
a
what
we
need
to
have
webhook
that
goes
to
my
live
dot,
whatever
slash
build
one
and
it
needs
to
have
the
payload
in
there
and
that's
actually
that's
a
lot.
What
we
should
be
using
computers
to
reduce
the
cognitive
load
that
we
have.
So
what
I
did
is
I
wrote
this
thing
and
let
me
see
if
I
can
find
it
all
right.
A
Great
github
webhook,
look
at
that,
so
so
what
we'll
do
here
is
we'll,
go,
let's
say
and
we'll
change
this
thing
to
be:
slash,
build
one
and
we'll
go
here
to
live
I,
don't
like
doing
things
manually
that
we
can
do
with
the
tool.
So
I
just
wrote
this
tool,
real,
quick
that
actually
will
go
out
to
the
github
and
if
I
go
reload
this
page.
No,
not
this
one!
Look.
We
have
a
new
web
hook.
So
the
neat
part
about
this
is
the
head.
A
There's
this
green
check,
which
means
that
it
was
actually
able
to
talk
to
something
and
then
so
that
means
that
we
have
wired
it
all
the
way
up
and
we're
able
to
talk.
So,
let's
see
if
we
can
actually
make
it
work.
So
what
I'm
going
to
do
is
I'm
going
to
go
to
my
website
and
I'm
just
gonna
not
have
changing
the
code.
I'm
gonna
break
anything
well
even
on
accident,
so
we
changed
it.
A
We
had
another
exclamation
point,
we're
super
happy
now
we're
gonna
do
a
tour,
our
dashboard,
and
what
we're
going
to
do
is
we're
going
to
see
if
there's
any
pipeline
runs
running
and
I
don't
see
any
so,
let's
actually,
let's
dive
in
and
figure
out
what
happened
here.
So
what
we
have
play,
the
way
that
we
diagnose
this
is
I,
go
to
the
L
deployed
listener
and
I
go
look
at
its
logs
and
it's
gonna
tell
me
real
quick,
what's
going
on
here
so
now.
A
What
it's
saying
is
that
well,
actually
it's
saying
that
it
might
have
worked
so,
let's
clear
that
so
it
ran,
and
what
it's
saying
is
no
task
runs
found
for
this
pipeline.
Yet
well,
that's
interesting
because
we
ran
it
from
the
command
line
earlier.
So
what
is
what
does
that
mean?
And
that
means
we
have
to
go
back
to
the
command
line.
So
okay,
get
pipeline,
run.
A
A
Do
you
see
what
the
problem
is
here?
The
problem
is
I
think
it
ran
the
wrong
thing.
So
what
I
want
to
do
here
is
look
at
my
trigger
to
play
again:
oh
wow,
okay,
so
I'm
just
going
to
get
rid
of
this
entirely
and
I'm
gonna
go
back
up
to
my
other
directory
and
copy
over
the
right
one,
this
time
and
by
looking
alright
copy,
no,
not
trigger
deployed.
A
Okay,
okay,
alright!
So
now
what
we
have
is
I
think
actually
the
right
one.
No
one
pointed
out
in
the
in
the
chat
that
I
was
using
the
wrong
one,
the
whole
entire
time,
but
how
about
youth
now
so
k5f
trigger
dot
Yambol?
Let's
actually
make
sure
that
it's
calling
the
right
one.
So
steve
says:
how
do
I
do
that
really
I'm,
just
using
Zemke
mate
Zim
them
commands,
so
I
use
lowercase,
Z,
lower
upper
case
R.
A
It
just
opens
the
whole
file,
so
I
want
to
go
up
to
my
template
and
see
a
service
count
name
and
build
pipeline.
Okay,
get
pipeline
build
pipeline,
alright,
so
k
apply
after
so
so
what
I
did
before?
Let
me
explain
to
you,
because
this
is
kind
of
fast
I
copied
over
the
wrong
pre-generated
template,
so
it
worked.
