►
From YouTube: GitOps Guide to the Galaxy (E37) | GitOps with Harness
Description
How do you Harness your CI/CD? Today Dave Nielson Director of Developer Advocacy at Harness.io takes our GitOps Guides through Harness' platform and how ArgoCD-our favorite subject-- is key to them.
Want to learn more?
https://harness.io/
Try Red Hat OpenShift: https://red.ht/3sMaW7O
Follow us:
Christian Hernandez https://twitter.com/christianh814
Hilliary Lipsig https://twitter.com/Int2Caffeinate
Dave Nielsen https://www.linkedin.com/in/dnielsen/
OpenShift https://twitter.com/openshift
Subscribe to Red Hat's YouTube channel: https://www.youtube.com/redhat/?sub_confirmation=1
What is GitOps Guide to the Galaxy?
Every other Thursday at 3pm ET host Christian Hernandez and guests dig in to discuss everything in the GitOps universe, from end-to-end CICD pipelines to creating Git workflows.
Learn GitOps https://red.ht/3saqlzS
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 on redhat.com https://red.ht/3LGCIeD
OpenShift on GitHub https://github.com/openshift/
#RedHat #GitOps #Kubernetes
A
B
A
A
A
We
he's
a
senior
director
community
and
developer
advocate
he's
a
director
there
at
harness,
come
to
talk
about
get
ops
on
harness,
but
before
we
go
there
right,
while
we're
trying
to
figure
him
out
there,
he
is
it's
before
we
get
get
started
on
that
a
couple
things
I
want
to
talk
about.
I
want
to
talk
about
argo.
Con
argo.
Con
cfp
is
still
open,
so
I
will
put
that
on
chat.
A
Cfp
is
still
open
on
argo
con,
so
this
will
be
in
the
bay
area.
I
I
don't
know
if
we
have
the
I'm
looking
at
the
see
we
haven't
announced
where
it's
gonna
be
no,
not
yet
okay,
so
I
still
need
to
keep
hush
hush
on
that
parting
of
the
the
I'm
part
of
the
program
committee
there.
So
if
you
have
any,
if
you
guys,
are
doing
anything
cool
with
argo,
want
to
talk
about
argo,
be,
you
know
be
connected
with
other
people.
A
There
it'll
be
in
mountain
view
right,
so
it's
actually
down
the
street
from
where
dave
is
out
right.
Now,
argo
con
there
submit
to
speak
register
now.
Early
bird
registering,
I
believe,
is
still
open.
You
may
have
to
check
the
link
and
that'll
be
in
september
and
also
what's
coming
up,
which
will
be
fun,
which
can
be
a
lot
of
fun
is
getupscon.
A
So
if
you
are
in
in
the
europe
area
right
somewhere
in
europe,
our
get
ops
con
will
be
a
day,
zero
event
from
kubecon,
so
it'll
be
co-located
with
with
kubecon.
So
if
you're
in
valencia
area
or
if
you
can
get
to
it,
and
you
want
and
you're
going
to
kubecon
make
sure
to
join,
get
ups
con.
There
check
out
the
schedule.
Hillary
and
I
will
be
there
we'll
be
speaking
it'll-
be
awesome.
A
We
should
probably
work
on
our
presentation
and-
and
so
yesterday
was
may
the
fourth
right
and
we
actually
hillary
and
I
had
a
conversation
of
what
is
what
is
the
best
star
wars
movie
right
hillary
and
I
are
both
nerds
star
wars.
Nerds.
We
had
a
conversation
of
what
is
what
is
your
favorite
star
wars
movie
for
those
of
you
watching?
If
you
have
a
favorite
star
wars,
movie
drop
it
in
the
chat.
Maybe
we
could
start
a
flame
war.
A
A
B
My
favorite
star
wars
movie,
and
I
have
a
laundry
list
of
reasons
for
why
I
think
that
is
spoiler
alert.
Everybody
dies
right.
I
love
that.
It's
not
an
over
engineered,
happy,
ending
right.
The
plot's
not
over
engineered
the
special
effects
are
not
over
engineered
from
everything
from
like
the
writing
to
the
cinematography
that
movie
just
like
flowed,
and
it
worked,
and
it's
why
it
is,
in
my
opinion,
the
the
best
of
the
movies,
and
I
was
actually
doing
some
research
on
this
movie
the
other
day.
B
You
know
making
sure
that
I
had
all
my
arguments
for
this
yeah
and
it's
funny
because
some
of
the
things
I
liked
best
about
like
the
way
the
plot
developed
were
not
the
original
writing
from
what
some
of
the
reading
I
was
doing.
