►
From YouTube: Tutorial: Everything You Need To Become a GitOps Ninja - Alex Collins & Alexander Matyushentsev
Description
Join us for Kubernetes Forums Seoul, Sydney, Bengaluru and Delhi - learn more at kubecon.io
Don't miss KubeCon + CloudNativeCon 2020 events in Amsterdam March 30 - April 2, Shanghai July 28-30 and Boston November 17-20! Learn more at kubecon.io. The conference features presentations from developers and end users of Kubernetes, Prometheus, Envoy, and all of the other CNCF-hosted projects
Tutorial: Everything You Need To Become a GitOps Ninja - Alex Collins & Alexander Matyushentsev, Intuit (Limited Available Seating; First-Come, First-Served Basis)
Please bring your laptop fully charged as we will have limited charging stations available in the room. A deep-dive on GitOps which will help you, even if you only have minimal GitOps experience, to get a total understanding of everything GitOps. Firstly you’ll get an introduction into what is GitOps and it’s key benefits, then we’ll walk through foundational techniques, such as tools and strategies, then we’ll take it to the next level with advanced techniques and best practices. Finally, you’ll get a chance to get your hands dirty with an accelerated GitOps lab. Parts: - Introduction - Foundations - Advanced Techniques - Hands On Argo CD Lab Prerequisites: Audience members should be familiar with core Kubernetes concepts, as well as comfortable using Git. Those interested in the lab should have a laptop with git and minikube installed. What to Bring: A laptop with Git installed. Optionally, you may wish to install Docker & Kubernetes (e.g. Docker for Desktop).
https://sched.co/Uaee
A
I'm
Alex
in
this
int
shirt
is
also
Alex.
We've
got
to
keep
things
really
simple.
We've
got
a
few
of
our
friends
around
here.
Now
they
all
have
their
own
names,
but
if
you
want
to
just
call
them
Alex
that
will
work
really
well
Alex
over
here,
alex
is
over
here
called
so
today,
we're
going
to
do
a
get
ops
tutorial
and
what
we've
found
are.
The
dry
runs
is
a
couple
of
things
that
are
really
useful
to
do
before.
A
A
A
If
you're
going
to
do,
the
hands-on
are
customized
and
get
I
mean
you'll
get
a
lot
out
of
this
session.
If
you
don't
do
the
hands-on,
but
you'll
get
even
more
out
of
it.
If
you
do
do
it,
okay,
great,
if
you
stood
at
the
back
I'm
just
going
to
reiterate,
come
to
the
front.
There's
plenty
of
places
to
sit
down,
there's
also
plugs
everywhere,
as
well.
Okay,
just
a
little
bit.
Ok,
so
just
I'll
go
back
a
second
I've
left
a
link
down
here
in
the
bottom
right
hand,
side
of
the
screen
here.
A
If
you
that's
going
to
stay
up
the
first
few
slides,
if
you
want
to
go
back
and
find
those
instructions,
click
on
that
link,
that'll
help
you
out
find
those
ones
and
we
just
that'll,
just
be
up
there
for
first
10
or
so
slides.
Okay.
What
are
we
gonna
do
today?
What
are
you
gonna
learn
today,
so
this
is
a
nice
kind
of
Venn
diagram.
I,
don't
know
what
you
would
call
this,
not
a
Venn
diagram,
we're
going
to
talk
a
little
bit
about
the
what
and
why
of
kin
ops.
A
We
already
want
to
give
you
a
really
fundamental
understanding
of
the
core
principles
of
get
ups
they're
really
useful,
because
once
you
have
a
good
handle
on
those,
then
you'll
be
able
to
abstract
from
them
and
understand
a
number
other
things
about
give
ups.
We're
also
going
to
do
a
hands-on
session,
where
you're
going
to
be
checking
some
code
out
of
get
deploying
that
code
into
a
cluster
that
we
have
set
up,
we're
forming
an
upgrade
on
that.
A
So
like
upgrading,
an
updating,
an
image
tag,
and
then
your
will
believe
some
troubleshooting,
your
of
your
app
when
maybe
it'll
go
wrong.
Maybe
and
also
the
opportunity
for
discussion
there'll
be
a
maybe
we'll
have
a
Q&A
at
the
end,
but
also
if
you've
got
questions
just
you
know
go
back
to
that
slack
channel
and
you
can
obviously
ask
there.
My
colleague,
Danny
over
here
will
be
asked
asking
your
questions.
Also,
maybe
introduce
yourself.
You
know
and
I'm
Alec
someone
from
Intuit.
A
You
could
maybe
introduce
your
stuff
and
say
who
you
are
and
where
you're
from
okay
moving
on
okay,
a
little
bit
about
who
the
hell
are
these
guys
now?
Why
should
I
care
Alex
I'm
a
principal
engineer
into
it?
I
work
on
into
its
modern
sass
platform,
which
is
around
a
two-year-old
kubernetes
platform
that
we're
working
on
previously
I
worked
as
the
EMEA
architects
in
on
QuickBooks
Online
in
Europe.
Anybody
here
use
QuickBooks
TurboTax.
A
B
Is
all
you
want
yeah
so
as
that
ex
mentioned
I'm
Alex
as
well,
and
we're
working
on
the
same
team,
we're
part
of
modern
says
not
for
Jim,
but
we
basically
working
on
our
CD
and
even
before
into
it.
I've
worked
on
Argo
as
well
at
a
different
company
which
was
acquired
by
into
it
and,
as
you
can
see
on
the
slide,
there
is
a
link
to
a
book
vision
co-authoring
with
a
few,
my
colleagues.
So
there
is
a
discount.
B
If
you
use
it,
you
cannot
I'm
sorry,
okay,
okay,
this
made
much
better
as
ice.
Let
me
start
from
beginning
so
I'm
my
name
is
Alex
as
well,
and
both
axis
I'm
working
on
are
BCG
projects
even
before
working
for
into
it.
I
used
to
work
on
our
go
at
a
different
company,
so
I'm
kind
of
very
experienced
in
that
area.
You
can
blame
me
for
any
bugs
you
see
and
awesome
so
I'm
after
as
well
I
have
blocks,
and
there
is
a
book
if
you
use
that
link.
B
A
A
Okay,
can
you
hear
me?
Okay,
cool,
okay,
okay,
I
also
want
to
get
a
bit
of
show
of
hands.
I
want
to
kind
of
get
a
bit
of
a
feel
for
how
well
people
know
kubernetes
and
get
offs.
So
I'm
gonna
get
you
to
do
one
of
these
hands
up
things.
I
know
you
guys
hate
it
get
your
hands
up.
Have
you
out
of
kubernetes
yeah?
If
you
haven't
I
mean
you're,
really
the
wrong
conference?
Aren't
you
okay
and
you've
you
out
of
get
ya,
get?
Have
you
heard
of
get
ops?
Okay,
okay!
A
Have
you
who
hears
you
see:
I
DS
who's,
written
a
CR
D.
Let's
get
the
level
know
a
few
people
written
CIT,
okay,
who
designed
the
CID
spec
anybody
here
there:
okay,
nobody!
Anybody
here,
Joe
barrier,
no,
okay,
fine!
Okay!
Thank
you
very
much
for
that.
