►
Description
This week on the GitOps Guide to the Galaxy, we're going over CI/CD deployment strategies that are GitOps friendly using Argo CD and Tekton. We'll talk about the fundamentals of what continuous integration and continuous delivery looks like in a GitOps environment and discuss the different approaches to building a GitOps-friendly continuous integration and delivery system. If you want to learn more about how to build a continuous delivery process on Kubernetes with Tekton pipelines and GitOps pattern managed by ArgoCD, this is the episode for you!
Follow us:
Christian Hernandez https://twitter.com/christianh814
Alex Collins https://twitter.com/alex_e_collins
What is GitOps Guide to the Galaxy?
Every other Thursday at 3pm ET hosts Christian Hernandez and Chris Short sit down to discuss everything in the GitOps universe, from end-to-end CICD pipelines to creating Git workflows.
Subscribe to Red Hat's YouTube channel: https://www.youtube.com/redhat/?sub_confirmation=1
About OpenShift:
Red Hat OpenShift is an open source container application platform based on the Kubernetes container orchestrator for enterprise application development and deployment.
Openshift.com https://red.ht/33Fh6uV
OpenShift on github https://github.com/openshift/
OpenShift on redhat.com https://red.ht/2o4ccDk
#RedHat #GitOps #Kubernetes
A
A
Welcome
to
get
ops
guide
to
the
galaxy
I'll,
be
your
captain
here
in
this
journey
christian
hernandez,
technical
marketing.
Here
at
red
hat,
as
you
see,
I
have
no
co-captain
today
right,
so
I
guess
the
co-captain
today.
Is
you
out
there
watching
this?
I
I
I
I
hope,
there's
no
too
many
cooks
in
the
kitchen
here
with
with
everyone
being
a
co-captain
here,
but
thank
you
for
joining
another
episode
here.
A
A
So
before
we
get
to
that
right,
I
like
to
steal
a
little
bit
from
from
andrew
talk
about
a
little
bit
on.
You
know
what
what's
top
of
mind
right.
So,
first,
first
and
foremost,
if
you
haven't
already
please
subscribe,
share
right,
it's
it's!
We
we
like
to
grow
the
show
I
I
love
hearing
from
everyone
all
around
the
world
on
twitter
on
linkedin
people
constantly
message
me.
I
like
that
so
go
ahead
and
feel
free
to
reach
out
to
me
either
on
twitter
right.
A
Follow
me
on
twitter.
I
believe
my
name
is
either
down
here
down
here.
I
always
forget
when
you
marry
their
christian
h814
twitter,
twitter,
handle,
get
get
handle
as
well
and
also
remember
subscribe
share.
You
know
again
all
that
that
cool
influence
influencer
stuff
here
so
so
today
I
knew
I
was
gonna
be
alone,
so
you
know
I
had
my
my
my
two
cups
of
coffee,
so
I'm
ready
I'm
ready
to
dive
in
there's
a.
A
I
have
a
little
weird
tradition
of
every
thursday
I
drink
two
cups
of
coffee
instead
of
one
it's,
it's
really
people
always
ask
me:
why
don't
you
drink
as
much
as
you
can
it's
like?
Well,
it's
I
don't
know
it's
just
weird.
If
you,
if
you
treat
something
like
like
a
special
occasion,
it's
actually
a
little
bit
more
special.
So
so
I
have
it
here
see
here.
So,
first
and
foremost,
let
me
share
my
screen
here
talk
a
little
bit
about.
A
What's
what's
top
top
of
mind
here,
let
me
make
sure
the
screen
comes
up.
I
think
I've.
Already
there
we
go
so
yeah
again,
redhat.com
live
streaming.
If
you
haven't
already.
This
is
our
one
stop
of
everything.
You'll
have
the
calendar
there
so
make
sure
you
know.
Never
miss
an
episode
or,
if
you're,
trying
to
figure
out
what's
gonna,
be
what
go
ahead
and
visit
that
so
a
little
top
of
mind,
like
I
mentioned
before
argo
khan,
is
coming
first
ever
inaugural
argo
con.
A
We
try
to
make
it
a
live
event,
but
actually
we
just
decided
to
make
it
virtual
so
still
time
to
register
still
time
to
event
to
attend
this
event.
There's
gonna
be
a
lot
of
cool
things
going
on
here.
So
let
me
let
me
click
on
the
schedule
here,
it's
multi-track,
so
this
is
really
cool
so
like
if
you're
interested
in
like
workflows,
argo
workflows,
argo
events
or
argo
cd
and
argo
rollouts,
there's
there's
a
couple
of
tracks.
A
The
reason
I'm
mentioning
this
here
selfishly
is.
I
do
have
a
talk
here.
Talking
about
the
staple
applications
right,
argo
city,
staple
applications.
I
did
do
a
stream
earlier
here
about
staple
applications.
This
is
kind
of
a
continuation
of
that.
So,
if
you've
seen
that
episode-
and
if
you
have
the
time
next
wednesday
1
25
p.m-
pacific
time
I'll
be
giving
a
talk
at
argo
con.
A
So
if
you
can
attend
that
it'd
be
great
I'll,
be
going
a
little
bit
further
in
in
that
conversation,
so
other
cool
things
going
on
argo
khan.
I
know
dan
garfield
is
going
to
do
the
keynote
here
at
codefresh,
really
cool
guy.
I
work
with
him
a
lot
in
the
in
the
open,
get
ops
working
group
and
you
know
they
have
some
argo
cd
production
best
practices.
A
Another
one
from
alexander
he's
been
on
the
show
as
well
talking
about
argo
cd
he's
going
to
talk
about
best
practices,
so
a
lot
of
cool
stuff.
If
you're
into
argo,
argo
city
using
you
know
the
article
projects
for
certain
things,
it's
going
to
be
a
lot
of
great
stuff
there
and
I
think
that's
it.
I
think
this
next
top
of
mind
thing
kind
of
just
goes
into
what
was
oh
before
I
forget.
Actually,
I'm
running
a
survey.
A
It's
the
next
top
of
my
thing
before
I
get
into
the
actual
today's
topic
here.
So
I'm
running
a
survey
here,
I'm
gonna
throw
this
in
the
chat.
