►
From YouTube: GitOps Guide to the Galaxy (Ep 53) | Akuity on OpenShift
Description
Akuity accelerates and simplifies the implementation of GitOps practices into Kubernetes cluster management by using the Argo ecosystem. In this episode we explore how Akuity's platform looks on OpenShift with Nick Morey, developer advocate at Akuity.
Follow Nick Morey on Twitter | @morey_tech
Follow your Hosts on Twitter:
Christian Hernandez | @christianh814
Hilliary Lipsig | @Int2Caffeinate
A
B
A
A
and
I
mean
if
you,
if
you
just
like
search
on
LinkedIn
or
Twitter,
the
hashtag
IBM
tech2023
like
you'll,
see
the
pictures.
The
event
was
I,
just
the
most
people,
I
think
I've,
seen
in
Forever,
just
like
1300
of
us,
and
so
it
was.
It
was
great.
There
was
I
forgot,
the
name
of
the
person,
but
her
she's,
a
space
architect
came
and
talked
to
us,
and
it
was
overall,
very
cool.
A
A
highlight
highlight
level
takeaways,
IBM
loves,
openshift,
no
surprise
and
IBM
loves
ansible,
and
so
that
was
also
not
really
a
surprise
for
me.
Being
from
from
the
service
delivery
and
ansible
organization,
and
also
you
know,
loving
ansible,
but
so
it
was
great
to
see
all
the
different
ways
of
using
ansible
and
then
they
talked
a
lot
about
get
Ops
and
I
took
the
security
track.
A
So
there
was
like
you
know,
seminars
on
security
and
devsec
Ops,
and
so
hopefully
we'll
get
some
folks
from
IBM
to
come
in
later
months
to
actually
talk
to
us
more
about
how
they're,
using
these
things
but
yeah.
So,
like
super
professional
voice,
super
professional
face
I
had
to
wear
like
actual
business
clothes
for
the
first
time
yeah
they
said
the
dress
for
the
event
was
business.
Casual
and
I
was
like
you
guys
have
not
been
to
San
Francisco.
Obviously,.
A
A
B
So
this
is
the
co-located
event
with
kubecon,
so
for
those
who
are
either
going
to
travel
to
kubecon
or
for
those
in
the
area
in
in
Europe
and
nearby
the
cfp
for
that
closes
on
Sunday,
but
just
pretend
it's
Friday
for
the
cfps,
so
it's
co-located
with
kubecon
and
you
know
we're
looking
for
you
know
now
that
the
kubecon
schedule
is
out.
You
guys
can
now
I
know
everyone.
Does
this
I,
don't
know
why
I
don't
do
it
so
hit
me
up
on
Twitter?
B
If
you
think,
if
you
disagree
with
me,
but
if
a
lot
of
people
wait
until
kubecon
schedules
out
and
then
they
submit
their
cfps
for
the
co-located
event
by
the
way
I'm
part
of
the
program
committee
for
argocon
and-
and
we
know
if
you
double
submit
so
you
so
you
waiting
it,
it
doesn't
do
any
good
so
so
that
closes
on
Sunday
Friday
get
Ops.
Con
cfp
closes
right,
so
get
Ops.
B
Con
is
now,
if
you
haven't
heard
co-located
with
open
source
Summit,
so
it's
going
to
be
in
Vancouver
I'm
excited
to
get
back
to
Vancouver.
It's
been
too
long
since
I
visited
Canada
Vancouver's,
probably
without
visiting
other
cities.
I
may
visit
all
the
major
cities
in
Canada,
but
Vancouver's.
It's
still
my
favorite,
so
I'm
super
excited
to
go
to
open
source
Summit.
B
If
you
have
an
Argo
or
anything
related
to
get
Ops,
please
submit
to
get
Ops
Khan
so,
and
you
know
before
he
sits
there
awkwardly
for
too
much
longer.
I
want
to
introduce
Our
Guest
today,
so
I
want
to
introduce
Nick
Mori
right.
So,
as
we
say
right,
if
you
could
say
in
French
right
so
who's
he's
a
developer
Advocate
at
Acuity,
so
Nick.
How
are
you
doing
today.
C
B
C
C
C
B
B
I
think
awesome
awesome,
so
so
yeah
so
Nick
comes
from
us
from
Acuity
right
so
cutie
by
the
way
I'm
wearing
which
side
of
my
own
there
we
go
I'm
wearing
the
Cutie
shirt
right.
So
I
got
this
at
the
the
argocon
North
America,
one
of
my
favorite
shirts,
the
red
one
and
the
blue,
one
that
you
have
on
I'm
gonna
collect
them
all
right.
