►
Description
We have been hearing a lot about the benefits of using the reactive approach to solving concurrency problems in distributed systems. While reactive programming refers to the implementation techniques being used on the coding level, on the systems deployment and runtime level, we can leverage on a robust yet very flexible and lightweight framework such as Vert.x to deliver. In this session, we will first learn about what the missions of a reactive system are, which, among many things, include handling multiple concurrent data stream flows and being able to control back pressure as well as managing errors in an elegant manner. We will also discuss the special polyglot nature of Vert.x and its use of the Vertical model. Live coding will accompany this session to illustrate how to program a simple use case using multiple JVM languages such as Java and Kotlin, and then we will build and dockerize it to be deployed as a serverless container to the Knative cluster in a delightful manner.
Speaker(s):
Mary Grygleski (IBM)
A
My
topic
today
is
on
deploying
a
modern
serverless
container
as
powered
by
vertex,
but
actually
the
a
lot
of
the
things
that
I'm
going
to
be
talking
about
today
is
will
be
on
the
the
reactive
side
of
things.
So
all
right,
let
me
let
me
go
and
progress,
and
I
know
I
only
had
30
minutes,
so
I'm
gonna
be
a
little
bit,
maybe
brief
too.
So,
first
of
all
too.
A
I
wouldn't
say
it's
all
simple,
but
it's
actually
sounds
simple,
but
it's
not
very
flexible
and
another
strength
of
it.
Is
that
it's
a
polyglot
nature
of
it
and
it's
very
interoperable.
The
idea
is
that
it's,
the
opportunities
are
unlimited
and
basically
to
today
I'm
also
kind
of
bringing
in
the
you
know.
How
do
you
kind
of
do
deployment
to
the
cloud
part?
So
it's
all
serverless
ready
like
that,
so
very
modernized,
and
let
me
then
progress
and
again
you
know
when
I
can't
talk
about
quarkx
and
vertex.
A
The
the
thing
is
that
in
corkus
you
cannot
see
vertex
because
is
the
the
layer
that's
kind
of
underneath
it's
like
when
you
go
on
the
street,
you
know
plumbing,
you
cannot
see
the
pipes
and
all
that
stuff,
that's
working,
but
that's
what
is
happening.
Vertex
is
working
underneath
the
hood
so
yeah.
If
you
look
into
here
so
basically
the
is
the
application
that
goes
into
the
routing
layer
and
basically
to
the
routing
layer
then
goes
through
vertex.
It's
event.
A
So
before
I
actually
get
a
bit
deeper
into
like
vertex,
because
this
is
a
more
beginner
talk
like
to
kind
of
talk
a
bit
too
about
reactive
java
and
just
just
kind
of
get
everybody
on
the
same
page.
So
bear
with
me
if
you're
already
familiar
with
it.
So
when
we
talk
about
reactive
systems,
you
know
we
cannot
avoid
saying
reactive
manifesto.
A
Now
again,
it's
reactive,
it's
there's
actually
the
fine
line
between
the
reactive
systems
and
then
the
reactive
programming
programming
part
as
well
as
kind
of
the
the
reactive
streams,
the
specification
so
reactive
manifesto.
It's
essentially
there
are
four
core
tenants
which
is
responsive,
responsive
and
meaning
that
you
know
requests
going
into
a
system.
They
need
to
be
handled
in
a
very
timely
manner,
that's
being
very
responsive
in
a
nutshell,
so
that's
being
enabled
to
essentially
too
in
today's
world.
You
know
in
the
cloud
infrastructure.
A
It's
basically
a
lot
of
the
you
know.
The
performance
is
improvement,
they're
clustering.
So
all
these
other
things.
So
if
you
look
you
look
into
the
let's
look
into
the
left
side,
the
elasticity
elasticity,
meaning
that
yeah.
If
there
are
a
lot
of
requests
coming
in,
then
the
system
needs
to
be
able
to
handle
it
and
be
very
you
know,
kind
of
like
flexible
elastic.