The
problem
is,
is
that
it
didn't
work
in
the
right
way
it's
before
it's,
it
was
trying
to
do
a
deploy
rather
than
in
a
build.
So
that's
all
that
was
about
so
now.
A
A
There
we
go
now
all
right
now
we
have
everything
all
set
up
so
now.
Let's
try
this
one
more
time.
Redeliver!
Yes!
Now
it
should
be
green.
Ok!
So
what
it's
saying
here
now?
Oh,
we
got
a
403.
That
means
that
there
is
some
interesting
off
issues
here.
So
let's
go
diagnose,
diagnose
what
that
off
issue
is
so
if
we
go
into
our
trigger
deploy.
A
What
is
happening
here
all
right?
Let's
try
this
one
more
time:
let's
try
it
with
an
edit
this
time,
so
we'll
just
edit
and
we'll
add
another.
It
will
add
another
exclamation
point
to
show
you
how
much
I
really
want
this
to
work.
So
now,
if
I
go
to
my
Tecton
dashboard,
we
have
my
build
pipeline
and
and
if
we
go
back
into
our
settings-
and
we
look
at
our
web
hooks
as
I'm
unhappy-
and
why
is
it
unhappy?
It
says:
I
got
a
403,
it
says
eventlistener
build
Lister
event,
ID.
Well,
that's
very
interesting.
A
All
right
so
I'm
gonna
skip
past
this,
because
there's
something
else
I
wanted
to
talk
about,
but
I
promise
you
I
have
a
solution
for
this,
a
solution
that
actually
makes
this
a
lot
easier.
So
let
me,
let
me
show
you
something
else
here
and
another
idea
that
we
actually
have
so
we
have
build.
There
is
something
weird
with
the
build
and
and
I
think
I
know
what
it
is,
but
I
don't
want
to
dig
into
it,
because
I
actually
have
a
better
solution
that
will
supersede
that
we
also
want
to
do
a
deploy.
A
A
Get
ops
requires
us
to
describe
and
observe
systems
or
the
clarifications
which
eventually
form
the
basis
of
continuous
everything.
So
what
exactly
is
good
ops
again,
I,
don't
know,
but
I
I'm
going
to
ask
question,
is
get-ups
a
methodology
or
is
it
a
thing
that
we
do?
Is
it
a
way
that
we
think
or
is
it
something
that
we're
trying
to
get
done?
I'm
going
to
leave
these
kinds
of
words
to
others,
and
what
I'm
going
to
say
is
that
get
OPS's?
A
We
store
our
configurations
and
our
source
code
repositories,
so
we
conversion
them
like
our
software
and
that's
it
I'm
sure.
People
will
probably
there'll
be
some
debate
and
somebody
will
challenge
me,
but
practically
that's
what
it
is,
and
it's
called
git
ops
because
we're
we're
using
a
lot
of
get
these
days,
so
we're
going
to
store
our
configurations
and
get
so
but
first
before
we
can
do
that.
We
actually
need
to
have
a
configuration
that
we
can
store
and
get
so,
let's
create
one.
A
So
the
first
thing
we
need
to
do
is
make
a
directory
I'm,
gonna
use,
customize
I,
don't
need
to
customize
all
the
time.
I
use
a
few
different
things,
but
I
decide
to
use
customize,
because
actually
it's
pretty
good
for
this
example.
So
customize
is
interesting.
It
has
this
a
customization
file
and
you
can
apply,
you
can
have
resources
and
you
can
apply
transformations
to
the
resources
and
and
then
it
can
spit
out
Yambol
that
you
can
apply
with
cube
control
or
something
else.
A
So
today
we're
just
going
to
stick
and
keep
control,
and
so
first
thing
you
want
to
do
is
initialize
our
customize.
So
I
have
like
C
customized
version,
know
version
I
have
3
something
on
my
machine.