Pretty
soon
here
is,
we've
been
doing
get
off
skype
to
the
galaxy
for
a
little
over
a
year,
almost
two
years,
almost
two
years.
So
this
will
actually
this
this
december
will
be
the
second
winter
break
shutdown
that
we
have
here
at
red
hat
that
we're
gonna
take.
A
So
I'd
like
to
think
of
it
as
two
years
and
we'd
like
to
see
where
the
the
what
where
you
guys
think
about
the
show
where
we
want
to
take
it
and
so
yeah,
so
just
make
sure
you
get
to
get
the
survey
and
take
that
survey.
You
know
give
me
your
thoughts,
be
as
mean
as
you
want,
maybe
not
as
many
as
you
want,
but
give
give
the
feedback
to
see
what
you
guys
want
to
see.
A
What
more
we
want
to
explore
in
this
getups
journey
right.
So
so
so
saying
that
right.
My
last
top
top
of
mind
thing
here
is
that
I
was
actually
gonna
go
to
devconf
right
speaking
of
of
of
events.
They
just
made
devconf
virtual
one.
Well,
that's
that's
that's
one
thing
and
two
is
stephanie.
The
producer
is
like,
please
don't
be
mean
in
the
chat.
A
So
you
know
workshops
are
a
little
harder
virtual
event,
so
I'm
actually
only
going
to
be
doing
a
talk
for
for
devconf
as
well
I'll
give
you
that
information
as
soon
as
I
I'll,
probably
tweet
out
that
information
as
soon
as
I
have
more
on
that.
So
so
that
being
said,
so,
let's,
let's
talk
a
little
bit
about
openshift,
ci,
cd
and
techdon
get
ops.
How
does
all
that
look
right?
And
so
I
I
believe
you
guys
seen
this
presentation
before
it's
been
shared
out.
A
This
is
a
public
presentation,
so
you
know
I'll
I'll
end
up
sharing
and
sharing
it
to
I'll,
probably
tweet
this
out
later
on.
We
shared
this
presentation
out
before
there's
nothing,
nothing
to
hiding
here,
I'm
not
going
to
go
over
to
the
presentation
right
bit
by
bit,
but
I
do
want
to
call
out
a
few
things
right.
A
So
when
we
talk
about
cicd
right
and
traditional
traditional
event-based,
ci
cd,
you
know
a
lot
of
those
same
concepts
still
exist
in
a
kubernetes
world,
the
cloud
native
world,
except
now
you
can
go
even
further
right,
extend
it
even
even
beyond
of
what
what
a
traditional
ci
cd
is.
Osu
ci
cd
looks
like
for
just
from
the
the
sole
purpose
or
the
the
sole
fact
that
a
kubernetes
is
a
declarative
infrastructure
right.
A
So
you
know
the
platform
that
is
delivering
the
application
can
also
be
the
platform
that
does
that
build,
and
it
does
that
check
for
you,
and
so
you
know
if
so
now
that
we
have
a
platform
like
openshift,
that's
declarative,
you,
then
you
can
use
some
of
these
tools
to
have
a
declarative,
continuous
delivery,
continuous
integration
aspect
of
it
right,
so
I'm
gonna
jump
around
this
presentation
right.
So
don't
let
me
make
a
little
bigger
here.
I
might.
A
A
A
Builds
happening
on
demand
only
when
you
need
them
right
versus
a
traditional
kind
of
like
ci
process,
where
there's
like
a
platform
that
does
the
the
builds
for
you.
Tekton
takes
the
the
serverless
aspect
of
it
to
where
it
only
comes
up
when
you
need
it,
so
you
know
you
need
you
want
to
do
a
build.
That's
a
task
right.
Each
task
has
an
individual
step
right
collection
of
tasks.
It's
called
the
pipeline.
A
You
know
we've
done
a
show
with
with
tech
tom
before
so
I'm
kind
of
gonna,
you
know
gloss
over
this
here,
but
kind
of
the
idea
is
that
you
have
a
step
that
does
a
thing
right
and
a
collection
of
steps
is
called
a
task
so
like
step,
one
is:
do
this
step
twos?
Do
that
that
sort
of
thing
in
a
collection
of
tasks
is
called
a
pipeline
and
I'll
be
going
over
kind
of
like
what
my
demo.
A
You
know
how
all
that
looks
right,
because
the
idea
is
that
these
are
reusable
right.
These
are
kind
of
like
yeah.
You
can
specify
a
parameter
right,
for
example
like
if
you
need
to
do
a
maven,
build
on
something
you
just
you
know.
If
the
maven
build
is
always
the
same,
then
maybe
you
make
it
a
parameter.
If
you
just
pass
it
the
get
repo
for
example,
and
so
so
we
take
that
concept
right.
A
That
makes
not
only
does
a
cd
right,
continuous
delivery
to
nearest
deployment,
but
makes
make
sure
that
that
drift
detection
is
always
happening.
So
not
gonna,
go
too
deep
into
that.
You
guys
should
already
know
right.
If
you
don't
watch
past
episodes
right,
stephanie
stephanie,
the
producer,
I'm
gonna
call
her
stephanie.
The
producer
asked
me:
hey
there
any
share
any
links,
and
I
actually
forgot
to
show
this
link.
A
So
sorry,
stephanie
all
right
put
that
in
the
chat,
red
dot,
ht,
slash,
get
ops
cash,
past
episodes
subscribe
and
like
all
that
good
stuff.
So
let's
go
let's
focus
in
on
how
does
ci
cd
look
like
in
a
good
ops
world?
So,
let's,
let's
look
at
this
guy
right
here
for
a
little
bit
so
best
practice
is
that
you're
always
you're
always
looking
at
you
always
have
at
least
two
source
repositories
right.
You
get
repositories
right.
A
You
have
the
the
source
repository
of
the
actual
application
that
shouldn't
be
any
different
that
actually
fits
nicely
in
the
git
ups
model,
but
that
doesn't
really
change
in
a
good
ops
model.
Actually
right,
because
we
say
that
you
have
you,
you
have
to
separate
your
source.
Git
repository
from
your
configuration
repository
is
what
we
call
it:
the
that
separation
I've
always
thought
that
you
would
have
it's
better
to
have
it
in
one
repo
write
the
code
and
its
deployment
manifests.