That's
good
I
think
we've
got
the
right
level
for
the
workshop.
Okay
just
want
to
give
you
a
little
bit
of
information
about
into
it
here
so
into
it's
been
on
a
kubernetes
journey
for
around
18
months
to
two
years.
It's
a
pretty
large
organization.
A
We
have
around
now.
Twelve
hundred
developers
using
kubernetes
on
a
daily
basis,
we
run
16
160
clusters,
which
are
all
managed
clusters,
and
we
have
around
60
mm
pods
and
we
do
about
1,300
deployments
a
day.
This
graph
down
here
will
just
kind
of
show
you
that
the
accelerated
line
of
adoption
of
kubernetes
it
into
its
every
time
I
do
any
kind
of
presentation
about
about
stuff
we're
doing.
I
have
to
update
this
slide.
It
always
goes
up
by
10
to
20
percent.
So
you
can
never
reuse
this
for
flight.
A
We're
gonna
talk
a
bit
about
Argo
shortly
in
this
presentation,
but
we
also
do
a
number
of
open
source
projects
Admiral,
which
is
a
plug-in
for
sto.
We
also
do
CACO,
which
is
a
selection
of
cluster
management
tools
for
managing
nodes,
and
if
you
want
hear
a
bit
about
that,
Sri
can
answer
your
questions
and
we
also
do
Argo,
which
is
a
selection
of
tools
for
kubernetes.
A
The
probably
the
most
well
known
is
our
go
workflows,
which
is
a
workflow
engine
which
is
very
popular
in
the
ml
community
and
some
quite
interesting
conversations
with
people
doing
weather
forecasting
using
at
this
conference
and
an
Argo
CD,
which
is
AG,
adopts
a
tool
for
forged
ops
and
also
a
new
one
called
Argo
rollouts,
which
some
of
you
may
have
seen
a
presentation
on
that
yesterday
morning.
I
hope
you
enjoyed
that
okay.
Well,
what
is
get
ops?
A
It
always
takes
me
a
couple
of
moments
to
think
about
how
to
describe
get
ops
to
an
audience,
because
you
can
always
describe
pretty
much
anything
in
one
or
two
ways
you
can
describe
it
as
though,
as
the
what
and
what
is
the
implementation,
and
you
can
talk
a
bit
about
them
about
the-
why?
Why
would
I
want
to
do
that?
And
those
are
both
kind
of
good
ways
to
describe
many
kind
of
concepts?
What
get-ups
really
boils
down
to?
A
Is
you
describe
your
software
system
in
git,
so
you
basically
put
all
your
files
in
to
get
and
that
describes
your
software
system
because
we
didn't
cloud
native.
Your
software
system
is
Cuban
Eddie's,
so
those
files
are
going
to
be
giving
Enys
manifests,
and
then
you
have
a
second.
You
have
a
piece
of
software
called
an
agent
which
just
means
a
bit
of
a
software,
and
the
job
of
the
agent
is
to
make
sure
the
live
system,
which
is
your
kubernetes
cluster,
is
the
same
as
your
target
target
state
describing
kit.
A
A
Ok,
so
what
does
that
really
boil
down
to
just
three
key
components?
Git
repository
for
your
manifests,
an
agent,
a
tool
to
get
your
manifests
from
get
into
the
third
thing,
which
is
a
kubernetes
customer,
and
these
are
just
the
really
three
three
key
things
you
need
to
know
about,
but
nothing's
ever
really
that
simple
in
practice
there
are
three
other
things
which
are
really
useful.
We
need
to
get
the
most
value
out
geared
up,
so
you
kind
of
need
these
three
other
things.
The
first
thing
is
a
manifest
generation
tool.
A
So
that's
why
it's
quite
a
good
helpful
to
have
it
installed
on
your
laptop
and
the
second
one
is
the
CI
system.
So
get-ups
is
very
much
a
seedy
process
about
getting
manifests
from
get
into
your
cluster,
but
it
doesn't
really
take
any
any
of
the
responsibility
for
determining
what
goes
into
those
manifests
and
a
really
common
pipe
get
opted
pipe.
Own
actually
starts
with
a
CI
job
with
who
which
builds
the
application
runs.
A
For
example,
you
could
group
them
by
cluster
or
by
namespace
and
the
one
that
we
like
basically
is
grouping
by
label
because
actually
extremely
flexible
and
the
reason
you
probably
want
to
group
your
resources
together
in
kid.
Ops
is
because
you
typically
won't
want
everybody
in
your
organization
to
be
able
to
create
and
delete
every
kind
of
resource
you
might
want
to
restrict
certain
sets
of
resources
to
certain
teams,
and
there
you
know,
there's
no
hard
and
fast
rule
of
rules
around
how
you
do
that.
A
You
know
for
a
small
organization
with
one
thing
you
might
want:
everybody's
have
read/write
access,
but
if
you're
a
large
organization
with
lots
of
automation,
then
you
might
have
a
lot
of
tooling
around
this.
You
might
be
quite
restrictive
in
how
you
allow
people
to
do
even
to
only
allowing
specific
individuals
within
teams
to
do
deployments.
A
Okay,
that's
the
what,
but
what
about
the?
Why?
Why
would
you
want
to
do
get-ups?
So
the
the
primary
reason
you
want
to
use
git
Ops?
Is
it
provides
you
a
single
source
of
truth.
You've
got
all
your
files
version
controlling,
gets
that
everybody
can
read
everywhere.
You
can
see,
rather
than
having
your
file
spread
across
different
people.
Machines
or
you
know
in
some
inaccessible
database
somewhere,
and
it
comes
with
a
lot
of
benefits
of
using
it.
A
So
you
get
all
the
security
you
get,
you
could
control
who
can
read
and
write
to
repositories
and
therefore,
who
can
deploy
applications
that
gives
you
compliance
and
also
you'll
get
an
audit
trail.
You'll
know
by
looking
at
the
get
history
who's
done.
What
and
when
it's
a
very
difficult,
developer-centric
workflow,
so
typically
lawn
developers
are
using
get
to
version
control,
their
application
code,
so
you're
just
using
the
same
tool
again
to
version
control.
Your
deployment
configurations
by
deployment
configuration
I
mean
manifests
one
nice
feature:
is
it
trivializes
rollback?
A
We
hear
this
feedback
a
lot
from
developer
teams
that
it's
just
really
quick
to
get
your
code
out
into
production.
You
get
that
fast,
fast,
feedback,
loop,
okay,
what
we're
going
to
do
now
is
going
to
do
a
quick
demo
and
then
this
code
is
in
github.
You
can
check
out
these
links.
I'm
gonna,
I'm
gonna
bring
it
up
now
in
my
other
window,
we're
gonna
do
poor-man's
get
ops.
So
what
is
poor-man's
get
ops?
Well,
that's!
Basically,
if
you
don't
want
to
install
any
kind
of
application
to
do
field,
ops
were
well
wolf.
A
This
is
what
you
could
do
just
using
standard
kubernetes
resources.
Okay,
in
my
apologies,
you
go
okay.
So
the
first
thing
we
do.
We
obviously
we
need
an
application
to
deploy
and
what
I've
got
here
is
a
very
basic
application.