It
just
was
stuff
that
evolved
as
the
movie
was
happening,
which
that
kind
of
I'm
gonna
go
with
agility.
B
A
A
I
think,
like
everyone
say
like,
where
do
you
start
watching
star
wars,
and
I
always
say
well
now
that
rogue
one's
out
you
start
with
rogue
one.
You
definitely
have
to
start
with
rogue
one
and
then
from
there
you
can
watch
in
whatever
I
mean.
I
guess
you
would
want
to
watch
episode
four
right
after
rogue
one
ends
just
because
it
ties
it
ties
it
all
so
well
together,
but
it
was.
A
It
was
almost
it's
not
even
just
like
a
good
star
wars
movie,
it's
a
good
like
just
movie
in
general
right,
just
like
just
how
everything's
written
and
anyways
as
as
hillary
says,
we
could
probably
do
the
whole
hour.
Just
talking
about
like
why
this
why
this
is
we
we
were
actually
you
know
we're
having
our
planning
meeting
and
we're
talking
about
star
wars
and
they're
like
well.
What's
your
favorite
movie
and
like
we
just
happened
to
agree
that
was
rogue
one,
so
dave.
Do
you
have
a
favorite
star
wars
movie.
C
I
do
well,
I
wouldn't
say
favorite,
but
I'd
say
the
one
that
sort
of
surprised
me
the
most
was
actually
solo.
I
think
a
lot
of
people
sort
of
panned
it.
I
don't
think
they
necessarily
felt
there
was
consistency
between
the
han
solo
characters,
but
I
I
liked
it.
I
like
the
character
itself
just
for
its
own,
his
own
personality
sort
of,
like
I
thought,
he's
kind
of
quirky
and
fun,
and-
and
I
thought
the
the
movie
had
some
really
interesting
characters.
So
so
it
was
more
from
a
surprising
perspective.
C
C
B
A
C
A
A
C
A
A
Chose
rogue
rogue
one
is
another
good
one,
someone
guess
we
need
to
give
props
to
yoda.
I
think
there
needs
to
be
more
of
a
yoda
focus
in
the
new
one,
so
cool
cool,
so
anyways
again
we
can
talk
about
this,
probably
like
the
whole
hour.
So
just
to
recap,
remember:
argo
khan
remember
get
up
con.
If
you're
in
in
europe
come
say,
hi
I'll,
be
there.
Hillary
will
be
there
well,
actually
doing
a
presentation
together.
So
it'll
be
kind
of
like
on
the
road.
A
C
C
Yeah
we
were
gonna
throw
a
happy
hour
then,
but
we
noticed
that
there's
actually
kind
of
a
happy
hour
as
a
part
of
that
event,
so
we
moved
hours
to
monday
night.
So
if
you
show
up
on
monday,
we're
actually
listed
on
the
kubecon
schedule,
so
you'll
find
it
there,
because
we
paid
the
sponsorship
fee
to
list
our
our
event
and
it's
actually
on
the
beach.
So
it's
it's,
of
course.
A
There
you
go
awesome
so
yeah,
so
there's
plenty
to
do
so.
If
you're
there
at
kubecon
there's
plenty
to
do
all
of
us.
Are
there
so
cool
so
so
yeah
so
harness
right,
so
I've
been.
You
know
I
reached
out
to
today
because
I've
heard
about
harness
doing
some
things
with
around
get
ups
going
doing
some
things
around
argo
city.
So
I
figured
hey,
let's,
let's,
let's
bring
on.
A
Let's
share
the
love
this
you
know
the
all
the
argo
love
we
know
we
love
the
argo
here
at
red
hat
and
so
yeah.
So
dave
talk
to
a
little
bit
about
like
just
hardness
in
general,
where
you
guys
are
you
know
it?
I
know
it's
a
it's
a
sas
offering
but
kind
of
just
you
know,
let's,
let's
explain
it
to
me,
like
I'm
five
sort
of
sort
of
situation
for
at
least
for
my
benefit,
maybe
others
out
there
will
benefit
from
his
as
well.
C
Yeah
well,
first
of
all,
if
you've
read
the
news
recently,
you
may
have
seen
that
we
just
raised
a
monstrous
round
of
funding,
something
like
200
million
dollars
at
a
3.7
billion
dollar
valuation.
So
we
just
just
went
big
and
what's
what's
really
interesting
and
we'll
get
to
this
later.