You
guys
have
cold
shirts.
The.
A
B
A
B
B
B
It's
gonna
be
an
episode
just
showing
Spike
so
yeah
so
from
Acuity
right
and
so
Acuity
right,
startup
Right
started
by
the
creators
of
the
Argo
project
right,
so
that's
kind
of
that's
kind
of
cool
I've
been
wanting
to
get
you
guys
on
for
a
little
bit
last
time.
I
last
time,
I
talked
to
an
Acuity
guy
Alex.
He
wasn't
into
it.
B
C
But
the
idea
is
that
they
wanted
to
provide
Argo
CD
as
basically
Enterprise
ready
and
production
ready,
similar
to
kind
of
how
openshift
provides
kubernetes
as
production
ready
environments,
except
for
you
know
it
has
all
of
the
security
considerations
the
scalability
and
it
you
basically
don't
have
to
think
about
it
as
much
and
just
to
answer
the
question
there
right
now.
The
Acuity
platform
is
all
Argo
CD
based,
but
the
company
is
still
generally
focused
on
the
Argo
project
as
a
whole.
C
B
Yeah
yeah,
so
what
what
I?
Really?
What
was
really
cool?
What
I
saw
on,
and
maybe
you
can
talk
about
it
a
little
bit
and
then
and
then
you
can
just
let
me
know
when
to
bring
the
screen
sharing
on,
but
on
Twitter,
and
it
was
mentioned
on
Twitter
and
you
kind
of
chatted
me
about
it.
It's
like,
as
you
were,
playing
with
openshift
and
Acuity
you
there
was
like
there
was
an
update
made
to
the
Acuity
platform
to
to
better
integrate
with
openshift,
which
I
think
is
really
cool.
B
I
I
told
you.
The
system
works
right,
like
it's
kind
of
cool.
To
get
that
you
know
in
open
source
right
like
you,
you
you
make
a
lot
of.
You
know
these
discoveries
on
these
fixes,
just
by
the
nature
of
just
like
interactive,
which,
if
you
talk
a
little
bit
about
a
little
bit
about
that,
and
then
just
kind
of.
Let
me
know
when
to
like
bring
the
screen
on.
C
So
you
know,
through
my
through
my
testing.
We
we
ended
up
finding
what
we
needed
to
make
the
agent
manifest
compliant
with
openshift,
and
we
just
applied
a
patch.
So
now
it
just
works
by
default.
You
don't
have
to
go
and
like
patch
the
agent
manifest
just
to
be
able
to.
You
know
for
all
the
pods.
To
start
up
correctly,
so
if
we
want
to
jump
into
the
demo,
I
can
start
showing
that
off
so.
A
C
Yeah
definitely
so
starting
off
we're
on
openshift
dedicated
here.
This
is
my
cluster
that
got
deployed
onto
onto
G,
which
was
a
really
fun
experience.
I
set
up
some
like
basic
terraform
to
do
like
the
project
and
surface
account
creation
and
then
just
providing
that
service
account
credentials
to
openshift
dedicated
and
it
went
and
did
everything
else
like
you
know,
I
I'm,
pretty
sure
this
isn't
a
basic
gke
installation.
This
is
like
compute
nodes
and
openshift.
Does
all
the
bringing
the
nodes
together
right.
A
B
A
A
C
Very
cool,
so
I
I
think
I
like
that,
because
I've
worked
with
gke
pretty
extensively
and
you
can
kind
of
hit
in
some
into
some
limitations
with
it,
especially
with
scaling
the
the
the
the
master
nodes
and
and
the
API
server
I've
run
into
some
limitations.
With
that
so
I
imagine
taking
the
just
taking
the
pure
compute
resources
gives
it
a
lot
more
flexibility
and
how
you
want
to
manage
kubernetes
or
openshift
on
the
on
Google
Cloud
yeah.
A
A
You
end
up
with
the
same
like
feature
parity,
but
there's
a
lot
of
development
in
various
teams
that
are
actually
going
into
actually
making
that
happen
to
ensure
that
we're
able
to
best
utilize,
whichever
Cloud
platform
or
hyperscaler
we're
interacting
with
and
all
that
goes
into
even
more
cooler
stuff
on
like
some
of
our
future
things,
and
if
people
are
paying
attention
around,
oh
I,
don't
know
open
source,
Summit
or
red
hat
Summit.
You
might
have
be
hearing
people
talking
about
these
things
from
Red
Hat
and
especially
at
this
coming.