Called
I've
got
it
in
my
app.
Maybe
I
should
have
got
something
else
and
if
I
just
run
this
application,
we're
gonna
we're
gonna
use
this
again
throughout
the
room
throughout
the
demo.
A
A
So
this
docker
file
regimen
soars
the
very
basic
tools
for
an
agent
installs
get,
so
you
can
get
your
files
from
Kipp
and
I
sort
of
get
ex
library
because
it
contains
a
command
line
tool
called
M,
sub
stuff
and
M
sub
store
is
the
most
basic
table
plating
tool.
I
could
think
of
it.
Basically,
if
you've
got,
if
you
take
standard
in
you
cat
its
own
substr,
and
then
it
replaces
any
dollar
dollar
value
names,
dollar
variable
names
with
environment
variables,
it's
a
very
basic
templating
tool
and
obviously
Cooper
CT
alphas.
A
The
third
component
is
the
ability
to
speak
to
a
kubernetes
cluster
and
then
I've
got
an
installation
gamal
file
here,
and
this
contains
a
couple
of
different
things
as
well.
So
go
a
cluster
role
binding
in
a
service
account
so
that
my
cron
job
can
speak
to
coop.
Ct
also
has
access
readwrite
access
to
those
api's
and
then
just
have
this
con
job
here.
So
I
make
the
font
a
bit
bigger
yeah.
Let's
do
that.
That's,
let's
tab
to
a
different
window
in
status,
so
basically
have
a
cron
job
here.
A
This
is
set
to
schedule
every
one
minute
and
what
it
does
is.
It
runs
this
short
shell
command
here
clones.
This
git
repository
it
cats,
the
content
of
this
yellow
file
through
2mm
substr,
and
then
it
passes
it
to
Clube
CTL
apply,
and
then
we've
got
to
do
a
couple
of
additional
arguments
here
and
including
a
label
and
approve.
So
the
label
is,
you
know,
talked
about
earlier.
They
need
to
kind
of
group
your
resources
together.
A
It's
really
useful,
so
we're
applying
a
label
there
to
do
that
and
then
we'll
delete
any
resources
that
we
that
we
don't
find
that
we
don't
want
using
the
pruning
I've
just
added
validate
faults,
because
that's
what
I
had
to
do
to
get
it
to
work,
don't
really
know
why!
So
just
so
this
is
this
is
in.
If
you
wanna
run
this
local
yourself.
This
is
just
in
our
github
get-ups
workshop
on
github
and
maybe
Danny
can
drop
a
link
in
maybe
go
for
it
Fanning's
on
Facebook
as
well.
A
Okay,
so
here
we
go,
we're
just
create
our
resources
and
we
can
just
watch
this
on
the
console.
I
think
we
can
watch
this
on
the
console
there
we
go,
and
so
it
runs
about
once
a
minute.
So
we
just
got
to
wait
for
it
to
run
and
when
it
runs,
it
will
run
that
cron
job
and
it
will
make
me
try
and
fill
time
and
try
and
avoid
any
dead
air
there.
It
goes
all
right
guys.
It's
basically
just
runs
across
your
bar.
B
A
The
problems
with
poor
man's
get
ups,
okay,
let's
just
go
through
five
problems
with
poor
man's
get
ups,
and
this
is
this-
may
be
sound
a
bit
salesy,
but
this
this
stuff
really
applies
to
any
kind
of
tools
and
we're
going
to
talk
about
some
other
tools
you
can
use
shortly,
and
one
of
the
problems
is
security.
Obviously
my
cron
job
has
readwrite
access
to
the
cluster.
Therefore,
my
able
you
can
commit
to
get
has
basically
read
about
access.
A
A
If
those
manifest
is
going
to
validates
us
that
there's
none
of
that
kind
of
features
in
there,
you
might
have
quite
a
big
impact,
so
I
might
exit
I
can
easily
make
kind
of
quite
a
big
mistake
using
this
this
tool
and
it
has
latency
so
as
I
showed
you
there
that
that
cron
job
only
kicks
off
every
minutes.
You
know
you
might
not
wanna
configure
in
a
minute
in
production
because
you
might
not
be
doing
your
deployment
every
minute.
A
You
might
have
three
minutes
or
five
minutes,
but
you've
got
basically
wait
for
that
cron
job.
To
kick
in
to
do
that
deployment
and
the
final
thing
is:
it's:
it's
been
essentially
a
single
tenancy,
so
there
is
only
really
one
team
in
this
in
this
model
and
we're
all
deploying
to
the
same
cluster.
So
we
can't
deploy
across
clusters
across
NATO
or
easily
across
namespaces
and
other
things
like
that.
So
these
are
kind
of
the
some
of
the
some
of
the
downsides
and
we're
gonna.
Perhaps
even
talk
a
bit
more
about
this
later
on.
A
A
So
double
you
guys
seen
the
blog
posts
announcement
recently
yeah,
so
we're
gonna
be
working
with
those
guys
to
produce
and
I
think
that
get-ups
engine,
which
would
be
a
kind
of
a
best
in
breed,
get
up
solution
and
we
both
have
two
products
and
they
have
slightly
different
use
cases.
I
won't
really
go
into
the
details,
but
we're
not
we're
not
experts
on
flux,
we're
experts
on
Argos
CD
and
that's
what
we're
gonna
show
you
today,
so
I'm
just
going
to
give
you
a
little
quick
that
we
can
add
a
little
oh
yeah.
A
So
so
what
does
I'll
go?
See?
I'll
go
see!
This
is
handsome
baby
zoidberg
here
who's
our
mascot,
and
he
basically
just
takes
that
role
of
the
agent
with
inside
your
with
inside
your
kid
ops
solution
here
are
a
few
of
the
names
in
the
Argo
community,
so
we've
got
kind
of
some
big
names
up
there.
Some
smaller
names
and
we've
met
some
people
here
today
who
aren't
listed
here.
So
if
your
are
using
Argo-
and
you
want
listed
here-
love
to
hear
I
love
to
hear
have
to
get
you
on
the
readme
file.
A
If
you
can
do
that-
and
you
know,
we've
got
a
couple
and
you
can
see
you've
got,
we've
got
nearly
2,000
stars
in
this
repository
now
saying
to
my
colleague
Danny
you
I
wonder
who
can
get
the
most
stars
at
cube,
Khan
on
their
repository,
who
will
get
he'll
get
the
most
new
ones
sitting.
There
challenge
that.
But
if
we
get
two
thousand
stars
into
koukin
I'll
play
you
the
only
animated
gif
you're
gonna
see
at
this
coupon
because
very
popular
anymore,
I.
Think:
okay,
quick,
a
quick
demo.
A
So
let
me
just
make
this
okay,
so
this
is
the
user
interface
of
Argos
CD.
This
shows
you
the
way
that
we
group
APNIC
group
group
resources
in
Argos
senior,
which
is
in
a
thing
called
an
application.
An
application
is
essentially
a
custom
resource
which
defines
which
repository
path
and
revision
to
get
the
manifests
from
and
which
cluster
and
namespace
to
deploy
them
into,
and
then
those
those
resources
when
they
are
deployed
they
get
labeled
with
their
applications
name.