But
it's
really
interesting
is
that
we
started
off
as
a
a
cd
company
so
like
we,
we
start
off
with
cd
first,
and
you
know
now
we're
providing
this.
C
You
know
platform
and
so
we're
kind
of
coming
back
full
circle
with
with
our
go
to
providing
that
as
a
service,
it's
kind
of
coming
back
to
our
roots.
So
but
that's
that's
sort
of
the
interesting
news
as
of
late,
but
I
guess
just
going
back
a
ways
by
the
way
I
head
up
community
and
developer
advocacy.
So
that's
that's
my
role
here.
I've
been
doing
that
for
quite
a
few
years.
C
C
You
know
a
programmer
software
engineer
and
he
was
one
of
the
early
folks
working
on
monitoring
and
agents
for
monitoring
that
kind
of
stuff,
and
there
was
this
kind
of
new
wave
of
innovation
around
monitoring
and-
and
he
left
his
previous
company
to
go
found.
A
company
called
app
dynamics
which
you
may
have
heard
of
and
app
dynamics
was
kind
of
app
dynamics
and
dyna
trades
were
like
the
two
big
fast
growing
sort
of
monitoring
companies
and
and
so
that
company
just
grew
and
grew,
and
they
were
about
ready
to
go
public.
C
Instead,
he
decided
to
take
the
money
he
earned
and
double
down
and
go
into
funding
startups,
including
his
own,
which
harness
was
one
of
them
and
harness
was
started
basically
from
the
idea
that
okay,
you're
monitoring.
But
what's
the
thing
that's
causing
you
the
most
pain
and
he
went
and
talked
to
his
customers
and
said:
okay,
I'm
you
know
looking
to
solve
some
other
problems
here.
What's
causing
you
the
most
pain
and
they
said
well,
ci
seems
to
be
working
cd,
not
so
much,
and
so
they
said
he
said.
Well.
C
If
I
go
solve
the
cd
problem,
that
means
you
pay
for
that,
and
they
said
yes,
and
so
he
went
all
in
on.
It
went
all
in
and
just
like
we're
just
gonna
go
solve
the
cd
problem
and
we
did
harness
has
an
awesome
cd
solution.
It's
got
a
lot
of
ai
based
in
it.
You
know
analyzing
logs
and
other
metrics,
to
use
machine
learning
essentially
to
identify.
C
How
can
we
tell
if
this
actually
deployed
successfully
and
if
it
didn't,
if
it
didn't,
meet
certain
criteria,
let's
roll
it
back,
and
so
that's
what
we
start
off
doing
we're
fantastic
we're
we're
the
best
in
the
business
at
it
and
that's
how
we
got
our
start
and
that's
probably
what
half
of
our
evaluation
is
is
is
the
fact
that
we
do
that?
Incredibly
well,
so
fast
forward.
C
C
A
C
A
Yeah
yeah,
it's
there
seems
to
be,
and
I
kind
of
want
to
get
your
get
your
opinion
on
this.
It
seems
to
be,
I
guess
now,
with
kubernetes
or
now
that
kubernetes
has
matured
a
lot.
A
lot
of
these
like
ci
solutions
based
on.
I
don't
want
to
say
serverless
but
kind
of
like
that
serverless
idea
of
like
hey.
You
don't
need
that
monolith
right,
you
just
kind
of
spin
up
a
container.
Do
your
build
or
do
whatever
you
need
to
do,
and
it
spins
down
right.
There's.
A
Tekton,
there's
things
like
argo
workflows,
there's
jenkins
x,
right
that
that
tries
to
do
some
of
that,
and
I
guess
drone
is
one
of
those
as
well
right,
and
I
just
I
just
find
it
interesting
that
all
these,
like,
I
guess,
as
it
shows
you
the
maturity
of
kubernetes
right,
it
kind
of
just
as
I
think
we're
at
a
point
where
kubernetes
matures
so
much.
I
like
all
these,
like,
oh
hey,
I
want
to
do
ci
now
with
containers.
Right
I
want
to
you
know,
do
my
builds
in
in
this.
C
B
C
That's
right
and
drone
was
the
first
to
recognize
that
containers
were
a
game
changer,
so
it
actually
launched
right
after
docker
did
it
was.
It
was
the
first
ci
product
to
take
advantage
of
docker
and
really
benefited
from
that
tremendously.
The
timing
was
excellent,
so
it's
interesting
go.
Take
a
look
at
it.
It
scales
out
horizontally
incredibly
well,
it's
very,
very
lightweight,
very,
very
thin.