A
I
quickly
found
out
that
that
is
really
hard
to
manage.
So
I've
always
said
split
it
out.
I've
talked
to
hundreds
of
customers
who
always,
if
there's
a
contention
where
there's
some
customers.
That
just
say
absolutely.
No.
I
want
it
all
in
one
repo-
and
I
say:
okay,
that's
fine.
I
used
to
think
like
you
too,
so
just
just
trust
me
on
that.
I
actually
think
you
have
multiple
ones
right.
You'll
have
a
config
repository
specific
to
the
cluster
and
a
config
repository
specific
to
the
deployment
of
the
application.
A
So
I
think
you
should
at
least
have
three,
but
we're
gonna,
keep
it
a
little
simpler
than
that,
so
you're
gonna
have
at
least
two
right.
You
have
a
config
git
repository
source
code,
repository
right
and
they're,
going
to
work
in
tandem,
and
so
the
the
idea
is
that
since
you
have
two,
they
each
kind
of
have
an
independent
life
cycle
source
code
repository
where
that
life
cycle
is
going
to
stay.
The
same
really
right,
you're
gonna!
That
process
is
gonna,
is
gonna,
stay
the
same,
whether
using
git
ups
or
not.
A
I
actually
have
a
an
opinionated
way
of
configuring,
those
whether
it's
the
config
for
the
cluster
or
the
config
for
the
deployment
of
the
application,
and
it's
something
called
a
trunk
based
development
right,
and
so
let's
talk
a
little
bit
about
trunk
based
development.
This
is
also
another
point
of
contention.
A
One
of
my
ex-co-workers
right-
we
still
keep
in
touch,
says
that
he
does
a
lot
with
in
his
current
job
with
customers.
He
does
a
lot
with
argo
city
and
techton.
So
I
wonder
what
he
thinks
about
that.
If
he's,
if
you're
there
nick,
I
know
you'll
pipe
up
and
disagree
with
me
if,
because
we're
both
opinionated
guys
but
just
in
short
trunk
base
development
is
something
that
I
use
with
the
config
repository.
A
A
You
need
to
the
yeah
yeah
yeah,
so
so
benjamin
says
ex
co-workers
still
keep
in
touch
kind
of
sounds
like
a
friend
yeah,
so
yeah.
So
I
guess
so,
I
guess
yeah
I
have
a
friend
of
mine,
exco
worker
and
still
friend
so
you're,
absolutely
right,
and
so
the
yeah
back
to
this
trunk-based
development,
so
back
trunk
based
development.
A
A
A
That
could
be
what
that
could
be
your
trunk
right,
whatever
the
trunk
is,
and
you
have
different
release
branches
right,
so
you
have
different
like
feature
branches,
so
free
feature
branches
are
basically
things
that
are
going
to
go
back
into
trunk
and
things
that
are
going
to
be
deleted
because
release
feature
branches
are
short-lived,
essentially,
is
what
so
so
with
with
the
trunk
based
development.
Here,
trunk
based
development
at
scale
is
best
that
was
short-lived
feature
branches.
So
that's
kind
of
like
the
the
idea.
A
Right,
like
you
have
the
trunk.
You
have
main
production
staging
whatever
right,
whatever
it
is,
I'm
going
to
say
main
again
right:
you
do
you
don't
commit
to
the
release
branches
right,
you
do
your
work
in
the
release
branches
and
then
it
gets
committed
back
down
to
the
trunk
and
in
the
trunk
that
gets
deployed
into
production.
A
So
so
this
is
like
one
of
the
things
that
that
I
that'll
that
I
find
a
lot
of
people
do
are
doing
with
going
back
to
the
presentation
with
the
config
git
repository
right
with
the
source
code
repository
that
has
its
own
life
cycle.
That
has
that
cycles
independent
and
different
than
the
config
repository
that'll
have
like
feature
branches,
long
live
feature
branches,
that'll
have
tags.
That'll,
have
you
know
all
you
know,
release
bundles
that'll,
that'll
that'll
have
that'll
have
its
its
own
life
cycle.
A
The
configurate
repository
usually
follows
what
I
call
a
what
they
call
a
trunk
based
development.
So
not
gonna
go
through
all
of
this
here.
This
is
a
great
read,
especially
if
you're
looking
into
your
git
workflows,
whether
you're
a
dev
on
the
dev
side
or
the
op
side,
or
you
know
sre.
This
is
a
great
read
about
how
well
what
the
idea
is
behind
trunk
based
development.
So
so,
if
you
haven't
seen
that
that's
what
this
config
repository
is
right.
A
So
so
what
ends
up
happening
is
that
someone
makes
a
a
commit
right
to
something
right
so
to
code.
That's
gonna
trigger
a
ci
process
right,
so
the
ci
process
does
something
with
it
right.
It
does
everything
that
you
would
think
with
ci
process.
It
builds
the
application,
does
a
test.
You
know
once
the
test
passes
it
pushes
that
image
to
image
repository
in
that
same
ci
process,
there's
going
to
be
a
change
to
a
config,
git
repository
right,
so
actually
in
the
ci
process.
A
The
actual
ci
process
is
committing
to
your
config.
Git
repository
right
sounds
a
little
scary,
but
what
we'll
go
through
in
detail
in
the
demo
once
that
is
done,
the
the
get
ops
controller
in
our
case
it'll,
be
argo
cd,
we'll
see
that
change
and
it'll
act
on
it
right
so
it'll
say:
oh,
hey,
there's
a
a
new
image
update
that
was
made
to
the
deployment
configuration.
A
Let
me
push
that
into
whatever
environment
that
you
push
that
into
ci
process
can
do
other
things
like
create
a
pull
request
instead
of
committing
directly
that's.
You
know
part
part
of
the
little
demo
that
I'm
going
to
have
here.
That's
going
to
happen
right
so,
but
really
here
for
in
a
get
ups
application
delivery
model.
The
magic
is
really.
The
ci
process
interacts
with
your
git
repo,
with
a
configuration
deployment,
git
repo
once
a
build
is
successful.
A
So
here
let
me
so
here
what
I,
what
I
have
here
is:
I'm
gonna
drop
it
in
the
chat
this
this
link,
I
did
give
to
stephanie
the
producer,
so
so
I
did
prepare
at
least
for
this
here.