C
Okay,
good
to
know
I
yeah,
it's
interesting,
the
the
there
must
be
some
pretty
like
intricate
nuances
between
deploying
to
the
different
Cloud
platforms
like
they
try
to
say
it's.
Oh
it's
all
just
Cloud.
It's
all
just
you
know
compute
resources,
but
interacting
with
them
must
be.
It
can
be
a
challenge.
A
B
C
All
right
so
yeah
we
we've
got
our
our
open
shift
cluster,
and
so
what
I'll
do
is
I'll
swap
over
to
to
the
Acuity
platform
here
so
I'm
already
logged
into
my
my
organization,
my
Mori
Tech
organization
and
basically
I'm
going
to
start
by
just
spinning
up
a
new
Argo
CD
instance,
so
that
you
know
this
might
be
one
of
the
first
differences.
You'll
see
between
the
open
source
and
and
a
SAS
platform.
C
C
And
then
that
way,
you
really
only
need
egress
traffic
out
to
the
internet
allowed
for
you
to
manage
that
connected
cluster
from
a
central
Argo
CD
control
plane.
So
it
could
be
across
multiple
Cloud
providers.
It
could
be
a
kind
cluster
on
my
laptop,
or
in
this
case
it
could
be
openshift
dedicated
in
gcp,
without
any
special
networking
required
to
expose
the
API
server
to
the
Acuity
platform.
C
So
you
know,
while
I
was
talking
through
that
I
provisioned
a
new
agent
for
the
cluster
and
for
the
sake
of
this
demo,
I'm
just
going
to
download
the
agent
manifest,
but
it
does
present
present
you
with
a
a
single
command
that
you
can
paste
with
your
Cube
CTL
context
already
configured.
You
can
paste
that
to
deploy
the
agent,
but
in
this
case
I'm
going
to
open
up
the
downloaded
file
and
yeah.
If
you
can
get
it
fast
enough,
steal
that
token,
and
then
you
might
get
something
fun
deployed
to
your
cluster.
C
So
now
that
I've
got
the
agent
manifest
downloaded
I'm,
just
going
to
go
back
to
the
openshift
dedicated
platform
here
and
I'm
just
going
to
import
the
yaml
directly.
So
these
are
all
the
manifests
that
that
make
up
the
agent's
deployment,
probably
no
need
to
poke
through
these
right
now,
because
once
I
hit
create
here,
it's
going
to
go
and
display
all
the
resources
that
that's
being
created
so
we'll
just
give
that
a
sec
they're
all
created
great.
C
So
if
we
go
over
to
workloads
deployments
and
I'm
going
to
filter
it
down
by
the
the
Acuity
project,
yeah,
that's
the
cool
part
is
that
each
agent
is
is
getting
its
own
token
to
connect
back
so
worst
case.
If
you
get
it,
I'm
just
gonna
go
and
delete
it
or
I'm
going
to
deploy
some
cool
stuff
to
your
cluster.
C
Yeah,
so
you
know
just
to
talk
about
a
little
bit
that
patch
that
we
created
to
make
the
agent
manifest
work
directly
with
openshift.
The
problem
I
had
run
into
is
that
redis
didn't
want
to
start
because
it
had
a
custom
uid
set
for
the
the
pod
which
the
default
security
context
constraints.
We're
not
happy
with,
and
I
had
to
either
go
and
create
a
service
account
and
assign
that
to
the
any
uid
security
context
constraint.
C
C
In
this
case,
I
can
just
go,
pull
it
from
the
UI
of
the
Acuity
platform
and
I'm
just
going
to
generate
a
new
password,
because
we
obviously
aren't
going
to
go
and
store
the
admin
password
for
you,
but
what
I
can
do
is
just
regenerate
it.
Every
time
I
need
access
to
the
admin
user
in
reality,
you're
going
to
go
and
you're
gonna
yeah
right.
That's
my
favorite
feature.
B
B
C
Here,
yeah
yeah
and
in
reality,
you're
gonna
go
and
you're
gonna
set
up
s's.
So
you
know,
if
I'm
going
to
leave
this
Argo
CD
instance
up
for
any
longer
than
this
demo.
We
would
be
configuring,
SSO
and
just
logging
in
that
way,
but
let's
just
go
and
copy
that
password
and
so
as
a
part
of
the
Acuity
platform.
C
B
C
That's
the
thing
is
it's
still
using
the
the
same
configuration
that,
if
you're
already
using
open
source,
so
basically
you
can
go
and
take
your
existing
config
map
for
for
SSO
and
just
go
paste
these
parameters
into
here.