So
we
know
which
resources
make
up
part
of
an
application.
A
I'm
just
going
to
pick
out
certain
manager
here
to
give
it
to
show
you
a
little
bit
around
this
user
in
toast.
Cuz
you're
gonna
see
a
bit
more
of
this
shortly,
but
it
might
be
quite
nice
a
bit
about
this
beforehand.
So
this
shows
you
my
application.
This
is
the
cert
manager
application.
You
can
see
it's
a
number
of
resources
in
there
and
we
can
see
that
some
of
these
resources
have
a
nice
green
tick
to
them.
A
Next
one
which
tells
us
there
what's
in
get
is
the
same
as
what's
in
our
cluster,
which
is
really
important
to
know,
and
some
of
them
have
small,
small,
yellow
X's.
That
tells
us
that
actually,
there
are
some
differences
between
that
and
what
we
expect
and
we
can
click
into
those
resources
and
they'll
actually
show
us
here
in
the
interface.
A
What
on
the
left-hand
side,
what
we've
got
in
our
cluster
and
what
on
the
right-hand
side
is
what
we
what
we
really
want
to
have-
and
this
typically
shows
you
the
difference
between
the
deployed
commit
and
the
commit
that's
at
the
head
of
your
branch-
and
this
is
this
information-
is
also
aggregated
up.
So
at
the
top
here
we
can
see
that
this
application
is
healthy,
which
is
indicated
by
the
green
heart
and
but
it's
out
of
sync
with
git.
A
So
we
know
that
there's
some
changes,
so
maybe
you
want
to
get
application
into
sync
with
with
git,
ok,
the
hands-on,
so
I
hope
this
is
the
bit
you
all
been
waiting
for.
Do
you
have
your
laptop's,
ready,
I,
hope
everybody's
managed
to
get
through
those
instructions?
If
you
do
have,
if
you
have
had
any
issues,
then
you
put
your
hand
up
if
you
haven't
be
able
to
log
into
github,
and
maybe
somebody
can
help
you
out
on
that-
should
I
just
remove
my
laptop
from
here.
Let's
do
that.
B
B
B
My
god,
how
come
it's
all
right?
Let's
get
says
it
so
before
I
cause
that
thing
I
want
to
say
thank
you
to
digital
ocean
because
they
give
us
all
some
kubernetes
cluster,
so
I
want
to
give
them
credit,
and
if
everything
is
not
working,
you
can
buy
into
it's
a
joke,
so
we
taste
think
it
went
well
so
far.
Okay
and
the
next
step,
please
open
that
URL.
B
B
Okay
and
I'm
going
to
open
it
as
well.
So
before
we
jump
into
the
exercise,
this
is
basically
you
can
use
the
depositary
a
later
one,
because
it
has
all
the
steps
which
we
are
going
to
do
in
the
workshop.
So
if
you
start
today
and
something
doesn't
go
right,
this
is
pretty
much
instructions
of
what
you
can
do
to
repeat
this
workshop.
B
Okay,
make
it
bigger.
Thank
you,
I
am
going
to
forget
things.
Please
bear
with
me,
okay
and
I'm,
going
to
give
you
just
a
brief.
You
know
overview
of
what
is
in
the
tip.
Oh,
it's
pretty
basic
because
we're
doing
key
tops,
it
has
bench
of
KML
files
which
I
kubernetes
manifests
and
because
we
want
to
make
it
like
a
real
life,
we
don't
want
to
just
have
just
Yama
files,
we're
going
to
use
some
templating
too,
and
our
choice
right
now
is
customize.
Later
on,
we
will
talk
more
about
helm
and
customizing.
B
What
works
better,
English
kisses
for
now
it's
customized
and
we
try
to
make
it
as
simple
as
possible,
because
it's
a
demo.
So
we
have
a
single
resource
which
is
pot
and
we're
going
to
deploy
it
in
few
minutes
and
we
have
customization
to
tml
file
and
all
it
does.
It
basically
tells
to
customize
that
it's
supposed
to
use
that
port
as
the
resource.
A
B
City,
big,
this
is
giant
whoo.
This
is
my
customized
I
used
t
dot
one
please
install
at
least
three
dot,
something
because
customize
changing
quickly
and
I
cannot
guarantee
it's
going
to
work
on
to
awesome,
so
I
think
we
needed
to
start.
Please
raise
your
hand
if
you
need
some
more
time,
see
no
hands.
So,
let's
get
started,
and
the
very
first
step
which
we
supposed
to
do
is
to
fork
repo
Wow
I
hit
92
fork,
so
I
see
energy
to
false
already.
This
is
good,
but
for
those
who
didn't.
A
B
So
and
the
reason
you
have
to
fork,
it
is
because
we're
going
to
make
changes
in
the
tipo
and
each
of
us
will
do
these
changes
and
basically
you
need
to
write
access
and
we're
not
going
to
give
you
a
password
of
our
repo.
So
we
will
have
to
use
yours
and
awesome,
so
everyone,
forked
I,
hope
I
hope
we
are
not
going
to
get
hundreds
of
works,
so
we
kissing
alright,
so
and
I.
B
Don't
really
need
my
readme
file
to
guess
what
is
the
next
step,
because
next
step
is
to
clone
the
tee
ball
before
you
call,
please
make
sure
you
it's
a
suggestion
to
use
SSH
URL
and
the
reason
is
just
easier
to
push
if
you're
a
github
user.
Maybe
you
have
SSH
key
configure
it
in
github.
If
you
don't,
you
have
to
use
HTTPS
and
use
usual
username
password,
that's
more
work,
but
that
works
too.
So,
okay
I'm
going
to
clone
the
repo.
B
And
great
so
we'll
have
the
repo
and
then
we
think
for
everyone
to
do
that.
Unfortunately,
I
cannot
verify
the
two
clones
very
poor.
There
is
no
information
and
we
are
ready
for
the
next
step.
It's
still
basically
preparation.
I
have
these
shell
scripts.
You
can
use
them
as
a
basically,
you
don't
have
to
type
you
can
just
copy
paste,
and
your
step
is
to
to
use
the
set
this
environment
variable
username.
B
It
is
not
only
not
only
just
to
clone
your
repo,
it
is
also
it
will
be
used
later
on
and
important
piece
here
is
to
use
your
github
username
as
a
username
and
please
use
lowercase
and
the
reason
is
we're
going
to
use
that
same
environment
variable
to
give
a
name
to
a
Cuban
agency
source
and
if
you
might
know,
community
resources
not
allowed
to
have
special
characters
and
uppercase
characters.
So
please
use
your
lower
case.
Username.
A
You
guys
probably
know
that
we're
going
to
do
it.
Sometimes
we
won
vacation
and
what
we
can
treat
this
crate
300
name
spaces
for
you,
so
we're
all
you're
gonna
share,
lovely
big
namespace
together
and
we're
gonna,
give
you
a
bunch
of
unique
names,
your
application
and
we're
gonna
use.
Your
recommend
use
your
github
using
thing
for
that,
because
we
can
pretty
much
guarantee
nobody.
B
B
B
B
Ok,
so
actually
the
next
step
is
to
verify
that.
So,
if
next
command
fails,
customize
is
not
installed,
or
maybe
it's
a
wrong
customized
version.