I'm
actually
so
this
is
this.
Is
the
csd
demo
that
I
have
you
guys
can
take
a
look,
there's
actually
a
helm
chart
for
it.
So
if
you're
running
openshift,
4.7
or
newer,
this
will
work
for
you
and
it
needs
to
be
a
a
freshly
installed
openshift
cluster.
A
But
this
is
kind
of
like
what
I
was
going
to
go
through
in
def
conf.
But
it
looks
like
that's.
You
know
the
the
that
workshop's
not
gonna
happen.
So
that's
why
I
decided
to
do
this
this
stream.
So
here
you
can.
You
know,
look
look
at
look
at
how
how
I'm
doing
some
of
these
some
of
this
demo.
Here
and
again,
it's
a
it's
a
helm,
it's
a
home
chart,
so
you
can
just
add
it
and
and
run
run
with
it
here.
A
A
A
All
right
there
we
go
cool
so
so
the
way
this
is
set
up,
so
I
had
I
actually
logged
in
as
a
developer.
I'm
not
I'm
not
a
cluster
admin
here.
I
actually
deployed
my
own,
so
here
I'm
in
welcome,
get
ops.
Let's
go
to
topology
here
I
actually
deployed
my
own
namespace
scoped,
argo
city
instance
right
so
there's
so
this
that's
part
of
of
the
argo
city.
So
there's,
if
you
go
here,
let
me
clear
this
get
argo
cd
right,
so
this
is
openshift.
A
Get
ops
is
the
argo
cd
that
manages
the
cluster,
and
this
is
a
deployment
of
my
my
you
know:
namespace
scoped,
argo
cd,
that's
what
we
call
it
or
I
guess
yeah
name,
space
scope
or
go
see.
We
had
another
name.
I
forgot
what,
if
you
guys
remember
we
had
another
name
for
it,
you
drop
in
there.
I
think
it's
called
project
based,
not
sure
anyway.
This
is
the
openshift
get
ops.
A
That
kind
of
just
takes
care
of
its
the
cluster
as
a
whole,
but
I'm
actually
gonna
focus
on
like
as
if
I'm
a
developer,
I
deployed
my
own
version
of
argo
cd
using
the
get
ops
operator
to
it
to
a
specific
namespace.
So
this
space,
I'm
gonna,
call
it
welcome
argo
cd.
So
that's
what
this
is
here
and
that's
what
this
is
here.
So
I,
as
you
can
see
that
I
have
different
environments
set
up
technically
too.
I
have
dev
and
prod
right
and
I
have
a
pipeline.
A
A
This
actually
takes
a
while
to
run
so
I
won't
talk
about
each
individual
step
until
I
actually
fire
it
off,
but
I
have
pipeline
here
that
that
that
basically
builds
and
deploys
the
development,
and
then
I
I'm
gonna,
promote
it
to
production
once
things
are,
are
going
are
going
smoothly
right.
I
have
this
app
here.
Dumb
app
right
now,
it's
shouting
green
right.
This
is
shouting
green
as
well,
and
so
this
is
production.
This
is
dev
and
I
have
a
get
t
instance.
Let
me
log
into
this
oops.
A
A
So
this
is
like
a
static
php
page
here,
and
I
also
have
the
deployment
right
so
now,
as
you
can
see
here,
I
have
two
repositories,
one
for
one
for
the
application
and
one
for
deployment
here
and
here
deploy
an
app
and
then
and
then
the
pipeline
right.
So
the
pipeline
actually
lives
in
the
deployment.
I
believe
yeah
techton
right,
so
I
actually
so
I'm
kind
of
using
using
everything,
declaratively
right
and
so
I'm
using
argo
cd
to
deploy
to.
A
I
use
argo
city
to
do
to
point
to
the
deployment
repo.
That
then,
creates
the
dev
the
pipeline
in
the
production,
and
it's
all
self-contained
here,
and
so
usually
the
first
step
here
is:
let's
make
a
let's:
let's
do
we
have
that
we
have
time.
So,
let's
make
this
a
little
longer
here.
Let's
now
this
deployment
right,
we
want
to
make
a
code
change,
so
let's
go
to
the
application
and
then
let's
make
a
code
change
here.
So
let's
can
we
create
a
branch?
A
A
Your
pull
request,
no
no
for
requests.
No,
because
you
need
a
branch
for
that
right.
I
thought
you
could
do
it
and
get
t.
Maybe
you
can't?
Oh
okay,
I
can
do
this.
Let's
make
the
edit
and
maybe
that'll
create
a
new
branch.
So
here
we
have
green
right,
see
we
have
green
and
let's,
instead
of
blue
green
we're
going
green,
blue
or
let's
go
yellow
right.
Let's
do
it.
Let's
do
another!
A
A
Right
for
profile
change,
so
here
I
create
a
pull
request:
right
change,
changing
to
yellow,
okay,
pull
request
cool
all
right.
So
now
we
have
the
application
right.
I
was
a
developer.
I
made
a
pull
request.
I
called
my
hat
right:
I'm
a
different
developer!
Now
right,
I'm
the
architect,
I'm
like
all
right.
Let's
this
looks
good
files
change
all
right,
yellow
yeah,
that's
what
we
wanted
looks
good
to
me
oops
and
let's
go
to
merge
pro
request.
A
So
this
this
is
merged
right,
and
so
what
this
does
is
this.
If
I
go
down
to
the
pipeline,
is
this
kicks
off
another
build
right?
A
lot
of
people
ask
me:
how
does
build
get
kicked
off
when
you,
when
you
commit
to
that?
Well,
as
you
as
you
probably
already
guessed,
there
is
a
where's,
my
repo,
let
me
go
back
there,
we
go.
There's
a
web
hook
right
and
so
let's
go
to
web
hook.
A
A
A
Let
me
go
back
to
token
here
and
that
pipeline
is
defined
here
right,
and
so
basically
it
just
takes
that
trigger
template
takes
that
template.
It
feeds
it
into
the
pipeline
with
those
parameters,
and
you
can
do
really
some
pretty
cool
things
with
these
parameters.
You
can
do
you
can
extrapolate
things
from
the
payload
right
so
like.