That's
pretty
cool
and
it'll
also
store
like
the
secrets
that
you
need
for,
like
the
client,
ID
and
the
client
secret.
You
can
go
pull
that
from
the
Acuity
platform
instead
of
having
to
manage
it
in
your
say,
management
cluster,
that's
hosting
or
go
CD,
so
yeah
going
back.
C
B
A
C
Yeah,
so
for
just
to
kind
of
talk
about
some
of
the
details
of
the
this
application
deployment.
So
in
the
Acuity
platform
we
saw
that
I
provisioned
a
cluster
with
a
specific
name.
Well,
when
I
go
to
deploy,
my
application.
I
can
use
that
name
to
to
describe
where
I
want
to
go
and
deploy
the
resources
for
this
application.
So
the
the
destination
is
the
same
as
the
the
agent
name
here.
C
C
So
the
the
interesting
part
here
is
that
the
actual
application
custom
resource
is
getting
deployed
to
the
Acuity
platform
and
then
only
the
resources
that
are
generated
by
that
application
end
up
in
the
connected
cluster,
so
you're,
not
adding
all
of
the
Argo
CD
custom
resource
definitions
to
every
connected
cluster.
The
Acuity
platform
will
host
all
of
that
and
then
it
just
deploys
the
the
resources
generated
by
that
application.
A
C
It
it's
a
hard
question
to
answer.
I
was
actually
discussing
with
some
people
in
the
cncf
Argo
CD
slack
about
how
do
you
measure
how
much
an
Argo
CD
instance
can
can
scale
on
its
own
and
currently,
in
our
stress,
testing
of
the
Acuity
platform
we've
tested
with
I?
Think
it's
something
like
5000
clusters
and
10
000
applications
and
haven't
run
into
any
limitations.
So
the
best
answer
is
yeah.
Well.
B
B
It's
about
350
clusters,
550
repos,
and
about
almost
3
000
application
per
Argo
City
instance,
and
that
the
usual
caveat
of
like
you
need
to
set
your
resource
limits
and
and
quotas
and
like
you,
need
to
like
adjust
those
but
like
so
that's
kind
of
like
what
Intuit
told
me
I'm,
pretty
sure,
there's
other
folks
that
have
either
reached
an
upper
or
lower
limit
of
that,
but
that's
adding
to
it
per
our
date.
On
average,
that's
kind
of.
C
Well,
so
what's
interesting
is
comparing
that
to
the
the
Acuity
platform.
Is
that,
with
with
the
agent
based
model,
you're
kind
of
Distributing
the
workload
of
Argo
CD
into
each
of
the
connected
clusters,
because
ultimately,
the
the
generation
of
like
the
the
components
that
you
would
need
to
scale
on,
a
single
Argo
CD
instance,
are
now
spread
out
between
all
of
the
connected
clusters
and
the
processing
requirements
for
each
cluster
are
self-contained
in
that
cluster,
with
the
with
a
caveat
that
you
can
dedicate
the
repo
server
to
a
specific
cluster.
C
So
if
you've
got
like
one
cluster
that
you
want
to
connect
to
your
private
git
repo
or
like
a
internally
hosted
one,
you
can
go
and
you
can
use
the
what's
called
the
get
delegate
feature
to
go
and
say:
I
want
this
cluster
connected
Target
CD
to
do
all
of
the
repo
processing,
and
that
includes
Helm
and
git.
Even
though
it's
called
get
delegate.
That
was
a
fun
debate,
yeah,
but
yeah.
So
honestly,
we're
gonna
do
some
scalability
testing
and
we're
gonna.
C
C
B
A
A
Is
a
completely
alter
your
numbers
and
then
you're
going
to
start
hitting
limitations
that
are
just
things
like,
like
you
said
like
wherever,
wherever
the
the
bottlenecks
end
up
in
like
okay,
we've
got
one
git
Source
or
we've
got.
You
know
only
this
much
throughput
on
our
networking
or
whatever
that
ends
up
being
yeah.
C
B
B
A
B
C
And
so
all
that
to
be
said
is
our
application
is
now
ready
to
be
synced
after,
after
all
of
our
conversation
and
we'll
go
ahead
and
we'll
do
it.
So
this
will
do
the
the
generation
and
deploy
the
resources
onto
our
openshift
cluster,
which,
if
we
go
back
here
and
we
change
to
our
Helm
guestbook
project,
you
can
see
that
we
now
have
a
deployment
and
we
now
have
pods
for
that
deployment.
C
Yeah
we're
gonna
break
it
so
that
I
can
show
the
pains
that
I
went
through
trying
this
out.