I
came
from
redis,
which
is
very
similar
story.
C
You
know
salvatore
built
redis,
just
obsessed
with
making
people
happy
focused
on
keeping
it
simple
and
drone
has
a
very
similar
path
that
it's
taken
to
become
successful.
Brad
was
not
a
marketer,
he
was
just
a
geek
and
he
just
loved
solving
people's
problems,
responding
to
people's
questions,
and
you
know
the
the
the
chat
rooms
and
all
that
and
just
made
about
made
a
great
product.
So
we're
really
lucky
to
have
it
as
an
influence
here
inside
of
harness.
It's
definitely
like
dramatically
influencing
our
our
harness
ci
product.
C
In
fact,
we're
in
the
process
of
actually
merging
the
two,
which
is
sounds
difficult,
but
drone
was
acquired
early
enough
in
the
harness
ci
development
that
it
already
impacted
the
development
quite
a
bit.
So
there's
not
as
much
to
it's
still
difficult
but
there's
still
not
as
there's
not
as
much
as
if
we
took
two
completely
different
products
and
try
to
merge
them.
C
So
that's
the
drone
story
and
then
from
there
it
was
again
just
listening
to
customers.
We
started
adding
more
features.
One
custom,
one
one
feature
I
thought
was
really
interesting
is
called
our.
It's
called
cloud
cost
management
it
I'm
a.
I
am
a
huge
fan
of
heroku
from
early
days.
I
thought
that
it
was.
I
still
to
this
day
think
it
was
the
most
influential
cloud
technology
ever
and
by
the
way
I
was
I
was
there
in
the
very,
very
infinitesimal
stage
of
openshift.
A
A
That
makara,
that
did
a
lot
of
and
and
I
agree
with
you-
I
think
kuroku-
doesn't
get
like
enough
credit
to
what
they
did,
especially
with
the
end
youth,
like
user
experience,
very
important
right.
Basically,
I
just
like
run
a
few
commands
and
I
can
see
my
application
up
and
running
like
at
that
time.
That
was,
like
you
know,
right,
mind,
blown
and
which
is,
which
is
why
I
like
openshift.
A
C
My
that's
my
background.
Yeah
you're,
probably
more
you're,
probably
more
accurate
as
far
as
influence
than
I
am,
but
but
there
is
a
side
story
to
this,
that
the
very
very
beginning
june
24
2008.
We
ran
something
called
cloud
camp
and
cloud
camp
was
a
super
early
event
in
the
cloud
space
and
there
was
a
guy
two
two
guys,
isaac,
roth
and
I'm
gonna
blank
right
now
I
don't
know
about
blank
on
his
name.
Oh
you
know
come
to
me
just
a
second
he's,
a
good
friend
of
mine.
C
Actually,
but
anyway,
so
isaac
roth
and
they
had
a
company
that
they
had
just
raised
some
money
for,
and
they
didn't
have
a
name
yet
so
of
all
things
they
they
said.
We
want
to
sponsor
your
event
and
they
didn't
have
a
name
for
their
company,
so
I'm
like
well,
you
gotta
come
up
with
a
name
and
they're
like
oh
tobias,
coons
there
you
go
that's
his
other
guy,
the
other
guy's
name,
and
so
they
came
up
with
a
name
right
there
on
the
spot.
C
C
C
C
To
give
you
a
free
cloud,
you
know,
or
free
instance,
was
that
if
you
weren't
using
it,
they
would
shut
it
down,
and
I
love
that
because
that's
how
everyone
got
to
have
free
right
and
instead
of
having
those
containers
always
running,
and
so
they
would
shut
it
down,
which
is
great
for
demos
and
students
and
all
that
kind
of
stuff
startups.
You
know,
so
we
have
a
feature
in
cloud
cost
management
which,
if
you
are
managing
your
instances
and
an
instance,
is
not
being
used.
C
You
know,
maybe
it's
something
that
you
don't
really
care
if
it
shuts
down
when
it's
not
being
used,
will
automatically
shut
it
down
for
you
based
on
whatever
time
constraints
you
you
have,
and
it
saves
people
so
much
money.
It's
amazing.
You
know
amazon
says
that
40
of
all
of
their
instances
are
not
being
used.
C
Size
of
the
market
that
our
product
can
go
after
right,
like
there's
that
much
waste
out
there.
So
I
love
that
feature
and
then
we
came
out
with
feature
flags,
and
then
we
acquired
companies
to
help
us
create,
let's
see,
what's
called
service
resource
management.