You
know
more
processors,
more
processes
or
whatever
is
needed
to
process
your
requests.
A
Let's
say
I
like
to
use:
you
know
like
valentine's
day
for
a
flower
shop
online
flower
ordering
so
on
valentine's
day
before
that
you
get
a
lot
of
orders.
All
of
a
sudden.
Your
system
is
flooded
so
the
system
in
order
to
be
qualified
for
reactive.
You
need
to
have
you
know
many
processes
being
able
to
handle
all
the
load
and
then,
when
it's
like
you
know
slow
time
in
the
summer
you
get
flowers
outside
then
there's
no
need
to
be
working
with
that.
A
Then
you
don't
need
as
many
processor,
so
it's
kind
of
like
returning
all
the
resources
back
to
the
pool,
so
it's,
in
other
words
very
scalable,
very
elastic,
so
that's
another
characteristics.
And
then,
if
you
look
into
the
right
side,
there's
resiliency,
meaning
that
even
you
know
in
any
error,
condition
too.
The
system
should
not
slow
down,
so
the
responsiveness
should
be
maintained.
So
then
you
need.
The
system
needs
to
have
some
mechanism
of
handling
kind
of
the
the
problem
with
it
and
then
being
able
to
bounce
back
pretty
quickly.
A
So
that's
also
error
handling
needs
to
be
very
responsive
as
well,
and
then
all
these
three,
I
also
always
look
at
then.
The
fourth
thing
in
here
I
call
it.
Fourth
thing
is
really
the
message
driven
mechanism
that
is
enabling
all
of
it
and
by
being
reactive,
your
system
has
to
be
very
loosely
coupled.
Nothing
is
kind
of
like
tied
together,
but
the
amazing
thing
is
that,
even
though
you
we
call
it
like
loosely
coupled
and
message
driven,
they
all
do
their
own
thing,
but
the
whole
system
as
a
whole.
A
They
need
to
be
all
kind
of
working
in
a
very
coherent
fashion
and
message,
driven
as
such,
too
is
essentially
messages
passing
between
different
endpoints
like
that
and
which
is
different
than
event
driven
too
because
event
driven
is
a
bit
more
like
your
messages
are
being
emitted
by
a
source
and
the
source.
It
doesn't
know
where
the
messages
are
going
to,
and
it's
really
up
to
the
client
who's
interested
in
the
in
the
the
messages
will
kind
of
subscribe
to
it,
for
example,
in
a
reactive
extension
kind
of
manner.
A
So,
okay,
so
here
just
kind
of
a
quick
thing
to
give
you
some
also
a
bit
of
broader
understanding
how
vertex
fit
into
this
larger
picture.
There
are.
These
are
like
some
of
the
reactive,
popular
reactive,
toolkits
and
libraries
open
source
that
are
kind
of
popular
in
today's.
You
know
kind
of
you
know
ecosystem.
A
The
rx
java,
well
of
which,
like
vertex,
has
also
like
very
good
integration
with,
and
I
have
some
example
that
I'm
going
to
be
showing
you
too
in
this
talk
later,
and
so
that's
you
know
basically
yeah
so
okay,
then
there's
also
then
akka
and
akka.
It's
basically
from
light
band
and
that's
also
a
very
highly
reactive.
A
In
fact,
light
band
is
the
one
that
leads
the
they
establish
they
establish
or
help
to
define
the
reactive
manifesto
and
their
kind
of
very
sophisticated
framework
library
making
use
of
the
actor
model,
and
it's
a
pretty
pretty
unique
kind
of
modeling.
That's
actually
based
on
erlang.
It's
not
really
all
new,
but
this
isn't
the
talk
about
it.
So
I
won't
go
into
details
and
there's
also
logon,
which
is
a
microservices
framework.
A
A
So-
and
this
is
also
a
quick
thing
too,
about
like
reactive
streams
too-
there's
also
our
socket,
which
is
actually
now
managed
by
the
reactive
foundation
underneath
the
linux
foundation.
These
are
the
six
companies
that
started
it
and,
of
course,
vmware.