So
next
step
is
to
use
customize
to
get
the
final
aeon
files
which
you're
going
to
install
onto
your
cluster,
and
the
reason
we
pick
customize
is
because
it
kind
of
it's
useful
to
make
to
customize
the
template
of
your
application.
B
So,
as
I
mentioned
in
the
repo,
we
have
a
resource
and
a
customization
document
and
those
two
files
in
a
base
directory-
and
this
is
the
typical
pattern
which
customize
encourages
developers
to
use,
is
to
create
a
base
template
of
your
application,
which
you
can
customize
later
on,
and
usually
you
want
to
customize
it
to
create
bunch
of
resources
of
so
a
bunch
of
environments
and
in
our
case
each
environment.
Is
you
an
applications
which
is
specific
to
you?
I
want
to
make
sure
my
base
is
actually
working
to
do
that.
B
I
need
to
run
this
customized
build
based
command
and
it
produces
a
mo
file
and,
as
you
can
see
right
now,
it
uses
my
name
so
my
app
as
a
name
of
my
pod
and
it
uses
some
predefined
image
is
a
pretty
fancy
version
with
no
changes
from
the
base
and
next
step
is
to
create
an
environment
and
push
it
back
to
your
fork.
And
you
can
you
don't
have
to
use
that
command,
but
it
will
help
you.
B
B
A
He's
doing
this
just
a
little
bit
about
the
the
layout
of
a
directory
structure
for
a
customized
application.
Typically,
I'll
have
a
base
directory
which
contains
your
kind
of
default
set
of
manifests,
and
then
you
have
a
series
of
overlays
and
overlay.
Basically
is
a
list
of
changes
to
be
made
to
the
base,
manifests
for
different
environments
so
and
there's
a
bit
being
a
director
called
overlays,
slash
name
of
environments.
A
You
might
have
overlay
slash
production
or
uswest,
one
or
QA
or
whatever
you
want
to
over
there,
and
for
that
there
that
to
be
able
to
run
customize
build
in
the
directory.
You
need
have
a
file
called
customization
ya,
know
in
that
directory,
and
that,
basically,
is
the
figuration
configuration
for
that
particular
customization.
And
what
we're
going
to
do.
What
we're
doing
here
is
showing
how
to
create
an
overlay
for
your
development
application
and
it'll
be
customized
for
for
you
individually
yeah.
So
this.
B
Is
just
a
cut
out
put
of
customization
of
the
mo
file
and,
as
you
can
see
it
uses
base
as
a
base
for
my
application
and
next
we
can
just
try
to
build
it
and
see
what
happens
so.
Please
copy/paste
this
customize
build
comment
and
execute
it
in
your
on
your
laptop
and
basically
basically,
you
should
see
exactly
the
same
output
as
output
of
a
previous
base
command
and
the
reason
is
you
made
no
changes
and
we
use
existing
base
directory
as
a
base
for
our
new
customized
application.
B
B
B
A
A
A
A
We
talked
a
bit
about
Hell
and
customize
for
moments,
so
you
guys
bring
noticing
we're
not
we're
not
using
helm
for
this,
and
the
reason
we're
using
customized
predominate
is
because
that's
something
we
use
internally,
partly
because
that's
what
we've
used
the
workshop,
we
find
they
fit
quite
a
quite
different
use.
Cases
and
customize
is
particularly
good
if
you're
consuming
your
own
software
than
your
own
team,
because
it's
very
easy
to
create
manifests
hell
is
actually
excels
at
packaging
applications
up
for
other
to
use.
It
can
be
a
bit
of
personal
preference.
A
B
Mentioned
before
that,
we're
going
to
deploy
an
application
into
the
same
namespace
and
obviously
we
don't
want
to
step
on
each
other
and
delete
resources
or
like
override
resources
of
someone
else.
So
the
next
step
we're
going
to
use
one
of
customized
features,
which
is
name
prefix,
and
it
is
a
very
convenient
for
that
workshop
to
make
sure
everyone
uses
a
unique
name
for
photo
shoots
or
you.
B
B
Yes,
yeah
I
think
it's
time
to
move
forward
and
in
case
you're
really
stuck.
There
is
a
chance
to
repeat
it
again
and
please,
you
know,
feel
free
to
reach
out
to
some
slack
we're
happy
to
help
you
visit
even
after
the
talk
so
I'm
going
to
continue
and
as
I
mentioned
just
now,
I'm
going
to
change
the
name
prefix
of
my
resource
using
customize
and
we're
at
the
step
6
and
as
usual,
you
can
just
copy
paste.
B
B
Interesting,
ok,
we'll
see
you
later,
ok
go
ahead
and
change
the
current
directory
to
overlays
first
death
and
next
step
is
to
change
name
prefix
of
your
resources,
which
you
going
to
deploy.
There
are
two
ways
to
like
multiple
ways
to
do
it.
Actually,
one
way
is
to
just
go
ahead
and
modify
the
AMA
file
or
use
customized
patch
or
use
a
feature
of
customize
code,
name
prefix,
and
you
might
guess
from
the
name
what
it
does.
B
B
B
Okay,
and
if
that
was
successful,
we
can
make
sure
that
customized
build
is
actually
producing
what
we
expected,
so
the
build
should
make
a
slightly
slight
change
because
we
have
just
one
resource.
Just
this
is
just
was
perfect.
This
is
my
username,
but
the
beauty
of
that
feature
is
that
if
you
have
ten
resources,
you
don't
have
to
go
in
when
you
find
all
of
them,
it
will
be
taken
care
by
customize
who,
so,
if
everything
is
going
smooth
so
far
for
you
it's
time
to
commit
the
change
and
push
it
back
to
get
her.
A
B
Yes,
the
reason
is
we
kind
of
quickly
we
didn't
so.
The
question
is:
why
do
we
set
the
board
name
instead
of
just
user
name
space
to
basically
name
space?
Is
it
a
question
right
yeah?
The
reason
is,
it
would
be
like
extremely
difficult
for
us
to
automate
an
Essbase
creation
for
like
all
of
us,
so
we
decided
to
just
use
customized
feature
of
name
prefixing
for
the
demo
purposes,
but
in
a
level
yes,
the
suggestion,
basically
into
it,
we
suggest
to
create
a
separate
namespace
for
each
application.
B
B
B
B
Okay,
I
think
it's
time
to
move
on.
Let's
continue
and
so
far
it's
been
kind
of
boring
every
one
of
you
edited
edit
files
before
and
push
it
back
to
get
I
hope.
Now
it
will
be
more
entertained.
So
it's
time
to
take
what
you've
done
and
deploy
it
into
real
cluster.
And
for
that
workshop
we
did
something
really
brave.
We
created
a
cluster
and
we
gave
everyone
on
github
admin
access
and
if
you
have
a
readme
file
open,
you
can
see
the
instant
URL
or
Fargo
City
instance,
which
we
created
for
that
workshop.
B
Please
go
ahead
and
click
on
it
and
you
need
to
login.
There
are
two
ways.
Basically
you
either
know
the
admin
password
which
we
are
not
going
to
tell
you
or
you
can
just
click
on
login
via
github
URL,
and
you
will
have
to
confirm
that
you
trust
us
and
basically
we'll
know
your
email.
