►
From YouTube: Lessons from Becoming Cloud Native [Keynote Part 8]
Description
At Kong Summit 2019, Kong Senior Engineer, Harry Bagdi discusses how going cloud native gives you the necessary agility that is critical to the success of your business, but the path to transition isn’t always clear-cut. Harry shares Kong’s story of becoming cloud native to help you avoid common pitfalls.
Learn more about Kong: https://konghq.com/
A
A
A
A
Now
a
tough
crowd,
I
thought
you'd.
Take
that
seriously
with
one
slide.
No
I
seriously
thought
that,
but
since
you're
not
I'm
going
to
ramble
on
for
10
more
minutes
now,
you
asked
for
it
all
right.
So
that's
true.
The
engineers
are
always
told
how
this
hot
new
thing
will
solve
all
your
problems,
but
nobody
tells
us
how
do
we
go
from
our
legacy
infrastructure
to
this
new
thing?.
A
A
A
A
A
We
want
to
change
one
thing
at
a
time
we
are
all
familiar
with
Kong
as
a
RESTful,
API
gateway
use,
Kong's
admin,
API
and
configure
come.
The
restful
part
is
important.
Here
you
can
change
Kong's
configuration
any
time
every
time,
multiple
times
during
the
day,
you
can
use
cuban
ettus
to
run
this
deployment.
A
lot
of
our
customers
already
do
that,
as
you
are
integrating
cuban,
it
is
more
into
cloud
native.
A
What
you
get
is
Kong's
ingress
controller.
You
have
automated
your
deployment
and
your
configuration
and
handed
over
that
duty
to
kubernetes
anytime.
You
scale
your
services
deploy
new
services,
everything
gets
picked
up
by
cuban
artists
and
the
controller,
and
it
reconfigures
calm
no
need
to
manually
change
anything
in
Kong.
You
configure
things
in
Kong
and
use
Kong
for
analytics.
Now,
as
your
organization's
scale,
your
business
grows.
So
will
your
services
and
hence
you,
will
move
forward
and
adopt
something
like
a
service
mesh
architecture.
A
You
can
use
Kumari
steer
here
and
you
can
keep
your
knots
out
traffic
running
through
Kong
ingress,
so
you
can
see
here
we
have
service
a
and
B
part
of
our
service.
Mesh
key
here
is
gradual
change.
We
want
to
change
one
thing
at
a
time.
We
run
complex
architecture,
so
we
introduced
one
service
at
a
time
into
the
mesh.
As
you
can
see
here,
you
can
control
the
traffic,
that's
coming
into
your
cluster
and
direct
it
towards
your
mesh
or
towards
your
traditional
architectures.
A
A
A
As
you
install
your
applications,
you
install
custom
resource
definitions
into
kubernetes.
Each
team
can
control
the
configuration
of
Kong
and
you
can
be
sure
that,
if
something
works
today,
it
will
work
until
you
roll
out
a
new
application.
You
can
be
confident
that
you
can
change
and
push
this
deployment
across
clouds
across
platforms.
A
A
A
This
I
think
is
the
hardest
problem
to
solve,
integrating
with
all
the
technologies
that
are
out.
There
is
hard.
The
devil
is
in
the
details,
but
I
think
the
industry
is
converging,
the
convergence
is
slow,
but
it
is
happening
and,
as
we
see
that
convergence
happening,
we
build
that
integration
inside
form.