Oh
so
interesting,
I'm
going
to
I'm
being
challenged
by
Cornelia,
Davis,
saying
the
get
and
storing
config
and
it
it's
just
a
small
part
of
get
ops.
A
Hi
Cornelia,
it's
nice,
seeing
you
and
I
know
I
just
challenged.
Cornelia
works
at
we've
worked.
The
inventors
of
get
ops
and
I
am
definitely
being
snarky
and
tongue-in-cheek.
Here,
I'm,
not
dismissing
any
bodies
of
work.
I
actually
do
think
there's
a
lot
more
to
it,
but
I
thought
it
was
funny
so,
but
really
not
trying
to
offend
anybody
here.
A
But
we're
gonna
do
a
tiny
little
piece
of
get
ops
today
and
we're
going
to
store
configurations
and
get
and
then
we're
going
to
set
them
up
to
deploy
automatically
so,
hopefully
I
do
not
offend
and
if
I
did
send
an
email
to
I'm.
Actually
just
just
tweet
at
me,
angrily
on
Twitter
I
will
apologize
profusely
there.
So
first
thing
I'm
gonna:
do
we
have
customized
3.5
dot
four,
so
we're
going
to
actually
do
customized
create
and
what
does
it
do?
A
It
creates
a
file
called
customization
and
it
just
gives
you
an
API
in
an
API
version
and
a
kind
smart
thing
here
that
I
want
encourage
everyone
to
do.
If
you
create
a
configuration
file
and
you
put
stuff
in
it
version
it
because
you're
gonna
change
your
mind
or
someone's
gonna
change.
Something
and
what
you're
gonna
have
is
configuration
that
you
don't
know
when
it
was
created.
So
breaking
your
configurations,
we're
gonna,
put
some
stuff
in
here.
A
So
the
first
thing
we're
going
to
do
is
we're
going
to
do
K,
create
deployment
and
I
already
have
this
command,
so
I'll
typed
out.
So
this
is
how
you
create
a
deployment
from
the
command
line,
so
we're
going
to
create
a
deployment
called.
Let's
say
it's
images.
Let
me
call
Brian
L,
slash
web
say
we're.
Gonna
do
dry
run,
so
it
doesn't
create
a
new
cluster
and
then
we're
gonna
pipe
that
to
ya
Mille
and
we're
gonna.
Put
this
in
a
file
called
deployment
yeah
mo.
A
We
can
also
do
the
same
thing
with
the
service,
so
we're
gonna
create
a
service.
A
cluster
IP
service
called
web
say,
and
it's
gonna
have
TCP
it's
gonna
map,
port
1323
and
the
container
support
80
on
the
service
and
dry
run
piped
again,
we'll
type
it
to
the
service
mo
file.
I
see.
Do
we
need
any
other
files
here
now
we
won't
use
any
other
files.
A
So
now
what
we
want
to
do
is
we
want
to
go
to
our
let's
see
we'll
go
to
our
customization
file
and
what
we'll
do
is
why
resources,
resources
and
we'll
put
our
deployment,
dot,
yeah
and
we'll
use
our
service,
animal
and
what's
gonna
happen
here
is
customize
is
going
to
do
all
the
smart
things
and
it's
going
to
build
our
configurations
so
combine
all
those
yeah
mold
files,
and
it
did
that
you're
thinking,
yeah
Brian,
that's
amazing!
I
can
use
cat
to
do
this.
A
Yes,
yes,
you
could
have,
but
I'm
gonna
show
you
something
neat
about
customize
now.
So
first
thing
we're
going
to
do
with
our
customized
configuration
is
we're
going
to
get
out
of
here
we're
not
going
to
use
the
editor
anymore.
What
we're
going
to
do
now
is
use
the
command
line.
So
this
is
a
very
interesting
thing
that
I
can
deal
with
customize
I
can
set
an
image
right
from
the
command
line,
and
the
first
thing
we
need
to
do
is
actually
go:
get
an
image
from
docker
hub
and
we'll
take
this
wine,
hello.