So
like
there's
our
service,
reliability,
management
and
another
one
is
a
security
testing
orchestration
and
then
we
acquired
another
company
called
chaos
native,
which
is
actually
a
cncf
company.
They
have
an
open
source
project
in
the
cncf
called
litmus
chaos.
C
We
just
recently
acquired
them.
So
it's
been
kind
of
like
this
build
and
acquire
to
growth,
and
we
have
now
this
full
platform.
But
you
know:
along
comes
argo
cd.
Actually,
along
comes
git
ops.
I
should
say
right.
So
along
comes
git
ops,
which
that
cloud
camp
event
that
I
ran
back
in
june
2008,
one
of
the
co-founders
of
cloud
camp,
is
a
guy
named
alexis
richardson.
C
Lexus
has
gone
through
many
lives
since
then
you
know
he
founded
rabbitmq
and
of
course,
then
weave
works
and
you
know
was,
was
one
of
the
key
people
behind
get
ups,
and
so
he's
always
had
this
sort
of
very
creative
career
and
I've
watched
what
he's
been
doing
so
I've
been
watching
git
ops
for
a
while.
I
you
know,
like
I
had
at
the
10-year
anniversary
of
cloud
camp,
which
was
in
2018,
had
one
of
their
engineers
come
and
talk
about,
get
ops
and
flux,
and
so.
C
A
A
That's
that
whole
and
then
there's
a
whole
auto
healing
self-healing.
You
know
reconciliation.
There
is,
you
know
a
lot
of
people
try
to
glom
together.
Like
you
know,
oh
what
you
know,
why
do
I
need
you
know
getups
if
I'm
already
doing
ci
cd,
it's
like!
Well,
it's
it's!
It's
that
extra
last
step.
It's
the
other
cd
right,
as
you
know
that
atlassian
talking
about
like
ci
cd,
continuous.
C
And
that's
that's
really
what
happened
like
all
of
the
ci
companies
decided
they
wanted
to
do
cd,
but
they
did
the
first
cd.
They
did.
You
know
continuous
deployment,
which
basically
means
I've
created
this
stuff
and
it
will
run
in
a
production-like
environment,
but
they
didn't
actually
do
all
the
testing
and
all
that
kind
of
stuff
right.
So
you're
you've
got
the
resources
and
the
assets
and
they're
ready
to
go,
but
you
still
have
to
go
put
them
over
there
and
and
make
sure
it
works.
C
C
C
There
but
I
like
it,
because
it's
like
a
forcing
function
for
ops
to
say,
wait,
wait,
wait!
No,
you
can't
do
that
and
then
you
say
why
not
right
and
then
ops
will
say,
but
but
but
and
then
the
reality
is
that
they
have
a
legitimate
concern.
You
know
there's
definitely
times
when
you
don't
want
to
just
yolo
into
production,
but
they
need
to
explain
themselves
right.
C
They
need
to
come
to
the
table
and
say
here
are
the
things
that
we
need
to
be
a
part
of
this
deployment
process
and
then
that
can
go
into
you
know
the
get
ops
process
that
could
be.
That
can
become
a
part
of
it
right.
Some
of
the
checks
and
balances
that
you
need.
So
I
love
it
as
a
forcing
function.
B
C
B
Now
we
basically
said:
there's
a
second
cd
right.
That's
like
a
really
great
way
of
putting
it,
because
there
was
not
cohesion
in
the
industry
around
what
cd
really
meant
is
it
that
you're
continuously
deploying
or
that
you're
continuously
delivering
that
code
to
end
users
and
tools
like
argo
cd
and
just
like
git
ops
in
general
kind
of
help
solve
some
of
the
most
technical
challenges.
To
doing
that,
one
is:
what
is
your
consistency
of
operations?
B
That's
really
important,
I
don't
know
about
other
people,
but
I've
actually
had
that
stuff
not
kept
anywhere
in
like
a
git
repository
and
then
it
actually
caused
production
failures
because
it
turns
out
there
were
resource
differences
between
the
environments
that
were
not.
They
were
not
codified
or
recorded
anywhere
and
github
solves
that,
and
that's
that's
very
nice,
because
you're
guaranteed
to
know
that
what
you're
testing
is
the
right
things
so
from
a
quality
perspective
and
from
a
reliability
engineering
perspective
I
can.
B
I
can
take
a
look
at
what's
in
there
and
say:
okay
yeah.
This
is
this
makes
sense.
This
looks
good
based
on
everything
we
know
about
these
workloads.