Actually
this
is
a
bit
of
an
older
slide
from
last
year,
and
so
vmware
is
the
one
that
provided
this
slide.
I
thought
it's
kind
of
nice
because
I
wanted
to
point
out,
because
our
socket
is
a
bit
of
a
lower
level
layer,
five
and
six
communication
protocol.
A
That's
also
based
on
reactive
streams,
and
it's
really
fast
too,
so
just
to
give
you
some
kind
of
kind
of
a
comparison,
okay.
So
now,
then,
let
me
kind
of
get.
Let
me
I'm
keeping
track
of
the
time
so
getting
a
bit
then
into
vertex
right.
So
when
we
do
vertex,
we
work
with
the
core
library
and
some
good
examples
of
use.
Cases
is
like
a
list
of
all
of
these
things,
but
I
want
to
highlight
it.
A
Has
this
event
bus
that
is
very
powerful,
and
the
thing
is
too
is
that
it's
also
quite
simple:
to
use,
for
example
like
in
terms
of
showing
you
like
setting
a
periodic
delayed
action,
kind
of
call
is
basically
set
periodic
and
that's
how
you
do
it
and
you
can
also
and
okay.
You
can
also
like
doing
the
deploying
and
undeploying
the
verticals
vertical.
Is
the
model
that's
being
supported
by
vertex
and
I'll
go
into
that
shortly,
but
it's
not
enforced.
A
You
know
you
don't
need
to
use
it,
but
it's
highly
recommended
because
it
provides
like
a
unit
of
deployment,
especially
say
in
today's
like
cloud
infrastructure.
It
actually
makes
sense
to
extend
your
abstract
for
vertical
or
you
can
write
your
own
too,
so,
okay,
so
these
are
again
some
of
these
examples
to
use
vertex
and
some
usage
and
dependency
on
this
is
that
you
you
can
like
build
just
like
anything.
A
This
is
dependency
for
the
core
library
currently
they're
up
to
392.,
and
then
gradle
too
is
you,
can
just
do
compile
and
then
give
this
directive
and
in
vertex
too,
actually
the
kind
of
very
basic
of
it,
and
that's
why
I
like
it
too,
is
that
it's
it's
simple:
it's
all
you
need
to
do
in
your
runtime
is
essentially
just
start
up
the
the
object.
Vertex
object,
control
center,
so
by
doing
a
vertex,
dot,
vertex
and
we'll
have
some
examples
too.
So
this
is
like
runtime
options.
A
How
you
do
it
is
to
add
an
option
you
can
pass
in
the
the
instantiate,
your
vertex
options
and
basically,
then
do
a
set
in
this
case,
setting
a
worker
pool
size
of
40..
So
you
can
also
then
have
other
options
as
well.
You
can
find
out
from
the
documentation
so
again.
Vertex
is
very
event
driven.
Some
of
the
example.
Events
will
be
like
a
timer
has
fired,
or
data
has
arrived
on,
the
socket
or
data
have
been
read
from
disk
and
exception
has
occurred
or
http
has
received
a
request.
A
A
Okay
and
again
I
mentioned
about
the
vertical.
So
vertical
is
similar
to
the
actor
model
in
akka
and
it's
not
a
requirement.
It's
a
unit
of
deployment
in
vertex
and
a
vertical
process
is
incoming
events
over
their
event.
Loop,
an
example
of
some
of
these
events,
which
I
mentioned
about
timing
events,
network
buffers
messages
sent
by
other
verticals
and
event
loop,
a
bit
more
about
it
event.
Loops
are
typically
used
in
asynchronous
programming
models
and
each
event
gets
processed
in
a
very
reasonable
amount
of
time.
A
So
then
you're
not
blocking
the
event
loop
and
every
event
loop,
2
is
attached
to
a
thread
and
by
default
the
vertex
attaches
like
two
event:
loops
per
cpu,
core
thread
and
and
okay
and
some
words
too
about
event,
loop
and
vertical.