A
A
Oh,
so,
actually
that's
interesting,
I'm
reading
through
the
comments
here,
so
it
says
Brian
L.
So
that's,
what's
a
miss,
make
stress
right
image.
So
the
comment
was
it
says
if
you
haven't
used
it
before
before
before
use
customize
edit
fix
it's
a
hidden
gem,
so
we'll
Tet,
our
customization
that
we
just
put
in
there
and
we'll
do
customize
edit
fix
and
didn't
save
it.
It
did
me
thing
what
it
do
and
do
anything
or
maybe
it
move.
Oh
it
probably
unrig
ripped
my
file.
A
So
that's
an
interesting
thing
about
about
nesting,
yeah,
Mille,
I,
don't
think,
there's
a
standard
and
depending
on
the
editor
and
I
I'm
in
sometimes
it
blow.
If
I
go
down
to
a
next
level,
it'll
indent
over
again,
sometimes
it
won't
and
I'm
not
here
to
fight
editors,
sub
I
might
usually
go
with
it
because
it's
easier,
but
no
that
is
pretty
cool
though.
So
what
did
we
just
do?
We?
A
We
set
this
image
so
now,
when
I
do
a
customized
build
notice
that
I
didn't
it
actually
replaced
the
image
right
in
my
right
in
my
in
my
deployment,
and
it
knows
how
to
do
that.
So,
if
I
did
a
que,
apply
hey
instead
of
an
F
so
play
K
applied,
so
cute
control
apply
K
for
customize
and
in
this
directory,
what
it's
going
to
do
is
is
going
to
say
that
you
can't
change
that.
That's
the
appointment,
that's
because
I
already
have
one
deployed
and
I
deployed
it
from
a
different
way.
A
Okay,
delete
deployment,
let's
say
all
right!
Now:
let's
try
it
see
you!
This
works!
Yes,
so
now
we
have
this
thing
running
in
our
cluster.
If
I
wanted
to
see
it,
I
can
actually
go
back
to
actin
and
use
a
feature
that
I
was
using
before,
but
I
use
it
again,
so
we'll
just
go
and
so
I
don't
have
to
configure
I
want
to
configure
it
in
grass
or
anything
like
that.
Oh
this
doesn't
do
target
ports,
alright,
never
mind
we're
not
going
to
see
this
one.
Just
trust
me
is
running.
A
Actually
it
says
it's
running.
Yes,
it
started
the
container,
so
you
want
to
trust
me.
It
says
it's
running,
so
what
we
want
to
do
is
we
want
to
have
automate
it
deploys.
So
how
can
we
have
automated
deploys?
So
first
thing
I
want
to
do
is
I
want
to
this.
Is
my
configuration
there's
a
lot
more
I
can
do
with
customize
there's
a
lot
more
transform,
as
I
could
run.
I
can
actually
do
some
labeling
and
some
namespace
stuff.
Not
gonna.
Do
any
of
that
for
time.
A
I'm
gonna
get
ad
see
you
later
believe,
I'm
gonna,
add
everything
I'm
gonna
get
check
in
and
I'm
going
to
check
it
in
yellow
Yolo,
yeah,
very
funky
for
Brian
and
typos,
and
then
what
I'm
gonna
do
is
I'm
at
the
hub,
create
and
I'm
going
to
use
this
hub
command
if
you're
not
familiar
with
the
hub
command.
I
use
it
for
a
lot
of
github
things
for
creating
issues
and
a
lot
of
other
things,
but
it
creates
a
crazier
posit
ory.
A
So
I
created
this
repository
and
then
I
can
do
git
push
that
you
originally
to
create
this
branch
remote
and
by
the
time
I
get
there
it's
here.
So
what
I
could
do
here
is
I
can
run
through
this
all
again
and
I
could
actually
create
a
pipeline
and
create
and
then
I
can
create
triggers
and
I
can
create
that
webhook
and
I
can
make
it
all
work.