This
is
going
to
be
sustainable
and
resilient,
and
so
the
last
part
one
of
the
hardest
parts
that
I
like
the
most
about
get
ops
tools
and
christian
can
speak
to
this
way.
Better
is
the
like
codification
of
like
tagging
and
cutting
and
naming
releases.
B
B
A
A
Which
is
a
get
ops
controller
right
for
us
it'll,
be
our
argo
cd
right,
a
lot
of
people
use
flux,
whatever
you're
using
doesn't
matter
it's
it's
your
you
have
that
reflection
right
and
it's
like
well
what's
what's
in
git,
especially
if
you're
using
things
like,
because
you
can
like
use
tags
to
deploy
right
with
argo
cd.
Not
just
like
you,
don't
just
point
to
a
branch
I
mean
you
can,
but
you
can
also
like
point
to
a
tag.
A
So
a
specific
tag
so
like
which
I
think
is
better
because
you
have
like
an
actual.
You
know
the
tags,
the
tag
right
and
you
know
right
like
in
the
only
way
you
know
to
update
that
is
to
change
that
tag
reference
somewhere
in
in
the
configuration.
So
so
we
do
have
a
question.
Don't
understand
the
question,
but
maybe
it'll
push
the
conversation
along.
The
conversation
is:
does
the
solution
have
any
intellectual
use
cases
and
then
they
correct
it?
Saying
intellectual
property
use
cases.
B
A
C
A
A
C
C
C
C
The
rest
is
going
to
go
into
production
and
so
that's
a
different
way
of
doing
things
and
while
for
a
while,
we
were
thinking
you
know
we
like
the
way
we
do
it,
because
it
gives
you
a
little
bit
of
extra
control
and
customization,
and
you
know
some
enterprise.
Customers
are
absolutely
not
ready
to
just
take
stuff
from
a
git
repo
and
put
it
into
production
so
like
having
some
extra
steps
that
require
intervention,
or
at
least
some
checks
and
balances.
C
You
know
throughout
the
deployment
process
is
really
what
they
feel
comfortable
with,
but
there's
just
a
ton
of
use
cases
out
there,
where
a
helm
chart
is
good
enough,
like
where
you
know
the
the
get
ops
process
will
work
perfectly
well
and
so
recognizing
that
we
feel
that,
like,
let's
not
miss
the
boat
like
this,
this
ship
is
often
you
know,
heading
off
into
sea.
Like
you
know,
we
better
take
off
now,
otherwise
we're
never
going
to
catch
it.
C
C
We
think
some
other
people
will
be
attracted
to
get
ops
because
it's
catching
on
and
they
think
that
it's
going
to
solve
their
problem
and
then
once
they
get
into
it,
they
realize
they
have
some
intricacies
that
gitops
hasn't
solved
yet
and
so
we'll
have
a
solution
for
them
as
well.
So
you
know
we
that's
what
we
think
and
I
think
that's
going
to
work
out
well
for
us.
I'm
part
of
the
more
of
the
get
ops
process
like
I'm
working
with
the
community.
C
That's
what
we
do
in
our
on
the
community
side
of
harness
we're
just
getting
started.
So
it's
sort
of
a
new
community,
but
we're
using
our
existing
drone
community
to
sort
of
kick
start
that
and
so
we'll
be
putting
together.
You
know
drone
plus
argo
sort
of
processes
and
educating
people
about
that.
C
Both
are
open
source
and
then,
if
you
want
sort
of
like
the
enterprise
offering
that
harness
offers
as
a
service,
then
of
course
you
can
use
that
from
us
as
well,
and
we
also
recently
very
recently
offered
our
cd
product,
which
is
not
the
git
ops
product,
but
our
regular
cd
product
as
software,
it's
not
open
source,
but
it's
source
available,
it's
very
similar
to
open
source.
But
it's
you
can't
use
it
to
like
go
offer
your
own
offering
to
other
customers,
but
you
can
certainly
use
it
for
whatever
you
want
for
yourself.
C
You
know
and
offer
it
to
customers
inside
your
company.
So
that's
called
harness
cd
community
edition,
but
that
is
not
get
ups.
That
is
that's
our
traditional
pipeline
product
and
I
think
in
the
future.
You
know
we'll
see
how
it
goes,
but
I
can
imagine
that
we
might
have
a
harness,
get
ops
community
edition
as
well.
C
So
anyway,
that's
kind
of
the
starting
point
for
the
whole
thing,
and
you
know
beyond
that.
If
you
want,
I
can
go
into
more
detail,
but
that
kind
of
gives
you
the
the
the
background
of
where
like.