So
what
we're
going
to
do
is,
let's
see
if
I
can,
if
I
can
find
it
here
through
the
let's
see,
I
think
I
want
the
it's.
The
helm,
guest
book,
namespace
and
I
want
a
roll
binding,
and
it's
going
to
be
named
any
uid
this
one.
Okay,
I
just
pretend
you
didn't
see
that
and
we're
gonna
go
and
I'm
going
to
go
and
delete
this
replica
set
to
to
show
the
demo.
C
So
the
helm
guest
book
is
going
to
go
and
create
a
a
replica
set,
and
strangely
we're
in
a
crash
loop
back
off,
and
if
we
look
at
the
events
here,
there's
nothing
really
specific
about
the
the
deployment
of
this
pod
that
that's
preventing
it
from
working.
But
we
go
to
the
logs
here.
We
can
see
that
it's
having
trouble
binding,
Port
80.,
now
I'm
wondering
if
you
guys
can
help
me
understand
why
we're
not
able
to
find
Port
80.
B
C
B
So
a
pod
is
is
nothing
more
than
kind
of
just
like,
like
c
groups
and
all
of
that
stuff,
wrapped
around
containers
right
and
that
container
is
nothing
but
like
a
process
right,
a
Linux
process,
but
that
Linux
process
has
to
run
as
a
user
right
like
because,
like
you
know,
Linux
Unix
right,
everything's
a
file,
and
you
know
everything
has
to
run
as
a
user.
You
know
that
that
sort
of
thing
when
openshift
spins
up
a
container
a
pod,
it
assigns
a
random
uid.
So
so.
C
B
It's
nice
I'm,
like
I'm,
I'm
talk,
I'm
gonna
make
a
pejorative
statement.
It's
it's
just
it's
just
true.
If
you
don't
know
like
if
you're
like
a
developer
or
stuff
like
that,
there
is
a
privileged
Port
range
right
that
you
cannot
bind
to.
So
if
you
have
a
process
like
Port
80
is,
is
a
port
that
that
a
regular
user
can't
bind
to
it's
just
it's
it's.
It's
marked
as
privileged
and
so.
B
B
B
So
that's
when,
but
since
you're
binding
to
Port
80
you
you
probably
you
know
this.
This
is
in
your
image
right.
This
is
just
like
a
Helm
chart.
The
the
real
fix
is:
don't
run
your
applications
on
privileged
ports.
There's
there's
no!
In
the
in
the
age
of
containers.
You
can't
make
an
argument
for
running
on
Port
80..
The
only
argument
is
it's
a
legacy
app
and
at
that
point
I'm
saying
you're.
Writing
it
into
a
container
change
it
then,
if
you're
refactoring
it
might
as
well
refactoring
it
to
run
on
80
80.
A
A
A
The
the
role
binding
stuff
is
actually
things
that
I
hear
a
lot
and
I
see
a
lot
that
people
hit
so
I
do
want
you
to
go
through
the
way
you
fixed
it.
Even
though
your
SRE
will
tell
you
no
don't
use
that
Port,
but
I
do
think
that
this
is
a
very
useful
thing
for
folks
at
home.
Maybe
just
getting
started
with
openshift
to
see.
C
Yeah
yeah,
certainly
so
I'll
go
through
kind
of
my
journey
of
of
solving
this
problem,
so
you
know
I
Port,
80,
I
I
started
to
get
it
so
after
I
found
out
that
we're
we're
setting
a
very
specific
user
ID.
Now
this
is
a
question
I
had
about
openshift.
Does
it
assign
a
specific
range
of
uids
to
each
namespace
or.
B
C
C
Great
yeah,
so
you
know
After
figuring
out
the
whole,
the
whole
uid
thing,
and
it
took
me
a
minute
to
relate
that.
It's
setting
the
uid
within
the
container
and
then
that
prevents
it
from
doing
a
privileged
Port.
So
once
I
got
through
that,
you
know.
The
first
thing,
I
thought
is
that
I
I
might
just
update
the
the
container
Port
here,
but
obviously
the
container
image
isn't
pulling,
which
port
to
run
from
based
on
the
spec
of
the
the
pods.
C
C
This
is
a
role
reference
to
the
the
neui
D
role,
which
my
understanding
is
that
it
will
then
let
my
container
run
as
what
whatever
uid
it's
default
set
to
run,
and
it's
not
going
to
Auto
inject
that
run
as
user
ID
parameter
and
then
the
subject
is
the
default.
Namespace
default
service
account
in
the
home
guestbook
namespace,
which
is
what
my
pod
is
running
at
and
so
by
creating
that
on
the
cluster
I
should
be
able
to
go
back
to
Argo
CD
here
and
I'm.