But
really
you
know
what
this
is,
what
I'm
typing
and
I
just
don't
want
to
type
anymore,
so
I'm
gonna
do
something
different.
A
So
now
I'm
just
thinking
here,
I
can
do
this
this
morning.
I
was
thinking
about
this
problem
and
I'm
actually
going
to
I'll,
come
back
in
a
couple
months
and
actually
do
a
whole
tea
tik
on
this.
Let
me
know
if
you
would
like
to
see
a
tea
GI
k
on
writing
operators
for
kubernetes
Eva,
either
in
the
chat
or
on
Twitter
or
on
LinkedIn,
or,
if
you
know
me,
on
Facebook
or
any
of
the
other
back
channels
that
we
may
hang
in
together.
A
Just
let
me
know
I
sat
and
thought
about
this
problem
where
I
was
like
alright,
so
to
get
a
deployed
going
or
to
get
up
to
get
a
build
going.
What
do
I
need?
I
need.
My
see,
map
configuration.
That's
like
six
steps.
I
need
to
have
all
these
paces
I
need
to
know
the
repository
need
all
this
and
I
said.
You
know
that's
a
lot
and
you
get
a
deploy
going.
What
else
do
I
need
to
know?
Wow?
That's
a
lot,
and
you
know
me,
experts
not
really
I
struggle
with
this.
A
So
I
was
thinking
well.
Why
don't
we
use
kubernetes
for
what
it's
good
for
to
actually
solve
this
problem
and
here's?
What
kubernetes
is
good
for
it's
a
platform
for
solving
for
making
platforms?
So
today
this
is
actually.
This
is
not
production
quality.
This
is
nothing
else,
but
I
want
to
show
you
how
easy
it
is
to
use
an
operator
to
do
all
these
things
that
I
did
and
the
benefit
is
when
we
use
the
operators
I
don't
have
to
mess
up.
A
A
Brian
I'm
looking
for
a
window
in
my
other
window,
so
give
me
a
second,
so
I
could
find
it
there.
It
is
there,
it
is
I
use
the
IntelliJ
stuff.
The
whole
IntelliJ
stack,
I
like
it
so
much
I
paid
I,
don't
know
much
I
money
to
pay
a
year
like
a
hundred
and
fifty
bucks
a
year
to
use
all
their
products
and
you're
saying.
Why
would
you
do
that
Brian,
because
it
makes
me
way
more
than
150
I
probably
get
10
2:22.
A
So
really,
let
me
show
you
this
and
I'll
make
it
bigger
there.
We
go
I
wanted
to
just
have
this
I
wanted
to
have
a
CRD
that
allowed
me
to
specify
a
service
account
name
specify
a
repo,
it's
public.
If
it's
private,
we
got
it.
We
can
do
some
other
things
we
can
do
with
the
service
account
name
and
I
wanted
to
run
in
this
situation,
where
I
have
an
ingress
rather
than
like
an
HTTP
proxy
from
Contour,
or
something
like
that
and
I
wanted
to
set
it
up.
A
So
what
I'm
going
to
do
is
I'm
actually
just
gonna
run
this
thing,
I'm
going
to
apply
this
thing
to
my
cluster
and
just
to
show
you
how
lazy
I
am
I'm
running
this
controller
in
local
mode,
so
the
controller
is
running
on
my
computer
and
it
is
listening
to
the
API
and
New
Jersey
and
and
dy,
and
do
my
C
3.
So
if
I
do
que,
apply,
dash,
F
and
I
will
config
samples
and
image.
A
What's
let
me
like
that
blaze
is
my
Big
Sam
holes
and
image
by
animal,
because
I
want
to
see
it
from
getting
alright.
So
now,
I've
created
this
thing
in
my
cluster,
you
see
scrolling
things
are
happening
over
here.
Even
if
it's
too
tiny,
don't
worry
about
it.