Why
would
a
company
who's
already
doing
cd
really
well
very
successful,
want
to
then
take
on
a
new
type
of
cd
which
the
get
ops
methodology
and
that's
just
because
of
customer
demand.
A
Yeah-
and
it
also
shows-
I
think,
like
like
you,
said
customer
demand
and
it
shows
I
think
more
again.
I
go
back
to
the
maturity
right,
the
maturity
of
of
of
kubernetes
right
and
maturity
of
cloud
native
software
right.
It's
it's
no
longer.
You
know
it's
no
longer
things
that
people
are
testing
on.
It's
things.
People
are
deploying
on
like
in
in
product.
I
mean
hillary
can
tell
you.
Hillary
deals
with
this
all
all
day.
Every
day
right,
I
don't
know
if
she
might
even
be
on
call
right
now.
C
A
You
know
you
have
the
you
know,
I
I
always
people.
We
always
talk
about
get
ops
as
a
new
thing,
and
it's
not
necessarily
a
new
thing.
It
just
has
a
new
name
to
it
right.
It's
just.
You
know
we're
just
kind
of
putting
a
name
to
something
and
we're
kind
of
just
packaging
it
up
now,
as
a
like
a
solution
now
versus
where
it's
like.
A
A
C
Reminds
me
a
little
bit
going
back
way
back
to
the
like
again
like
2008,
when
we
were
starting
up
cloud
camp,
there
was
a
another
person
you
might
recognize.
You
should
know
him
named
john
willis,
who
was
running
around
as
a
part
of
that
early
gang
and
and
he
loved
to
tell
the
story
about
how
he
was
an
I.t
person,
and
you
know
the
I.t
people
kind
of
come
out
of
they're
like
cicadas.
B
I
I'm
actually
talking
about
exactly
that
in
our
upcoming
get
ups
contact.
I
actually,
I
have
a
slide
about
this
like
and
then
like
I'll,
be
posting
like
who's
deployed
production
code
by
hand
right
one
of
my
old
make
files,
which
is
still
like
very
modern,
compared
to
how
we
did
it
at
times
right,
which
would
would
go
do
like
okay.
If
I
want
this
to
build
here,
I
have
to
do
these
these.
You
know
bash
things.
If
I
want
this
to
do
go
over
here
and
in
production,
it's
better.
B
C
C
You
know
and
if
you're
lucky
they
put
it
on,
like
a
network
drive
or
something
like
that,
so
that,
if
you
know
anything
ever
happened
to
them,
that
you
can
never
get
in
and
deploy
your
application
again.
So
there
was
like
some
script
somewhere
and
then
all
of
a
sudden
people
were
like
you
know,
maybe
there's
a
better
way
and
let's
take
that
script
and
put
it
into
a
git
repo
or
you
know,
and
now
it's
kind
of
like
back
in
the
the
cf
engine
or
or
puppet
or
chef
days
right.
But
at.
C
C
A
B
Was
like
a
feature
that
got
added
like
oh,
you
can
trigger
another
job
from
this
first
job
right.
It
was
had
to
do
all
of
this.
You
had
jobs
running
on
like
oh
well.
This
will
take
roughly
this
amount
of
time
and
we
want
to
do
it
this
often
so
we're
gonna
have
this
other
job.
That
runs
like
every
three
hours
to
do
this
right.
B
C
Right-
and
I
think
you
know
from
a
harness
perspective,
we
we
have
a
conference
in
september
14th
that
we
just
announced
a
call
for
papers,
for
it's
called
unscripted
and
the
whole
reason
for
unscripted
is
that,
like
writing?
All
those
scripts
was
great.
You
know
10
years
ago
that
was
like
cutting
edge,
but
do
you
really
want
to
create
a
whole
bunch
of
scripts?
C
These
days,
like
there's,
been
enough
evolution
where
the
platform
keeps
rising
and
eating
up
more
of
your
scripts
and
so
like
there's
just
not
as
much
need
to
write
scripts
anymore,
because
you
know
we've
got
kubernetes
and
all
that
which
solve
a
lot
of
those
problems
just
out
of
the
box,
and
so
you
know
we
we
think
it's
time
to
target
having
no
scripts.
You
know
it
might
not
be
a
reality
today,
but
it's
certainly
something
that
we
think
should
be
a
goal.
C
B
A
B
A
C
A
A
Yeah
yeah,
I
know
we
have
we
have
some
time
here
and
while
while
dave
is,
is
doing
that,
if
you
have
any
questions,
feel
free
to
drop
that
into
the
chat
here.