If
you
want
certain
information
from
the
payload,
you
can
take
those
as
well
and
pass
that
to
your
pipeline
as
well.
So
here
so
first
thing
it
does
it
just
clones.
A
The
repo
locally
right
first
step
is
always
the
cloning.
So
once
it
clones,
the
repo
locally
it'll
set
the
image
tag.
So
this
is
basically
it'll
set
a
set's,
a
variable
right,
so
it'll
say:
okay,
I'm
gonna
set
this
commit
as
a
tag
right,
because
you
want
be
able
to
differentiate
the
tags
of.
What's
in.
A
What's
in
dev
versus
the
next
environment,
right
between
environments,
I
don't
like
floating
tags
things
like
dev
things
like
you
know,
colon,
dev
or
colon
prod,
because
that's
kind
of
an
anti-pattern
forget
right
because
in
git
you
want
the
the
source
of
truth
right
because
floating
tags
can
change,
whereas
things
like
either
the
sha
or
like
a
specific
uid.
Don't
so
you
have
that
once
it
cloned
and
sets
that
variable
it'll
actually
do
a
a
build
right.
So
here
here's
here's
an
aspect.
A
A
So
here
I'm
tagging
latest
right,
even
though
I'm
not
using
latest
in
my
pipeline,
I
figure
I'll
tag.
The
latest
in
case
someone
wants
to
you,
know,
play
around
with
this
image
and
they
just
want
the
latest
doing
some
testing
right
once
I
do
that,
I'm
gonna
clone
the
deployment
repo
right.
So
here
I
just
do-
I
was
just
here
to
four:
I
was
just
working
with
the
the
software
application
here
now
I'm
actually
cloning,
the
deployment
repository
and
what
I'm
doing
here
is.
A
I
want
to
have
it
here
so
that
way
I
can
interact
with
that
repo
right.
So
first
I
did
a
build
on
the
the
application
repo
and
now
I'm
cloning,
the
deployment
repo,
I'm
gonna
work
specifically
with
this
guy
here,
I'm
gonna
patch
dev.
Essentially
I'm
running
a
customized
edit.
What
this
does
is
essentially
is
gonna
update
the
image
to
the
new
version
here
and
then
I'm
gonna
commit
it
to
deb.
A
A
A
This
image
I
wanted
to
use
this
new
tag
which
it
should
say
in
here
yeah
one
dd,
one
dd
right
so
it'll
it
it
updates
a
tag,
and
so
when
it
updates
the
tag,
it
actually
deployed
it
right.
So,
like
you
go
refresh,
I
think
yeah,
it
already
did
it
while
we
while
we
were
talking.
I
wonder
if
you
can
see
the
yeah
so
here
it
says
the
last
sync
was
done
one
one
dd
right,
so
it's
there
so
that
actually
already
deployed.
A
So
if
I
go
here,
this
should
say:
yellow,
let's
cross
our
fingers,
it
says
hello
sweet,
but
in
production
it's
not
it's
not
yellow
right.
Why
is
that?
And
so
it's
because
in
my
in
my
workflow,
I'm
actually
going
to
patch
prod
right.
Do
the
same
thing.
Instead
of
doing
it
in
dev,
I'm
going
to
do
it,
I
do
it
in
the
in
the
prod
folder.
A
You
know
1dd
there.
It
is,
but
I'm
actually
going
to
create
a
branch
right.
This
is
my
feature
branch.
This
is
the
branch
that
I
want.
I
don't
want
it
directly
in
production.
Just
yet
right
I
want
to
I
want.
I
want
it
and
I
want
it
in
dev.
I
want
the
changes
in
place,
but
for
prod.
I
want
to
make
a
pr,
because
I
need
other
people
to
look
at
it
right.
So
this
is
the
process.
Here
is
the
process
here
is
making
a
a
a
pr.
A
So
that
way
you
kind
of
have
like
stop
gaps
or
feature
or
gating
right.
This
is
kind
of
how
you
would
do
it
gating.
So
pr
to
prod
is
here.
Oh,
it
looks
like
we're
getting
raided.
Nice
welcome
everyone,
so
here
I
make
a
pr
to
prod,
and
so
we
got
a
pr
to
prod,
and
here
we
can
go
to
the
the
repo
and
you
can
see
the
pr
is
already
there.
So
we
have
a
pr
to
oops.
Let's
go
back
here.
A
So
this
here,
it'll
it'll,
show
you
that
in
my
production
folder
in
the
customization,
I
that's
the
file.
I
want
to
change
right.
This
is
this
is
yeah
release1dd.
We
want
that
here
and
then
the
converse
to
go
back
to
the
conversation
here.
So
here
I
would
go
into
you
know.
Switching
my
hat.
You
know
I'm
an
sre,
I'm
you
know
whoever
a
release
manager
all
right.
I'll
look
at
this.
This
looks
good
now
here
here.
This
is
this.
Is
this?
Is
the
part
where
it
gets?
A
You
know
where
we
can
kind
of
think
about
this,
a
little
bit
the
here
we
may.
We
may
or
may
not
want
to
merge
this
right
so
part
of
the
pipeline.
Let's
go
back
to
the
details
and
see
where
this
here
you
know
part
of
pr
to
prod
can
then
trigger
yet
another
pipeline
that
tests
this
right
in
production.
It
doesn't.
Maybe
you
know
this
is
kind
of
like
a
simple
use
case
right
so
before
I
click
merge.
A
I
want
to
you
know
kind
of
talk
about
how,
like
other
things,
you
can
do
right
so
here
when
it
does
a
pr
to
prod
it
can
trigger
yet
another
pipeline
right,
because
it's
just
text
on
right.
This
is
just
you
know
it.
It's
all.
It's
all
web
hooks
right.
So
here
you
can
just
web
hook
into
it
yet
another
pipeline
that
does
yet
even
more
testing
right.
So,
like
hey,
it
passed,
dev,
that's
cool.
A
It
made
a
pr
to
prod,
but
we
don't
want
to
just
merge
it
right
without
testing,
so
this
can
then
trigger
yet
another
pipeline.
I
just
want
to
mention
that,
because
you
know
every
time
I
kind
of
just
you
know
talk
about
this
from
a
high
level
view
people
go
well
wait.