That's
not
important,
but
let
me
show
you
what
this
thing
does
and
and
you're
gonna
I'll
tell
you
why
I
think
this
way-
and
this
is
why
I
love
kubernetes
I,
really
love
the
idea
of
kubernetes
I
love
communities.
So
what
I
did
here?
A
So
if
I
go
to
images
like
literally
look
at
this,
look
at
this
name,
its
images,
op
X
and
that's
like
operational
experiments
from
from
my
personal
domain.
I'll
click
on
this
thing
called
web
seg
and
I'll
go
to
the
gamal.
This
is
what
we
just
saw.
I'll
go
to
this
resource
viewer
and
what
we
see
over
here
is
there's
a
lot
of
things
going
on.
A
A
It
created
this
web
hook
and
it
says
that
it's
failing
right
now,
but
I
bet
if
I
push
it
again,
it'll
work.
If
not,
that
means
I
just
broke
the
ingress,
and
all
this
is
for
nothing.
Oh
there
we
go
so
create
a
web
hook.
It
actually
manages
the
web
hook
in
the
cluster,
all
right
at
github,
for
you
and
then
what
does
it
do?
A
A
Well,
so,
well
now
create
this
bad
boy
in
cluster.
Look
mom,
no
errors,
no
stack
traces,
but
now,
if
I
go
to
octant
once
again
and
I
go
look
at
oh
there,
it
is
deploy,
stop
X
and
I'll.
Wait
for
that
to
load
and
I
will
click
on
this
one
and
I'll
click
on
this
resource
viewer
still
more
things
in
there
the
Arab
we're
having
a
forest
because
we
we
flew
barred
the
we
flew
power,
I
fubar,
the
ingress
before
that's
on
me.
So
not
a
big
deal.
A
A
We
know
that
it's
actually
their
fault
and
not
our
fault,
that
it
broke
so
kind
of
cool.
So
if
we
go
look,
I
just
want
to
see
we'll
go.
Look
at
my
Tecton,
and
this
is
my
website.
Deploy
run
it.
Actually,
it
runs
in
like
10
seconds
it's
not
very
slow,
so
what
it
did
is
it
ran
my
cute
control.
So
if
I
go
look
in
my
cluster,
so
okay
get
Todd.
What
we'll
see
now
is
we
have
this
website?
A
That's
been
running
for
oh
there,
it
is
it's
website
deploy,
run,
that's
been
running
and
somewhere
and
my
cluster.
We
have
this
thing.
Oh
there
it
is,
it's
actually
don't
know
where
it
is:
okay,
yeah,
it's
appointment.
Let's
say
all
right
so
and
if
I
do
okay
and
it
actually
want
to
go
fix
this
ingress
issue,
it's
bothering
me
so
if
we
go
figure
out
what's
wrong
in
our
ingress,
so
what
it
did
I.
A
A
The
weird
part
about
this
is
that
go
is
super
wordy
and
I
and
I
really
do
agree
people
whenever
they
say
it's
super
wordy,
but
that
wording
this
gets
me
allows
me
to
sleep
at
night,
so
I'll
take
a
couple
extra
lines
of
typing
if
that's
what
it
takes.
So
how
do
we
move
forward
from
here?
And
so
why
did
I
start
thinking
about
this
way?
This
is
what
I
think
about
this.
I.
Think
that,
as
I
keep
on
saying
that
kubernetes
is
a
platform
for
platforms.
We
should
start.
A
A
My
I
do
know
that
from
people
who
thought
they
can
lift
their
data
centers
and
move
them
to
the
cloud
that
didn't
work,
they
had
to
change
the
way
they
fought
and
with
kubernetes.
You
get
a
lot
of
benefits,
with
changing
the
way
that
you
think
about
approaching
infrastructure.