That's
it
I
think
else.
So
if
you
click
authorize,
oh
my
god,
you
should
have
seen
an
empty
screen
here.
We're.
B
Yes,
we
can
yes
likely,
like
a
hundred
people,
already
managed
to
finish
workshop,
or
at
least
get
very
far
there,
but
I'm
going
to
kind
of
reiterate
what
Alex
demoed
a
little
bit
earlier
and
I'm
going
to
walk.
You
step-by-step
and
explain
like
in
a
you
know
more
details.
What
I'll
go
see
the
application
is,
and
what
is
that
you
know
all
these
things
on
this
UI
means.
B
So,
basically,
if
you
want
to
deploy
something
using
our
guchigi
you're
supposed
to
create
an
application
which
is
just
Argosy
abstraction,
it's
a
way
to
grow
a
bunch
of
resources.
You
can
group
your
whole
cluster
into
a
single
application,
or
maybe
you
want
to
group.
You
know
just
set
of
resources
which
you
care
about
and,
as
everyone
did
before,
I
mean
what
we
just
eat.
We
just
created
a
source
of
manifests
for
our
application
and
we're
going
to
try
to
create
deploy
that
set.
B
B
Don't
worry
about
project
for
now.
Usually
project
is
basically
team.
We
are
one
team
right
now
and
please
do
not
destroy
your
teammates
applications.
So
the
sync
policy
is
for
demo
purposes:
I
want
to
set
it
to
manual.
You
can
set
it
to
automatic
if
you
really
believe
it's
not
really
recommended
in
production
and
next
meaningful
input
is
the
source
of
manifests.
So
I
suggest
you
to
use
your
github
repo
URL
and
please
make
sure
to
use
HTTP
URL.
If
you
use
SSH,
you
will
have
to
configure
keys
just
because
SSH
clone
requires
some
keys.
B
It
is
possible,
but
for
the
demo
purposes,
let's
just
use
HTTP
URL
and
next
is
a
cluster.
So
our
guchigi,
this
particular
instance,
is
configured
to
deploy
into
the
same
cluster
where
it
is
running
and
named
Sedna
space
for
simplicity
is
default.
You
can
choose
any
other
namespace,
but
basically
you
will
have
to
guess
the
name
of
the
namespace
and
I'm
just
telling
you
that
default
is
available,
and
this
is
it
that's
all.
You
need
to
deploy
an
application
and
if
everyone
gives
me
I
see
that
everyone
worried
here,
please
click
create
button.
B
So
the
repository
URL
supposed
to
be
in
the
repository
field-
and
there
is
a
path
and
path-
is
just
a
you
know.
Directory
inside
of
Europe,
oh
I'll
go
CG
kind
of
tries
to
help
you
here.
It
searches
all
the
repo
all
the
directories
instead
of
Europe
or
which
a
possible
application.
So
it
understands
customer
has
helmet
a
sonnet,
so
it
notices.
There
are
two
customized
files
in
my
repository
and
please
use
overlays
def,
because
it
is
it's
if
ik
to
you,
it
has
net
name
prefix,
who
now
I'm
ready
to
create
it
and.
B
B
Okay,
and
so,
as
you
can
see
here,
two
of
my
applications
in
so
the
resource
of
my
application
has
this
cross
next
to
it,
which
indicates
that
it
is
something
is
not
right
with
it.
I
can
click
on
it
and
I
can
see
exactly
what
is
different
and
the
difference
is
is
just
missing
in
my
cluster
and
if
you
following
me,
you
should
see
the
same
picture
and
the
way
to
fix
it
is
to
sink
and
sink
is
also
it's
a
basically
a
charm
introduced
in
algo
CG.
B
We
call
deployment
processing,
because
this
is
what
it
is.
Actually
it's
just
thinking
your
github,
sorry,
the
state
defined
and
get
into
the
cluster.
So
please
go
ahead
and
click
sync
button,
and
you
will
see
this
panel,
which
kind
of
showing
you
what
exactly
you're
about
to
do
in
my
case,
I'm,
going
to
push
a
deployment,
which
name
is
the
name
so
push
a
port
which
has
the
name
and
I'm
going
this
resource
was
taking
from
the
positive
from
revision,
wishes
her
division
of
Myra
positive.
B
B
B
At
least
now
I
have
some
feedback.
I
can
see
that
chart
is
still
changing.
That
means
people
keep
clicking
sync
button
and
meanwhile
I
kind
of
wanted
to
point
out
that
I
personally
feel
like
this
is
impressive
already,
even
though
all
we
did,
we
just
deployed
it,
but
basically
I
was
really
impressed.
How
like
hundred
people
managed
to
do
it
without
guidance
and
yeah?
B
B
So
in
addition
to
doing
key
tops,
our
goal
is
pretty
much
a
two
is
that
it's
basically
a
kubernetes
dashboard,
so
you
can
use
it
to
see
what
is
in
your
question
and
you
can
use
it
to
troubleshoot
issues,
and
if
there
are
no
issues,
you
still
can
get
some
useful
information.
You
can
see
what
resources
I
include
it
into
the
application.
So
in
our
case
it's
just
one
port
and
still
you
can
click
on
that
port
and
you
can
get
some
summary
information
about
the
port.
B
Let's
go
back
to
our
chart
and
see
how
many
applications
they
still
and
synced.
Okay
I
feel
like
whoever
not
synced
it.
Maybe
you
know
maybe
stack
and
maybe
need
help,
and
because
that
number
has
not
changed
much.
So
let's
continue.
And
finally,
let's
do
some,
you
know
github
deployment
and
unfortunately,
we'll
have
to
go
back
and
change
the
gate
again
and
it
might
be,
it
might
be
boring,
but
this
is
I
think
sso
key
tops.
B
Basically
it's
very
robust
once
you
make
changes
and
get
you
should
be
able
to
deploy
to
that
stage
and
later
on,
you
can
easily
roll
back.
So,
first
of
all,
let's
deploy.
Let's
make
a
change.
This
is
step
number
nine
and
it
will
be
really
easy
for
you.
If
you
go
back
to
your
repository,
open,
well
and
scroll
down
to
step
number
nine
and
next
we
can
use
another
customize
feature
to
make
changes
in
your
application.
B
So
basically,
as
you
can
imagine
to
upgrade
an
application,
maybe
you
want
to
build.
You
want
to
change
an
image.
I
already
created
an
image
for
that
workshop,
so
you
don't
have
to
build
it,
and
all
you
have
to
do
you
just
have
to
copy
this
batch
command,
run
it
and
around
get
teef
to
make
sure
something
meaningful
happened.
B
It
worked
for
me,
my
customization,
that
Yama
file
got
a
new
section
called
images
and
just
want
to
explain
what
it
is
supposed
to
do.
It
explains
to
customize
that
customize
should
check
every
resource
included
in
to
customize
application
and
what
search
for
images
which
name
is
it
get
absorbed
shop
my
app
and
make
sure
that
image
version
is
version
specified
in
a
new
tag
field?
So,
in
our
case
it's
v2
and
next
step
is
to
verify
that
it
actually
worked
and
image
is
replaced
by
customize.