You
know
what
what
about
other
pipelines
like
well
yeah!
You
can
do
other
pipelines
here
as
well.
You
can
chain
these
together.
A
You
can
have
either
one
big
pipeline
run
or
many
little
ones
right.
You
can
have
slack
ops
where
it
tells
you
if
something
passed
or
something
failed,
all
that
good
stuff.
So
but
here,
let's
just
say,
assuming
that
it
triggered
another
pipeline
and
it
did
all
the
testing
and
assuming
all
of
that
passed
assuming
that
you
know
someone
did
peer
review
on
this
stuff
here,
maybe
I
want
a
merch
pro
request
right
and
you
should
do
looks
good
to
me
all
right.
Let's
merge
pro
request.
A
Right
here,
what
I
like
doing
is
like
deleting
this
branch
right.
You
may
not
want
to
delete
it,
I'm
following
trunk
based
right
as
soon
as
I
merge
that
feature
branch,
I'm
deleted
shortly,
feature
branch-
I
I
I
do
that
I'll
leave
that
up
to
you
right
now,
I'll
I'll.
Let
you
do
that.
You
know
that
I'll
I'll,
let
I'll
I'll
I'll.
A
Have
you
your
your
devops
team
members
here
argue
about
whether
or
not
to
keep
that,
but
for
me,
since
we're
following
trunk
base
I'll
just
delete
that
we
don't
need
that
hanging
around,
and
so
once
you
do
that,
I'm
just
gonna
refresh
this
right.
So
usually
the
reconcile
loop
takes
three
minutes
for
argo
cd
right.
So
argo
city
has
three
minutes.
That's
adjustable
right.
A
You
can
adjust
that
so,
instead
of
sitting
here
for
three
minutes
I'll
just
hit
the
refresh,
so
it
checks
and
as
you
saw
it,
it
deployed
that
application.
So
here
this
is
product.
Sorry,
this
is
dev,
and
this
is
my
my
faux
production
here.
Now
it's
yellow
right.
So
now
we
are
now
they
match
right,
and
so
this
is
kind
of
the
you
know
way
too
quickly.
Do
iterative
application
development
with
get
ops
right.
A
So
here
I
actually
so,
if
you
notice
here,
I
didn't
actually
interact
with
argo
cd,
all
that
often
other
than
hitting
refresh,
and
I
didn't
really
interact
with
openshift
all
that
much
everything
was
driven
through
git
right,
so
I'm
using
git
t
here,
for
example,
but
here
I'm
using
git
yeah
all
the
this
process.
If
you
think
about
it
was
done
via
git
and
I
can
even
distill
it
even
easier
than
that.
I
could
have
actually
technically
used
something
like
vs
code
and
just
started
committing
through
vs
code.
A
So
and
all
the
automation
is
happens
that
happens
in
the
background
there
see.
So
so
this
is
kind
of
how
you
would
use
here.
You
can
see
the
history
right,
succeed,
succeed.
I
had
some
issues
here.
I
don't
know
if
you
guys
had
issues
with
the
other
day
we
had
an
outage
from
our
our.
I
think
it
was
one
of
our
registries,
registry.redhead.I,
o
right.
I
was
like.
Why
is
this
taking
so
long?
Why
is
that?
A
Why
can't
I
pull
these
images,
I'm
trying
to
prepare
for
a
stream
and
then
I'm
like?
Oh
yeah,
it's
I'm
kept
getting
404s
for
whatever's
anyways.
They
fixed
that.
So
that's
just
another.
Another
thing
to
look
out
for
right
is
your
your
your,
where
you're
keeping
your
images
right
a
lot
of
people
like
to
use
something
like
quay
or,
if
you're
in
england
key
right
as
they
call
it
there
in
australia.
I
think
in
canada
as
well.
I
don't
know,
I
think,
there's
some
canadians
watching.
A
Do
you
guys
call
it
key
or
que
drop
that
in
the
chat,
if
you,
if
you
guys
know,
but
some
of
them
like
to
have
that
in-house
right.
So
that
way,
you
don't
run
into
the
problem
that
I
did
and
I
think
it's
a
good
idea
is
to
have
your
your
in-house
red
registry
server
right,
even
if
it's
just
mirroring
right,
so
you
can
even
have
it
as
a
mirror.
I
think
that's
that's
another
cool
idea.
Actually,
that's
another!
A
A
cool
show
idea
is
to
have
mirroring
and
mirroring
git
repos
right.
So
the
the
idea,
someone
brought
this
up.
The
idea
of
having
a
git
repo
mirror
your
actual
git
repos,
so
that
way
in,
like
when
you
have
like
a
network
cut
out
or
like
a
have
the
source
of
truth
closer
to
where
you're,
where
your
clusters
are
right.
So
this
person
I
was
I
was
talking
to-
is
a
company
can't
say
the
name,
but
they
manage
hundreds
of
clusters
all
around.
A
You
know
north
america
and
they
decided
to
use
git
mirrors
right
because
they
want
they
don't
want
like
a
an
outage
in
their
in
their
get
system.
I
think
they're
using
a
github
or
gitlab
internally.
They
don't
want
that
going
down
to
you
know,
make
the
other
systems
going
down
or
network
ad
outage
right
to
you
know,
mess
anything
up,
so
they
actually
had
set
up
get
mirrors.
A
So
maybe
I'll
do
an
episode
of
that
that'd
be
kind
of
kind
of
cool
to
to
explore
so
anyways
back
at
the
topic
at
hand.
So
here
this
is
kind
of
what
what
I
had
in
mind,
some
other
things
to
consider
now
I'll
actually
bring
up
the
really
really
cool
slide
that
I
wanted
to
show
in
the
beginning.
But
I
thought
I
think
it's
it's
better
to
show
this
from.
You
know
a
high
level
with
how
of
you
so
nick
if
you're
watching.
A
This
is
probably
what
you
wanted
to
see
the
other
stuff.
Probably
it
was
probably
too
low
too
high
level
for
you,
but
here
this
is
how
to
use
some
of
these.
Some
of
this
offering
some
of
the
openshift
offering
right.
So
here
you
can
see
how
all
the
openshift
offerings
the
red
hat
offerings.
All
the
technologies
that
we
have
here
can
be
used
together
right
and
so
still
same
idea
with
the
application.