So,
as
a
as
a
you
know,
as
an
ops
team
or
as
a
dev
team
who's
supporting
things
that
run
in
kubernetes,
why
are
you
now
you
understand
that
your
things
will
ultimately
create
pods
and
services
and
ingresses
and
different
types
of
volumes?
A
But
do
you
have
to
work
at
that
level?
I,
don't
know
in
a
lot
of
cases,
probably
not
so
what
you
should
do
is
make
it
easier
on
yourself
and
then
you
make
it
easier
on.
Everyone
else
is
that
you
create
higher
level
abstractions.
So
it's
it
is
all
yam
low
end
of
the
day
and
we
can
actually
dive
through
we're,
not
hiding
any
of
that
yeah
mo.
A
But
what
we
need
to
do
is
to
say
that
for
someone
who
wants
to
build
an
image
on
kubernetes,
they
can
use
an
automated
service
like
Kay
pack
from
the
CMB
team
that
actually
can
automatically
build
images
or
they
can
do
it
in
this
imperative
way.
Actually,
it's
an
imperative
declarative
way.
They
can
do
that
as
well,
but
we
want
to
meet
people
where
they
are
rather
than
delivering
on
these
crazy
ideas
that
don't
work
with
anything
else,
but
we
also
don't
want
to
bring
out.
A
We
don't
bring
in
so
many
Lola
ideas
that
they
don't
really
it's
too
hard
to
actually
figure
out
how
to
put
them
together.
So
there's
the
the
winning
spot
is
in
the
middle
now.
What
does
that
mean?
I,
don't
know
we
need
to
figure
that
out
and
hopefully
that
what
I
was
able
to
show
you
and
if
anyone
wants
to
see
this
code,
just
paying
me
I'm,
not
going
I'm
really
published
it.
But
if
you
really
want
to
see
it,
let
me
know:
I'll
make
it
available
to
you.
A
What
I
wanted
to
show
you
is
that
there
are
lots
of
parts
for
building
applications,
even
easy
ones.
Building
an
app
is
filing
doctor
files
or
using
something
a
little
cloud
native
built
packs
or
using
something
else,
because
you
know
what
you
can
use
em
code
from
the
Google
project,
Scott
Nichols
was
on
earlier
and
I
know
Matt
Moore
from
the
connect
a
native
team
that
they
really
have
thought
about.
Building.
A
How
do
we
build
things
so
use
code,
use
cloud
native,
build
packs,
use
some
use
doctor
files,
but
think
about
how
do
we
make
it
easier,
more
accessible
and
then
what
we
need
to
do
is
we
need
to
understand
how
things
work
and
we
can
have
a
dive
in,
but
we
need
to
automate
what
we
understand.
So,
if
you're
doing
something
and
it's
prone
to
breaking
do
hand
that
to
the
computer,
computers
should
do
that
kind
of
stuff.
Humans
are
fat-finger.
This
whole
entire
thing.
A
The
next
piece
is
that
we
need
to
understand
parts
of
our
domain,
so
there's
a
reason
why
I
split
up
bills
from
deploy,
they're,
very
simple
and
they
map
you
know
loosely
the
CI
and
CD
I
might
even
talk
about
testing
or
multiple
environments
or
anything
like
that.
But
I
could
actually
create
another
deploy
and
actually
yeah
I
could
I
could
create
another
deploy
and
I
can
just
yank.
This
and
I
could
put
this
here
and
I
can
go
like
this,
and
I
can
say.
Call
this
one.
A
A
We
create
don't
create
big
things,
create
small
composable
things
and
then
take
those
small
composable
things
and
create
things
that
are
slightly
bigger.
We
have
this
idea
that
we
jump
from
small
to
super
large
and
what
happens
is
that
we
get
this
thing.
This
looks
very
cool,
but
at
the
end
of
the
day
you
can't
use
it
so
with
that
I'm,
tired
talking
and
it's
hot
in
here
under
these
lights,
so
happy
Friday
to
you
all.
Actually
I
would
love
your
feedback
on
any
of
this
through
email.