A
regular
vertical,
always
processes,
events
on
the
same
thread,
and
so
that's
nice
thing
too.
You
use
a
reactive
library
and
I
actually
got
asked
this
question
like
people
are
saying.
Well,
why
reactive?
Why
is
it
so?
A
The
thing
is
that
before
right
before
all
of
this
reactive
kind
of
a
new
way
for
reactive
revolution,
basically,
everything
was
simpler.
Life
was
simpler,
but
they're
not
as
powerful
is
that
when
we
deal
with
like
trying
to
do
multitasking,
then
we
have
to
deal
with
threats
ourselves.
So
all
of
these
threats
it
happens
in
the
kernel.
A
So
can
you
imagine,
then
that's
why
things
are
not
as
fast
because
then
every
time
you
have,
I
o
you're
holding
up
that
threat,
but
now
with
reactive
is
that
you
basically
still
has
threats,
but
they
would
get
moved
into
user
space,
and
so,
but
in
this
case,
do
vertexes
make
use
of
the
event
the
the
event
bus,
and
so
it
handles
all
of
the
threat
coordination.
So
then,
as
an
application
developer,
we
don't
need
to
deal
with
this
like
lower
level
kind
of
concern
and
a
vertical
can
be
passed.
A
And
also
too,
though,
in
highlight
of
vertex,
it
says
it's
very
polyglot
in
nature.
It
supports
like
multiple
languages.
All
these
are
listed
here
and
I
think
the
python
enclosure,
apparently
is,
will
be
in
in
earlier
version,
although
maybe
python
is
already
like
ready
in
392.,
but
it
doesn't
force
the
usage
of
any
languages,
that's
the
beauty
of
it,
and
so
basically,
that's
my
thinking
is
that
you
know,
in
terms
of
you
know
some
practical
usage.
You
know,
how
can
we
address?
A
So
anyway,
that's
just
some
thinking
about
you
know
kind
of
how
what
can
we
make
use
of
you
know
a
library
such
as
this,
you
know
make
it
even
more
like
powerful
in
that
sense,
so,
okay,
so
here's
an
example
too,
which
I'll
show
quickly
is
that
it's
just
illustrate
right.
This
polycloud
nature.
I
have
a
simple
case
in
which
I'm
basically
sending
http
requests
and
I'm
writing
a
vertical.
A
I
call
it
like
an
api
kind
of
gateway
that
funnels
the
request,
and
then
I
have
different
worker
verticals
that
handles
they're
written
in
different
flavor,
java,
kotlin
and
javascript,
and
basically
it's
the
event.
Bus
would
take
care
of
kind
of
you
know,
doing
all
the
work
and
and
and
the
doing
the
work
and
then
telling
each
of
the
vertical
to
do
their
work.
Accordingly,
you
know
based
on
the
requests,
so
the
javascript
side
is
basically
have
a
sock.js
bridge
over
it,
so
I'll
be
showing
that
shortly.
A
I
think
I
only
have
about
10
minutes
or
so
so
now,
so
it
gets
into
some
example,
but
okay,
so
these
are
just
some
of
the
kind
of
state
transition
diagram,
but
I
will
quickly
show
you
okay,
so
now
I'll
get
into
the
that
part
later.
But
let
me
I'm
feeling
a
bit
rushed,
so
I'm
sorry
kind
of
looking
at
the
time
so,
okay,
so
I
was
kind
of
talking
about
this.
Well,
let's
first
look
into
a
simple
case,
so
this
is
a
a
you
know.
A
Like
any
kind
of
demonstration
right,
we
want
to
show
something
kind
of
more
basic.
So
in
this
case
is
a
hello
world,
hello
vertical,
so
I
extend
abstract
vertical,
and
so
that's
why
I
kind
of
like
about
it,
because
it's
it's
a
bit
kind
of
simpler.
All
you
need
to
do
is
extending
it
and
basically
implement
your
functionality
from
the
start,
and
then
you
can
kind
of
handle
the
rest
of
the
thing
so
over
here,
vertex
set
periodic
in
this
case
shown
earlier.