A
Git
repository,
config
git
repository
you'll
have
one
or
many
application,
git
repositories
and
one
or
many
config
repositories,
and
oh,
hey
tiger's.
Here,
hey
tiger
tiger
the
intern-
I
guess
you're,
not
the
intern
here.
So
so
we
got
here
the
config
here
you
have
one
or
more
many
many
config
repositories
here
and
so
the
ci
our
vision
is,
we
see
it
as
techton
right
and
that's
kind
of
like
your
ci
process
outside
of
get
ops.
A
Tecton
can
do
ci
cd,
but
in
in
the
in
the
get
ups
world
right
in
our
in
our
journey.
Here
we
focus
on
the
the
ci
aspect
of
it
here.
You
could
also
use
our
argo
workflows
right
if
you
don't
use
tecton
and
then
doing
argo
cd
for
the
reconciliation
of
it
as
well.
Right-
and
you
know
we
kind
of
went
through
that,
but
on
top
of
that,
you
want
to
do
something
with
security
right.
A
So
that's
when
advanced
cluster
security
stack
rocks
comes
into
play
right
and
that's
how
this
interacts
with
your,
not
only
your
image
registry
right
for
your
ci
process
right,
because
you
can
have
scanning
happen
in
your
ci
process.
You
can
have
scanning
happening
in
an
image
registry
as
well,
and
you
can
make
sure
that
those
clusters
that
you're
running
are
in
compliance
right.
So
this
is
kind
of
where
acs
fits
in.
A
Here
we
had
michael
foster
from
acs,
really
really
cool
guy
came
to
came
on
the
show
talk
about
acs.
This
is
kind
of
where
the
acs
aspect
fits
into
it
right,
so
the
the
ci
process
interacts
with,
or
I
guess,
acs
interacts
with
your
ci
processing
and
vice
versa.
And
then
you
have
argo
cd
here
doing
the
reconciliation,
the
for
your
applications
right.
A
So
you
want
to
be
able
to
deploy
like
we
saw
right,
have
argo
cd,
making
sure
everything's
in
constant
sync,
making
sure
that
there's
no
there's
no
drift
detection
as
well
on
top
of
doing
the
the
cd
process
of
it
as
well,
and
then
you
have
the
advanced
cluster
management.
The
advanced
cluster
management
here
or
acm
right
right
at
acm
does
lifecycle
management
of
your
of
your
clusters
as
well
right.
A
So,
whereas
at
red
hat,
we
have
the
view
of
argo,
cd
is
meant
for
getup
workflows,
which
kind
of
overlaps
with
acm,
because
acm
does
have
a
git
ops
component
but
acm.
I
always
tell
people,
and
it
always.
You
know
whether,
whether
it's
people
here
at
red
hat
right,
whether
you're
in
the
field,
the
red
hat
or
really
just
customers
here
you
know,
I
always
say
calling
acm
a
get.
Ops
tool
is
really
short
selling.
A
It's
really
really
cool.
So
then,
here
you
kind
of
have
a
bigger
picture
of
how
that
looks
like,
and
so
you
can
have
acm
interact
with
your
ci
cd
process
to
spin
up
even
additional
clusters.
This
is
kind
of
think
about
that
for
a
second
when
you
have
like,
if
you
want
to
test
the
feature
branch,
you
can
have
acms
spin
up
a
cluster
right,
interact
with
your
ci
process
to
spin
up
a
cluster
and
do
the
test
that
way
right,
deploy
the
application
on
a
test
cluster
before
actually
promoting
it.
A
You
can
do
that
with
acm.
You
can't
do
that
with
argo,
but
you
can
definitely
do
that
with
acm,
and
so
this
is
kind
of
how
that
bigger
picture,
how
everything
fits
together
in
that
bigger
picture.
So
everything
here
in
the
middle
is
batteries
included
right.
So
for
those
people
always
ask
me
like
oh
how
much
you're
charging
for
how
much
you're
charging
for
tech
time
is
like?
No,
that's
it's
included,
use
it
supported
batteries
included
in
openshift.
A
If
you
need
anything
beyond
that
right,
you'll
you'll
definitely
need
a
something
for
a
for
acs
and
acm.
That
does
have
a
skew,
but
you
can
always
test,
for
example,
open
cluster
management.
That's
the
upstream
of
acm
and
stack
rocks
is
the
upstream
for
acs.
So
this
is
kind
of
the
idea
behind
that
right.
So
that's
kind
of
all
I
wanted
to
share.
I
don't
know
if
there's
any
questions,
I'm
looking
looks
like
the
raiders
may
or
may
not
have
left.
A
Thank
you
for
joining
anyways
a
tiger.
Is
there
so
tiger
the
intern?
Now
co-worker?
So
there
you
go
so
we
like
hiring
our
interns
here.
One
thing
I
did
want
to
try
is
that
so
notice.
How
so
it
you
may
or
may
not
have
noticed
it,
but
if
you
did
notice
that
how
how
I
had
to
click
refresh
for
for
this
to
do
the
reconciliation
right.
So
basically
it's
like
hey,
you
know
I
click
refresh
like
you
can
actually
have
argo.
A
Do
that
for
you.
So
let's
kind
of
look
into
that
for
the
let
me
see
we
can
do
some
exploring
the
last
last
few
minutes
here.
Web
hook
there
we
go
there,
we
go
get
web
hook
configurations.
So
here
we
have
so
here
it
says
you
can
do
a
webhook
configuration
for
the
application
here.
So
let's
go
so
the
idea
being
that
in
the
pipeline,
let's
go
back
to
the
pipeline.
A
Once
the
patch
we're
gonna
commit
where's
the
commit
to
yeah
commit
to
dev
the
commit
to
dev
step
that
can
also
hit
a
web
hook
right,
and
so
I
was
thinking.
Maybe
we
can
hit
that
web
hook
to
do
the
on-demand
reconciliation
instead
of
waiting
for
three
minutes
we
may
run
out
of
time,
but
let's,
let's,
let's
kind
of
just
start
that
that
journey.
I
haven't
done
this.
Yet
I'm
doing
it
live
so
we'll
see
if
I
fail
miserably
here
so
blah
blah
blah
argo
cd
api
web
book.