C
C
Okay,
okay,
I'm
glad
it's
not
just
my
Chrome
browser.
So
if
I
restart
it
we'll
get
a
new
replica
set
which
will
regenerate
the
Pod
deployment
and
it's
happy,
it
looks
like
we
were
able
to
bind
onto
Port,
80
and
we'll
notice.
If
we
go
and
we
look
at
the
the
Manifest
here,
the
security
concept
context
no
longer
has
the
run
as
uid
specified
so
that
that's
that
was
my
journey
of
of
working
with.
B
Lot
of
fun
well,
what's
what's
interesting,
is
that
you're,
with
with
containers
right,
kubernetes
and
containers
like
you're,
like
forced
to
care
about
security,
a
little
bit
more
and
I
and
I
only
say
that,
because
the
point
of
isolation
no
longer
is
like
a
VM
because,
like
you
know
like
from
VM
to
VM
like
if
there's
a
kernel,
vulnerability
one,
it
probably
won't
be
in
the
other.
If
you
upgraded
the
other
one
right,
but.
A
B
B
Actually,
don't
right
so,
like
you
know,
like
it's
they're,
just
processes
right
they're,
just
you
know,
they're
just
Linux
process
is
running
running
on
a
on
a
system
and
it
changed.
You
know
change
rooted
system,
so
it's
like
you
actually
do
have
to
think
about
security,
a
little
bit
more,
which
is
kind
of
what
we
try
to
do
with
openshift
right.
We
try
to
give
it
like
the
Easy
Button,
meaning
like
you're,
secure.
A
B
Default
and
if
you
want
to
do
anything
else,
you
kind
of
have
to
like
modify
it
to
your.
You
know
to
your
needs
like
using
any
uid.
Isn't
this
inherently
insecure?
But
you
know
it's
better
than
using
privileged
right,
because
there's
an
SEC
security,
conscious
constraint,
called
privilege
which
is
basically
like
you
have
access
to
the
whole
node
like
if
you
give
so
never
use
privileged.
B
A
A
Oh
yeah,
I
need
a
postgres,
and
you
know
this
is
me
hacking
around
at
like
10
30
at
night
and
I
wasn't
awake
so
I
just
just
grabbed
a
copy
of
you
know
some
yaml
for
a
postgres
instance
off
of
their
what
other,
tutorials
or
whatever,
and
then,
of
course
it
would
not
start.
I
was
like.
Why
would
you
start
there's
nothing
wrong
with
this?
This
should
be
perfect
right.
A
This
yellow
lens
I
didn't
like
screw
up
my
copy
pasta
like
everything,
should
be
fine,
I
go
look
at
the
logs
and
it
says
complaining
about
root.
I'm
like
dang.
It
right
I
need
to
go,
get
the
open
shift
version.
So
then,
I
went
and
I
think
the
the
key
or
Quay,
depending
on
whose
side
of
the
debate
you're
on
on
that
they
they
had
one
so
I
just,
went
and
used
that
one
which
was
well
and
good
and
easy,
and
then
everything
just
worked
like
magic.
A
The
way
I
wanted
it
to,
but
it
was
an
important
note
that
you
can't
just
use
any
image
on
openshift.
It
actually
has
to
have
one
that's
been
through
certain
security
protocols
like
removing
the
root
user,
which
removing
root
users
from
Linux
images
is
actually
like.
A
very,
very
big
thing:
I
came
from
iot
right
before
I
came
over
to
Red
Hat
and
we
removed
the
root
user
from
our
iot
servers
as
well.
You
just
did
not
that
didn't
exist.
A
B
There's
no
root
in
that
it's
like
in
the
iot
devices.
I.
Remember
like
hacking,
some
of
my
my
Internet
devices,
hopefully
spectrum's,
not
listening
and
like
it's
like
a
it's,
a
busy
box
image
and
like
like
it's
like
hilarious,
like
there's,
there's
like
there's
no
root
right.
It
was
imaged
with
this
operating
system
with
no
roots,
so
very
important.
Yeah.
A
C
C
Is
that,
like
start
with
nothing,
and
then
you
have
to
be
conscious
of
what
you're,
giving
your
your
services
and
and
and
to
speak
to
your
earlier
point,
is
kubernetes
is
taking
all
of
these
VMS
that
used
to
be
individual
components,
and
that
was
just
one
big
system,
so
least
privileged
matters
more
and
more
as
we
get
into
these
giant
distributed
systems
of
like
sure
a
container
over
here
and-
and
you
know,
Asia
could
now
affect
the
rest
of
the
system.