B
B
So
in
in
the
real
life
algo
cg
would
notice
the
change
already
and
if
I
go
back
to
my
application,
it
is
still
in
sync
and
the
reason
is
usually
we
can
configure
a
web
hook
from
your
repo.
Now
we
have
taken
the
cheapest,
it
will
be
difficult
to
come
to
configure
a
web
hook
so
to
fix
it,
you
can
simply
click
refresh
button
and
tell
to
algo.
She
did
that.
You
know
there
was
a
change.
Please
check
it
now.
If
you
don't
click
refresh,
you
would
have
to
wait
for
like
30
minutes
and
now.
B
B
I
guess
it
would
be
easier
if
I
go
back
to
any
file.
So
if
you
could
not
complete
the
steps
there
on
the
screen,
please
go
ahead.
Do
it
basically
copy
these
batch
commands
make
a
change
in
yogi,
tipo
git,
push
it
back,
and
click
refresh
button
on
your
application
page
to
make
sure
our
go
kind
of
refreshes
to
stay
attend.
Show
you
the
difference.
B
B
All
right,
we
have
limited
time,
so
we
have
to
continue
I
want
to
show
you
how
Argos
ejek
can
help
you
to
understand
what
exactly
is
different
in
your
application,
as
I
mentioned
before,
you
can
just
click
on
resource
and
inspect
the
differences
between
state
defined
and
get
and
then
a
cluster.
So
in
this
case
the
differences
is
image.
Tag
has
expected
and
I
want
to
go
ahead
and
fix
the
difference.
Basically,
I
want
to
sync
again
so
please
click
the
sync
button
and
confirm
the
syncing
on
this
panel.
So.
B
Sync
has
started
and
application
syncs
way
state
switch
to
green,
meaning
it's
in
sync
right
now,
and
we've
got
this
red
mark
indicating
that
application
is
in
a
bad
state
and
degraded
state,
and
that
is
pretty
much
I
think
the
second
important
feature
of
Argos
CD,
which
is
enabling
a
lot
of
use
cases
related
to
CI
I'm,
not
going
to
talk
much
about
it,
but
I
think
you
kind
of
might
guess
what
that
means.
So
our
go
CD
assess
the
help
state
of
each
resource
and
cool
it
into
your
application
and
basically
trying
to
answer.
B
Is
the
resource
healthy,
unhealthy
or
maybe
it's
still
progressing
from?
You
know
healthy
state
to
something
else.
So
in
our
case,
state
is
degraded
and
you
can
use
alga
GUI
to
find
what
exactly
happened
and
the
basically
most
useful
tabs
in
this
case
is
events
and
events
tab
just
showing
you
normal
kubernetes
events.
You
can
use
Capezio
to
get
the
same
data,
but
with
Argos
CD
you
have
in
your
eyes
you
can
just
click
a
button
and
you
can
see
that
some
container
is
failing.
B
A
We
find
developers
sometimes
introduce
bugs
into
the
application.
This
is
the
early
clipping.
Sometimes
they
don't
they
don't
some
don't
test
them
before
they
push
them.
Sometimes
one
kind
of
like
hi
anything
to
do
is
actually
always
run
your
customized
build
or
your
hound
template
commands
locally.
We
do
get
surprising
them
or
people
who
don't
understand
why
the
applications
won't
deploy
in
anxious
because
the
application
is
110
place.
That's
always
a
good
bit
of
feedback
is
run
that
run
that
before
you
commit
and
push
yeah
so.
B
B
Okay,
let's
go
but
I
can
try
to
fix
it,
so
we
do
not
recommend
to
do
that
in
production,
but
alga
said
you
cannot
give
you
an
urgency
a
way
to
fix
energy
emergency
situation.
You
don't
really
have
to
change
anything
can
get.
You
can
just
use
our
go
UI
to
roll
back
to
most
previous
commit
which
was
deployed,
so
we
just
did
couple
deploys.
So
that's
why
I
know
that
previous
commit
was
successful
and
I'm
just
going
to
click
on
this
history
and
roll
back
button.
A
The
nice
thing
about
this
is
it.
You
know
we
talked
about
a
little
bit
about
how
it
trivializes
roll
back.
So
you
know
just
the
ability
to
go
back
to
a
previous
commit
that
we
know
is
good
rather
than
having
to
kind
of
fish
around
and
try
and
figure
out
what
what
it?
What
is
good
this
is
gonna,
have
on
that.
One
of
the
nice
benefits
of
get
up
stir
the
way
that
I
kind
of
whether
I
imagine
this
is
you've
just
done
a
release
if
I
say
4:00
p.m.
A
on
a
Friday
afternoon,
and
you
know,
colleagues
lost
a
bet
and
it's
gonna
buy
you
a
beer,
maybe
his
name's
Danny,
and
you
know
you
need
to
you-
need
to
quickly
fix
this.
All
you
need
to
go
is
go
into
the
interface,
you
click
rollback
and
you
just
go
back
to
previous
version
and
you're
golden
at
that
point.
B
Great
and
I
guess,
because
we're
really
getting
close
to
that
piece
of
a
workshop
I'm
going
to
execute
the
last
step
and
I'm
going
to
do
a
proper,
get
rollback
and
pretty
much
as
Alex
mentioned
these
get-ups
rollback
is
this
trivial?
All
you
have
to
do.
You
can
just
commit
a
revert,
can
it
create
a
revolt
commit,
and
this
is
the
git
command
which
you
can
use
to
save
time?
B
B
Get
it
off
this
images
section,
so
my
application
no
longer
overrides
the
image,
so
it
it.
It
will
fall
back
to
the
image
tech
which
is
defined
in
base.
So
it's
going
to
be
version
1,
and
this
is
it
that's
that's
rollback,
so
all
I
have
to
do.
I
just
need
to
push
it
back
and
make
sure
my
application
is
in
a
sink
state,
so
you
did
it
before.
B
Basically,
you
can
delegate
to
your
application
click
refresh
button
and
see
that
both
statuses
agree
now,
so
application
is
healthy
and
synced
just
before
we've
finished
this
part
of
a
Hanzo,
hands-on
I
just
wanted
to
reiterate
on
what
page
it's
off
and
what
we've
learned
in
it's
part
of
that
exercise.
So
basically,
almost
every
one
of
you
created
an
application
and
most
of
you
managed
to
customize
it
and
deploy
it
using
Argos
regime,
and
we
made
few
changes
in
get
to
upgrade
application
to
a
new
version.
A
B
A
A
A
Anybody
who
used
Dvorak,
keyboard,
layouts,
yeah
I,
learnt
this
about
five
years
ago
and
now
I
can't
type
on
a
normal
keyboard
which
is
really
annoying
when
you're
trying
typing
passwords
into
machines.
Okay,
so
that's
a
hands-on
Oh
took
a
bit
about
some
of
the
templating
tools
because
they're
you
know
they're
so
key
to
the
success
of
your
kind
of
kid
ops
solution
and
the
two
big
tools
out.
There
are
customizing
ham
and
customizes
the
one
we
just
use
in
the
workshop,
a
customized
just
a
command
line
tool.
A
So
you
typically
use
that
by
you
have
a
set
of
default
manifests
and
you
apply
overlays
to
them.