A
Hopefully
that
wasn't
too
loud,
so
let's
curl,
oops
curl
dash,
k
and
see
here
it
says
to
do
api
a
web
hook
unknown
web
pokemon.
That's
a
good
sign.
Let's
go
to
see
what
it
says
here.
Oh
bad
request
mark
bundle
not
supporting
multiples,
it's
probably
expecting
a
payload,
but
that's
good
it.
It
looks
like
looks
like
it's
there,
so
it
looks
like
the
webhook.
Is
there
it's
listening
for
a
web
hook?
A
It's
not,
but
we're
not
sending
it
the
right
information.
So
that's
good.
So
at
least
we
know
that
the
white
hook
is
there.
It's
just
we're
not
sending
it
the
right
information,
so
I
probably
is
expecting
a
a
trigger,
but
I
wonder
what
what
the
payload
looks
like
I'd
like
to
see.
If
I
could
just
curl
it
once
to
see
if
we
can
make
it
go
see
here
yeah,
it
doesn't
tell
you
what
the
payload
looks
like
so
json.
A
A
A
A
A
A
Let
me
see
here-
and
this
is
just
I
think
it's
just
the
yeah.
This
is
just
the
documentation
yeah
I
may
have
to.
We
may
have
to
do
I'll
it'd
be
interesting,
because
I
think
this
is
a
it'd,
be
a
good
update
to
that
csc
demo
for
it
to
actually
hit
the
web
hook.
So
this
might
be
homework
for
me
because
I
don't
think
we're
gonna
figure
this
out
here,
because
here
it
actually.
So
if
we
go
back
to
the
documentation,
where's
the
documentation,
I
always
lose
the
documentation
there.
A
It
is.
This
is
talking
about
actually
on
a
commit,
but
I
actually
want
to
use
the
reason
I'm
using
pearl.
So
this
kind
of
my
thought
process
go
back
a
little
bit.
The
reason
I'm
using
curl
is
because
I
actually
don't
want
to
put
it
in
at
least
not
initially
in
get.
What
I
want
to
do
is
I
want
to
put
it
in
tecton,
because
in
tecton
commit
to
dev.
A
I
want
another
step
or
another
task,
I'm
not
too
sure
another
step
that
hits
that
web
hook
and
then,
and
maybe
for
the
the
second
one,
the
second
reconciliation
yeah.
When
someone
commits
to
maine
or
whatever
branch
in
in
my
my
production
instance,
then
yeah
you
know
then
then
maybe
I
want
that,
but
I'll
kind
of
want
to
see
what
the
payload
that
looks
like
so
that
might
be
might
be
homework
on
my
end,
we're
running
at
a
time
here,
let's
see
here.
A
So
if
there's
any
questions,
I
don't
see
any
questions
here,
so
I
just
see
just
me
and
tiger
interacting
right
so
again
tiger
the
intern.
I
I
wonder
when
that
intern
is
going
to
drop
off
your
title
tiger.
Is
I'm
used
to
calling
you
the
tiger
the
intern
because
you're
actually
not
an
intern
anymore
you're
in
the
tiger?
The
engineer?
How
about
that?
How
about?
Let's
do
that
tiger?
The
engineer:
well,
I'm
gonna
transition
you
to
tiger
the
engineer,
yeah
being
captain
here
and
get
off
sky
to
the
galaxy.
A
I
can
make
those
promotions
here
so
tiger
you're
you're,
no,
no
longer
the
intern
you're.
Now
the
type
for
the
engineer,
so
cool
cool,
all
right
so
doesn't
look
like
any
questions
coming
down
we're
coming
up
short
on
time,
maybe
next
time
I
show
this
whenever
I'm
on
the
road,
probably
the
next
time.
I
show
this
I'm
on
the
road
I'll
have
those
web
hooks
showed
up
I'll
I'll
post
this
on
on
on
social
media.
A
So
speaking
of
social
media,
make
sure
to
follow
me
christian
christianh814
on
twitter,
I'm
also
on
github
the
same
handle
so
be
sure
to
follow
me
there.
I'm
always
posting
updates
to
shows.
Maybe
I
should
I've
been
wanting
to
start
a
blog.
The
thing
is,
I'm
a
better
talker
I
mean
you,
you
may
have
a
different
opinion,
but
a
better
talker
than
I
am
a
writer.
A
A
If
you
haven't
already,
I
appreciate
everyone
who
has
we've
gotten
some
good
good
feedback
so
far,
I'm
gonna
leave
that
open
to
the
end
of
the
year
updates
in
terms
of
shows,
let's
see
if
we
can
find
it
here
on
the
calendar.
I
don't
think
it
goes
that
forward.
Oh,
it
does
so
thursday
december
16th
get
ups
guide
to
the
galaxy
that'll
be
the
last
show
of
the
year.
A
We
are
going,
we're
we're
taking
a
a.
We
have
a
company
shutdown
here
at
red
hat
is
something
you
know.
We
do
a
recharge
day
every
few
few
days,
but
then
towards
the
end
of
the
year.
We
have
like
a
week
where
we
basically
do
a
shutdown,
skeleton
cruise
sort
of
thing
here
at
red
hat,
even
after
ibm,
we're
still
contr
continuing
that
for
that
tradition.
A
So
this
will
be
the
last
show
of
2021..
I
have
something
special
lined
up
for
you
right
so
for
those
who
are
watching
here,
I'm
gonna
be
doing
the
top
10
best
things
in
getups
been
going
through
kind
of
the
kind
of
some
of
the
cool
things
that
we've
done
through
the
year.
I
figured
kind
of
close
out
the
show
doing
something
really
really
cool
about
just
kind
of
talking
about
the
year
right
and
then
hopefully
talking
about
the
year
ahead.
So
so
with
that,
thank
you.
Thank
you.
A
Everyone
again,
member
reach
out
twitter.
If
you
ever
have
a
question
or
anything
like
that,
I'm
also
on
the
kubernetes
slack
same
same
username,
christianh814
find
me
using
that
handle
everywhere
and
yeah.
I
think
that's
it,
and
so,
like
I
close
out,
every
show
if
it
doesn't
get
it's
just
a
rumor.
Thank
you.
Everyone
bye
enjoy
your
day.