That's
running
over
in
North
America,
just
because
it's
one
big
kubernetes
cluster.
A
Yep,
a
colleague
of
mine,
Kristoff
blecker
people
in
the
kubernetes
community
will
know
him
from
steering
I'm
just
going
to
call
him
out.
He
said
this
one
time
on
a
meeting.
He
said
most
amount
of
actions,
least
amount
of
access
and
I.
Don't
know
if
he
originated
that
or
he
stole
it
from
somebody
else.
But
I
love
that
and
I
say
it
all.
B
Yeah
well,
I
mean
it's
whatever
you
know,
whatever
analogy
works
for
you,
so
sort
of
thing,
so
yeah
so
as
as
I
was
actually
kind
of
looking
in
as
I
were
kind
of
rounding
up
for
the
last
10
minutes.
By
the
way,
if
anyone
has
any
questions,
you
know
drop
into
the
chat,
you
know
we'll
we'll
talk
about
it.
You
know
we'll
have
some
fun
with
this,
but
I
kind
of
want
to
give
some
time
to
the
to
Nick
about
talking
about
like.
B
What's
what
is
what
what
can
you
tell
me?
I,
don't
know
if
Hong's
watching
so
I
don't
know
if
you'll
get
in
trouble.
But
what
what
can
you
tell
me
about
like?
What's
up
and
coming
with
Acuity
and
with
you
know
some
of
the
things
right
now
it
seems
like
very
Argo
CD
based,
as
you
said,
but
like
there's,
there's
other
functionality
coming
down
the
road
map.
I.
Imagine
you're
gonna
start
implementing
something
like
Argo
rollouts,
which
is
quite
popular
or
you
know.
You
know
something
in
that
space
But.
B
C
Yeah
yeah,
it's
a
it's
a
great
question.
You
know
there.
The
Argo
ecosystem
is
quite
Broad
and
obviously
the
Acuity
platform
is
focused
on
the
Argo
CD
side
right
now,
and
there's
still
a
lot
of
effort
that
we
want
to
put
into
that.
Mainly
we
were
we're
adding
first
class
support
for
the
broader
Argo
CD
ecosystem.
C
So
so,
there's
that
and
there's
there's
some
some
upcoming
open
source
contributions
that
were
we're.
We're
really
excited
about.
There's
it's
further
down
the
road
map
and
I
can't
say
anything
too
specific,
but
we
are
still
very
committed
to
contributing
back
to
open
source
and
and
maybe
proposing
new
things
that
can
can
go
and
and
broaden
some
of
the
the
get
Ops
tooling.
So
you
know,
if
you're,
watching
no
secrets
spoiled
a
little
anticipation
built
yeah.
B
A
B
Twitter
yeah,
that's
one
of
the
things
I
really
really
liked
about
Acuity.
You
know,
and
you
know,
Hong
and
those
guys,
Jesse
and
Alex
is
like
they've,
always
been
open
source
first,
and
you
know
us
at
Red
Hat,
you
know
that's
kind,
that's
our
thing
right,
so
you
know
I
I.
We
always
appreciate.
We
always
appreciate
that.
That's
you
know
something
definitely
watered
right
up
our
alley.
Yeah.
A
It's
I
I
was
gonna,
say
it's
almost
a
callback
to
my
Arab
mystery
earlier,
where
I
was
like.
You
know,
go
look
at
this
thing
for
this
eventual
announcement
that
you're
gonna
care
about
like
I'm
like
okay.
Well
now,
where
what
thing
do
I
go,
follow
to
look
for
so
the
announcement
inevitably
I'm
going
to
care
about
right.
So
your
Twitter
is
fine.
I'll
I'll
stalk
you
there
I
can
do.
C
And
and
speaking
of
my
Twitter,
if
there's
any
Argo
CD
users
out
there
that
have
what
they
think
is
a
really
interesting
use
case
hit
me
up,
I
want
to
learn
about
it,
I'm
not
going
to
sell
you
on
anything.
You
know
I'm,
not
too
concerned
about
that.
I
want
to
learn
about
cool
things
that
Argo
CD
users
are
doing,
because
you
know
that's
my
favorite
part
of
this.
This
role
that
I'm
in
is
is
getting
to
learn
about.
C
B
Yeah,
that's
that's
where
that's
by
the
way
for
anyone
interested
like
in
the
Argo
Argo
project,
ecosystem,
cncf
slack
is
a
must.