Customize
is
actually
tool,
we
use
it
into
it
and
the
reason
we
use
that
is,
it's
really
good
for
customizing
apps
that
you're
going
to
consume
yourself,
because
it's
super
simple
to
use
another
tool.
You
might
look
at
and
we
actually
know
that
a
number
of
companies
use
this
as
well
is
helm.
The
helm
is
a
way
to
package
an
application
up.
A
You
typically
take
a
set
of
manifests
and
group
them
together,
and
you
provide
some
additional
metadata
about
those
manifests,
such
as
the
name
and
the
version,
and
also
some
some
values.
What
those
helmets
are
is
unlike
customize,
which
is
just
plain,
Y
Amal.
The
helmet
charts
consist
of
manifests
that
are
templated
golang,
which
is
a
delightful
templating
language
to
use,
but
allows
you
a
lot
more.
Flexibility
than
customize
you've
got
things
like
if
and
you've
got
loops
and
other
kind
of
things
in
there
as
well.
A
A
There's
no
secret,
that's
one
of
the
difficult
things
in
get
ops
is
doing
secret
management
and
the
reason
behind
that
is
because
you're
putting
your
manifests
in
to
get
and
obviously
a
Q&A
secret
contains
base64-encoded
values.
So
what
I'm
going
to
do
now
is
I'm
going
to
give
you
a
little
demo
of
how
to
do
secrets
now?
You
can
actually
do
this
yourself,
because
this
is
a
you
know
laid
out
in
a
tutorial
for
that.
I
will
not
be
waiting
for
you
to
to
do
that,
though.
Well
I'll,
just
crack
on
with
it
as
well.
A
If
I
can
type
on
my
keyboard
after
joking
about
it
a
minute
ago,
I
can't
now
time,
okay.
So
what
I'm
going
to
demo
is
a
thing
called
sealed
secrets.
So
what
sealed
secrets
does
it
is
an
operator
that
sits
within
your
cluster
and
it
looks
for
these
resources
called
sealed
secrets
and
sealed
secret
is
essentially
a
comp.
A
secret.
That's
been
encrypted
using
a
certificate
from
that
repository.
A
A
A
Here
we
go
so
what
I've
done
is
create
a
secret
where
the
key
is
a
foo,
and
it
reads
from
standard
in
so
it's
bar
and
I
can
see
in
my
secret
that
I've
actually
got
that
encoded
as
a
base64
value,
and
we
can
actually
just
easily
to
code
that
using
in
pretty
much
a
million
different
base64
encoding
to
us
just
decode
that
we
can
see.
That's
just
who,
in
the
bottom
left
hand,
side.
A
Okay
and
so
in
here,
I've
got
a
block
of
encrypted
data,
which
basically
is
parallels
the
the
date
of
block,
but
that
data
is
encrypted
now
I
can't
you
still
I
can
get,
and
even
people
who
could
read
that
can't
decrypt
that
secret.
But
what
else
have
I
created
here?
Well
I've
also
created
my
secret
file,
which
I
don't
really
want
to
keep
in
get
because
you
know
it's
just
just
standard
base64
encode
fast.
We
can
remove
that
file.
So
that's
clean
and
I
can
just
commit
that
file
up
in
to.
A
B
A
A
Just
run
I'll
go
locally
in
and
do
that
it
takes
a
few
moments
to
startup,
locally
I.
Think
I,
don't
know
if
I've
mentioned
this
before
our
goes
open-source
software,
so
you
up,
so
you
can
just
download
this
we're
not
here
we're
not
a
vendor
here
today
to
sell
you
anything
we're
just
here
to
demonstrate
our
software.
Maybe
it
something's
interesting
you
and
you
might
we'll
get
involved
in
or
use
yourself
when
he's
started
up
it
takes
about
a
minute,
or
so
you
all
will
have
a
little.
B
A
Context:
okay,
okay,
cool,
okay,
so
what
I'm
gonna
do
now
is
I'll
just
create
that
application
in
my
local
cluster.
You
think
this
this
agro
CDC
Li
tool.
So
we've
shown
you
a
lot
about
the
the
user
interface
during
this
presentation,
but
there's
actually
other
ways
to
create
applications
that
one
way
to
create
an
application
is
using
the
CLI
which
you
could
just
brew,
install
the
agro
CLI
if
you've
tapped
it
and
otherwise
called
declarative
management.
A
B
A
A
There
we
go
this
guy
here
we
got
an
app's
appeared
in
the
user
interface
here,
okay,
so
you
can
see
in
this
application.
I'll
just
make
this
a
little
bit
bigger,
so
you
can
see
a
bit
more
easily.
We've
got
a
sealed
secret
in
there
and
we
just
click
on
to
this
guy,
we'll
just
see
it's
the
the
diff
here
says
we
need
to
create
seal
secret.
It's
quite
nice.
If
I,
just
click
sync
it'll
it'll
do
it'll.
Do
a
Red,
Square
I'll,
try
again
other
demo.
Gods
are
not
with
me
today.
A
I
have
like
races
special
there
we
go,
I
have
great
sit
in
the
other
cluster,
so
I
like
it.
Luckily
I've
crazy,
see
it
by
accident
in
another
cluster
here.
So
what
we
can
see
that
this
here
is
created,
my
actual
secret
here
and
I
can
open
up
my
secret
and
see
what's
going
on
there,
I
like
actually
see
that
I've
got
some
encrypted
data.
This
is
another
feature
of
Argos,
see
that
actually
redact
that
encrypted
data
from
from
the
user
interface.
A
So
you
can't
see
that
just
to
kind
of
close
out
on
secrets
you
know
this
is
this
is
a
particularly
tricky
area
to
do
in
get
ops
because
you're
keeping
your
secrets
inside.
If
this
is
not
the
only
solution,
we
know
about
five
or
six
other
different
solutions
that
you
can
use,
for
example,
getting
your
secrets
from
AWS
using
their
controllers.
What
underlies
sealed
secrets
is
actually
set
of
controls
that
live
in
the
coop
system,
namespace.
That
provides
you
with
this
information.
Okay,
what
about
them
with.
A
Okay,
so
what
I
wanted
to
do
is
just
see
how
many
stars
we've
got
now,
whether
I'm
going
to
make
my
beer,
which
I
hate,
doing
2014,
new
or
brilliant
right,
I
promised
you
the
only
gift,
you'll
see
I,
find
this
gif
hilarious.
I.
Don't
really
know
why
I
don't
know
what
I
think
it's
like
she.
You
know
she
gets
that
we're
celebrating
something,
and
then
she
smashes
something
on
the
floor.
A
I
don't
know
if
anybody's
got
kids
and
news
reminds
you
of
your
kids,
okay,
just
just
a
kind
of
wrap
up
if
you
want
to
come
and
find
out
more
about
doing
get-ups
with
Argo,
and
if
you
want
to
find
out
more
about
the
gear
at
our
Argos
eating
flux,
collaboration
of
what
we're
doing
if
you've
got
the
questions
asked
about.
How
do
you
secure
it?
How
do
you
deploy
it?
How
do
you
have
you
set
up
your
teams?
How
do
you
organize
your
name
spaces?