►
Description
Keynote: The Future of GitOps - Evan Shortiss, Developer Advocate, Red Hat
The GitOps practice continues evolving and becoming more accepted and integrated into teams daily. Now that we have a more clear path to GitOps, how can we continue to evolve the practice to make it as easy as possible to integrate across the application development life cycle? In this session, we’ll explore possibilities to integrate across the application development lifecycle to further coordinate releases.
A
I'm
just
here
to
have
a
little
fun
speculate
on
the
future
and
kind
of
reflect
on
the
past
of
kubernetes
get
Ops
and
where
we
are
today,
I
should
take
off
my
mask
before
I.
Do
that,
though
there
we
go
so
yeah,
the
future
is
declarative.
I
would
say
the
future
is
now
based
on
what
I've
seen
today,
but
let's
dive
in
and
reflect
so
I'm
sure.
A
Most
of
you
are
familiar
with
these
logos,
kubernetes
Docker
for
sure
we
we
all
know
and
love
them,
we're
all
familiar
with
kubernetes
because
it
made
it
possible
for
us
to
Orchestra
containers
and
Docker
made
it
possible
to
easily
build
them
for
app
developers.
So
these
were
kind
of
a
Genesis
set
of
Technologies.
I
would
say
the
goddess
where
we
are
today.
A
A
You
build
your
container
using
Docker
or
maybe
podman,
now
I'm,
red
I'm,
a
red
hat,
so
I
have
to
mention
podman
you
design,
you
design
your
sorry,
you
define
your
desired
state
in
a
declarative
fashion,
which
we
all
like,
because
we're
Argonauts
here
today
and
then
you
apply
it
to
your
kubernetes
API
server
with
cubectl
and
then
kubernetes
will
reconcile
that
state
and
make
sure
you
have
the
correct
number
of
PODS
and
the
connect
version
of
the
Pod
or
container
image
running.
So
it's
kind
of
magic
right.
A
It's
kind
of
amazing.
You
can
just
declare
something
in
a
text
file
or
a
yaml
file,
and
it's
suddenly
available
on
three
nodes
in
a
cloud
and
highly
available.
So
I
think
it's
worth
reflecting
on
that
and
how
far
we've
come
to
get
there.
So
I
think
we
can
all
say
we
really
really
like
this.
It
gives
us
superpowers
and
it's
kind
of
abstracted
things
in
such
a
fantastic
way
that
we
can
deploy
more
efficiently.
A
But
it's
not
all
fun
and
games
right.
You
think.
Oh
I
have
kubernetes
I've
I've
Argo
I've
get
a
Docker
I.
Have
all
these
amazing
Technologies
I'll
have
five
nines
of
uptime
no
down
time
ever,
but
the
reality
is
often
not
quite
the
case
right,
so
you
think
you're
going
to
be.
You
know
the
going
left,
but
actually
you're
the
guy
in
the
right,
covering
and
screaming
about
STD
and
I.
A
Don't
know
some
of
your
kubernetes
nodes
being
done
and
that's
why
companies
like
these
came
along
and
started,
offering
managed
kubernetes
right
kubernetes
managing
it
is
not
your
it's
not
your
main
job.
Your
main
job
is
to
deliver
value
to
your
customers
and
using
kubernetes
is
a
means
to
an
end,
so
using
managed.
Kubernetes
is
probably
something
lots
of
people
in
this
room
do,
but
it
still
doesn't
solve
all
your
problems.
A
There's
tons
of
work
to
do
you
have
to
configure
manage
your
deployments,
your
audits,
rollback
logging
observability,
all
these
other
things
that
I
couldn't
fit
on
this
slide,
so
we're
all
super
busy
and
we
need
more
help
and
then
Argo
came
along
and
helped
with
some
of
that
right.
It
helped
with
a
lot
of
it.
A
Actually
so
we're
using
Argo
customize
Helm,
all
these
other
Technologies
to
make
sure
our
kubernetes
kubernetes
clusters
are
actually
in
the
state
we
told
them
to
be
in
because
we
we
get
sloppy
right,
we
run
commands
and
we
accidentally
run
them
on
prod
and
then
everything's
broken
so
kubernetes
with
Argo
has
helped
avoid
that
drift
and
we
even
went
further
right.
So
we
all
know
the
default.
Crds
and
kubernetes
are
amazing,
but
we
can
do
better.
So
we
decided
to
extend
kubernetes
with
our
own
crds
right.
A
So
you
can,
you
know,
do
that
with
Argo
right.
You
have
the
application
crd,
you
can
do
a
Kafka,
the
streams,
the
operator
so
obviously
again
and
what
Red
Hats
we
have
to
mention
the
operator
framework
which
came
from
core
OS,
and
it
allows
you
to
do
fantastic
things
like
create
your
own
custom
controllers,
run
them
on
kubernetes
and
keep
your
application
in
a
desired
state
or
do
installs,
which
is
a
day
one
type
operation
and
then
upgrades
as
part
of
your
day
two
activities.
A
So
one
of
my
favorite
examples
actually
before
I
move
on,
so
we
enhanced
Cube,
so
yeah
Cube
by
default
is
probably
already
like
the
one
on
the
right,
but
we
made
it
even
bigger
and
better
so
yeah.
You
can
take
something
like
a
Kafka
cluster
to
find
it
again
in
yaml
and
suddenly
you
have
a
highly
available
Kafka
cluster
running
across
multiple
nodes
in
your
kubernetes
cluster
and
you
can
change
the
settings.
A
You
know
the
number
of
gigabytes
assigned
to
those
Brokers
easily
using
declarative
interface
with
yaml,
and
you
can
even
go
as
far
as
managing
off-cluster
resources,
Now
using
things
like
crossblane.
So
maybe
people
are
using
cross
plane
with
Argo
to
manage
I,
don't
know,
S3
buckets
or
sqsqs
or
SNS
queues
are
the
equivalent
in
whatever
your
preferred
Cloud
vendors
environment
is
so
we've
gone
from
managing
kubernetes
to
managing
things
off
of
kubernetes
and
I.
Think
it's
it's
pretty
amazing.
A
So
the
question
is
what
now
and
I
think
you
know
we've
seen
a
lot
of
today.
So
today
we've
seen
a
lot
of
what
now
or
now
what
I
think
it's
amazing
seeing
Numa
project,
observability
and
AI
tools.
I
think
the
future
is
pretty
bright.
Looking
for
Argo
in
the
community
and
I
would
like
to
see
get
Ops
becoming
even
more
adopted
when
I
interact
with
Cloud
vendors.
You
know
I
use
the
CLI
I,
don't
use
declarative
fashion
to
interact
with
them.
A
Often
I
do
it
with
my
kubernetes
cluster
on
the
vendor,
but
I
use
the
CLI
for
the
getting
started
piece.
So
maybe
in
the
future,
we'll
we'll
use
gitups
to
actually
interface
directly
with
services
that
aren't
just
kubernetes
I,
don't
know
speculation
and
that's
it
I'm.
Out
of
time
so
I'll
say
thank
you
very
much
for
listening
and
I
hope.
We've
all
had
fun
at
our
Khan
I
know.
I
did
check
out
developers.redhat.com
and
I'll
talk
to
you
all
soon.