You
have
to
you,
have
to
be
there
because
that's
like
kind
of
where
all
the
Argo
project
users
hang
out
since
it's
a
graduated
project.
Now
we
get
to
say
that
it's
part
of
the
cncf
ecosystem,
so
yeah,
you
know
for
sure,
you'll
catch
me.
There
you'll
catch
Nick
there
Hillary's
there
at
times.
B
Yeah,
so
so
yeah
cncf
stack
is
a
must,
especially
if
you
want
to
talk
to
you
know
the
engineers.
You
have
a
question
right
if
you
also
I,
also
like
hearing
how
people
are
using
Argo
City,
because
we
have
our
opinions
right.
We
have
our
point
of
view,
but
like
we
want
to
see
how
you're
you're
using
it
right,
that's
the
only
way
to
make
the
product
better.
C
A
C
And
I'm
always
shocked
at
how
active
the
the
Argo
CD
slack
channel
is
on
the
cncf
slack,
because
it's
like
every
day,
there's
people
just
sitting
in
there
happy
to
answer
questions,
no
matter.
If
it's
a
really
strange
Edge
case
you
have,
or
you
just
can't
find
the
right
page
in
the
documentation,
which
is
me
half
the
time,
yeah.
A
100
100,
it's
it's
the
worst,
you
don't
you
read
the
thing.
You
don't
remember
where
you
read
it
at
that
point,
you're
just
frustrated
and
you
need
somebody
to
tell
you
where
the
thing
is
that's
in
front
of
your
face
is,
and
that
group
is
so
willing
to
do
that
and
I
love
everybody
in
that
slack.
For
that,
thank
you.
B
Awesome
awesome:
all
right.
Is
there
any
other
questions
in
the
tall?
Do
you
have
any
other
questions
or
anyone
that
that's
happened
to
watch?
Remember
drop
it
on
we're
on
for
a
few
more
minutes
here
and
I
I.
Actually.
So
what
what
I
like
about
this
platform?
Is
that
you
kind
of
have
a
central
location
for
managing
your
Argo
City
instances
right?
B
So
a
lot
of
so
there's
there's
this
big
debate
versus
a
hub,
Hub
and
spoke
architecture
right
where
you
have
kind
of
like
a
central
Argo
City
instance,
or
have
an
Argo
City
instance
per
cluster
or
have
a
subset
of
that.
What's
really
cool
about
Acuity
is
you
can
you
can
have
it
all,
and
on
top
of
that,
you
can
have
you
have
like
a
central
console
to
kind
of
manage
all
of
it.
So
that's
that's
kind
of
pretty
cool
and
I.
B
C
Yeah
yeah,
we
just
recently
announced
our
self-hosted,
offering
which,
if
you
come
from
the
home,
labbing
Community
self-hosted
might
be
a
little.
You
know
misdirected,
it's
you
can
host
it
on
on
really
any
platform
that
you
want,
but
it's
it's
really
well
supported
on
on
like
AWS
and
gcp.
So
if
you
just
want
to
host
it
yourself
and
and
take
all
the
components
and
have
full
control
over
the
environment,
you
have
that.
C
C
B
Yep
awesome
awesome,
that's
really
cool!
That's
that's!
Something!
I
wanted
to
get
my
hands
on,
but
I
I
ping
neck,
and
he
goes
that
it's
not
public.
Yet
it's
just
Acuity
customers.
Only
so
I
think
you
know,
which
is
which
is
fine
but
like
I'm
itch
I'm,
always
the
one
itching
to
get
my
hands
on
things
to
test
them
out,
but
that
that's
really
cool.
That's
really
cool
that
it's
self-hosted
is
available.
A
Yeah
I
agree:
I,
think
that
that's
one
of
the
things
we
talk
about
a
lot
you'll,
see
it
across
a
lot
of
the
red
hat
stuff.
Is
the
hybrid
Cloud
right,
a
hybrid
Cloud
bare
metal,
self-hosting
on-prem,
none
of
that's
going
anywhere,
it's
just
getting
augmented!
Really,
so
the
words
failed.
You
know
what
I
don't
know
why
I
try.
The
jet
lag
is
just
very
real.
Today.
A
B
Yeah,
no
for
sure
for
sure
cool
all
right
looks
like
we
are.
We
are
at
time
right
we're
at
top
of
the
hour.
I
actually
have
another
meeting
to
get
to,
or
you
know,
I'd
be
happy
to
stay
on
so
Nick
yeah
tell
people
where
to
find
you
where
it
was
the
best
place
when
I
know
you.
You
said
that
before,
but
you
know
kind
of
tell
them
any
parting
words
here.
Yeah.