People
talking
about
not
even
sftp
but
ftp.
That's
that's!
That's
how
that's
all
old
school
some
people
have
done
their
deployments.
Ftp
put
I've
written
a
lot
of
a
lot
of
ssh
loops
in
in
my
in
my
time.
So
all
right
cool
got
pictures.
Pretty
pictures.
C
A
C
Apologize
for
not
knowing
exactly
which
components
but
we've
taken
those
components
within
argo,
cd
and
we've,
basically
just
wrapped
it
up
right,
because
we
don't
want
to
change
argo
cd.
We
want
to
be
able
to
evolve
with
argo
cd.
So
basically
it's
a
very
simple
process.
Well,
it's
simple
on
the
outside
it's
hard
inside
right,
because
we've
got
to
match
up
the
argo
cd
functionality
to
our
ui
right,
so
our
existing
users
will
find
it
familiar
and
also
we
want
argo
cd
users
who
want
to
use
ui
to
find
it
familiar.
C
So
it's
kind
of
actually
magic
from
that
perspective,
and
I
don't
have
a
screenshot
of
that
to
show
you.
But
you
know
that's
what
you'll
see
from
within
the
harness
application
and
then
from
then
from
there
on
out
you're
you're,
basically
filling
out.
You
know
forums
and
not
too
many,
thankfully,
but
you
fill
out
a
few
forms
you're
specifying
you
know
the
the
repo
the
the
cluster
you're
you're
authenticating.
C
You
know
all
that
kind
of
stuff
that
helps
and
from
a
visual
interface
inside
of
our
application,
provide
the
content
or
the
the
the
attributes
and
the
and
the
data
that
you
need.
The
metadata
that
you
need
in
order
to
let
argo
cd
do
its
job
and
then
then
we
do.
What
cd
does
we?
We
monitor
the
desired
state
in
the
git
repo.
We
we
pull
from
that
git
repo.
We,
you
know
push
it
into
the
target
cluster,
which
one
thing
that
we
do,
which
is
kind
of
interesting,
is
from
within
harness.
C
C
It's
a
single
panel:
that's
what
customers
want
they
don't
want
to,
like.
You
know,
bounce
around
to
different.
You
know
tabs
and
all
that
kind
of
stuff.
So
you
know
we've
done
that's
where
a
lot
of
our
work
has
been
is
really
creating
like
a
multi-tenant
argo
interface,
so
that
they
can
quickly
and
easily
see
all
their
different
applications.
C
A
C
Add
that
at
the
end
of
the
get
ops
like,
maybe
it
goes
into
production,
but
it's
maybe
in
like
a
blue
green
environment
or
something
like
that,
so
it
did
the
cd
process
or
the
argo
cd
process,
but
it's
still
not
live
as
far
as
to
end
users.
Yet
and
then
it
goes
through
those
checks
like
then
we
can
add
in
additional
checks
and
then
that's
when
it
gets
exposed.
I
think
that
might
be
one
way
to
to
have
the
best
of
both
worlds.
A
So
one
one
thing
about
the
design
that
I
have
a
question
about
what
hardness
is
is
that
is
that
something
that's
hosted,
meaning
that,
like
let's
say
I
have
like
three
kubernetes
clusters.
I
don't
have
to
install
argo
on
those
clusters
right.
I
can
just
use
the
the
platform
that
has
argo
in
it
and
it'll
deploy
the
application.
C
A
C
A
A
Okay,
okay,
I
was
under
another
impression,
all
right
cool.
Thank
you
for
that
clarification,
awesome
and
also
you
said
easy
on
the
outside
hard
on
the
inside.
I
think
that's
hillary's
life
right
to
to
to
to
end
users
right
to
like
end
users.
It's
just
like
it
just
looks
like
magic
and
then
hillary's
out
here
pulling
her
hair.
C
A
B
C
C
A
A
Yeah-
let's
see
here-
oh
there,
it
is
yeah
awesome
I'll
put
that
also
here.
So
that's
the
get
up
beta
check
that
out
here
running
argo
cd
again
we
love
rwc.
That's
why
I
wore
the
shirt
we'll
be
in
valencia,
so
we
are
actually
on
break.
Usually
we
do
it
every
other
thursday
here,
but
that
phone
falls
under
kubecon
and
neither
hillary
nor
I
will
be
in
the
states
at
that
time.
So
we're
skipping
that
date
so
and
also
we'll
give
you
the
week
off.