A
So
that's
how
you
you
set
a
timer
and
then
also
like
set
the
the
you
know,
create
a
http
server
listener.
In
this
case,
I
have
I'm
using
this
lambda
expression
and
and
basically
two
it's
I'm
setting
the
request
handler
and
then
also
doing
it
binds
it
to
port
8080
in
this
case.
So
in
order
to
start
this
again,
the
earlier
I
showed
the
the
the
control
center.
A
You
start
off
the
control
center
and
then
all
you
need
to
do
is
doing
a
deploy
vertical
and
then
basically,
then
instantiate
so
over
here
and
that's
how
you
deploy
the
vertical.
So
all
these
are
not
cloud
yet.
You
know
and
I
which
I'll
bring
up
in
it
shortly,
so
I'm
not
going
to
have
time
to
run
it,
but
let
me
then
quickly
show
the
api
vertical
that
I
was
showing
earlier
so
over.
Here
too,
I
extend
abstract
vertical
and
basically
two
in
my
start
right
I'm
I
talked
about
a
configuration
so
over.
A
Here
too,
I
basically
am
making
use
of
the
config
retriever
to
handle
this.
So
over
here
I
read
from
a
property
file
of
my
config
json
so
over
here,
just
a
quick,
simple
thing
and
just
have
port
and
then
delay-
and
you
know,
work
accounts.
Some
simple
configuration
data
go
back
here,
and
so
so
I'm
making
use
of
the
config
store
option
and
then
do
the
retrieve
retriever
and
then
basically
then
add
this
file
store
and
then
basically
too,
the
nice
thing
is,
you
can
actually
have
it.
A
The
retriever
listen
for
any
changes,
and
so
then
you
don't
need
to
to
do
these.
You
know
get
kind
of
like
handling
of
updating
your
your
runtime
configuration
parameters.
So,
okay,
so
that's
that
that's
the
configuration
part
and
then
here
here
it
is
too
I'll.
Show
you
it's
in
over
here.
Well,
this
okay,
so
from
so
from
the
start
part.
If
everything
is
good,
then
I
process
my
configuration
and
I
also
then
set
up
my
router
over
here,
for
example,
router.
A
A
You
know,
processing
to
handle
it
so
and
again,
this
part
is
basically
the
soft.js
bridge
that
I
talk
about
and
that
this
part
is
for
handling
the
the
javascript,
but
as
such
too
right
now,
I'm
still
in
the
process
of
kind
of
working
on
this,
and
basically
it's
my
twitch
stream
that
I'm
showing
this.
So
if
you're
interested
to
follow
me
on
my
twitch,
which
I
have
some
information
about
so
okay,
so
let
me
then
quickly,
look
at
it.
So
all
of
these
methods
are
different.
You
know
kind
of
ways
of
handling.
A
All
the
requests
in
here,
but
I
wanted
to
quickly
show
is
that
is
the
how
I
do
the
deployment
so
over
here
right
I
talk
about,
I
can
deploy
my
java.
You
know
some
worker
working
vertical
to
do
stuff,
that's
my
java.
So
all
we
need.
We
will
need
to
do.
Okay,
all
these
setup
of
configuration
then
again.
This
is
the
line
that
I
basically
can
then
deploy
my
vertical,
and
this
is
actually
the
java
worker
vertical
class
and
that's
how
I
do
it
and
that's
all
I
would
need
to
do.
A
But
of
course,
then
I
have
to
handle
some
of
these
other
things.
Like
you
know
the
the
deploy
vertical
has
this
variations
of.
You
also
handle
any
kind
of
error,
or
if
it's
good,
you
go
ahead
and
actually
deploy
vertical
in
this
case.
In
this
example,
I'm
actually
measuring
doing
some
metrics
too,
and
over
here
too,
like
if
it's
kotlin,
then
all
I
need
to
do
is
basically
deploy
vertical,
give
it
the
kotlin,
vertical
class,
dot,
get
name
and
same
thing.
A
As
for
javascript
all
we
need
to
do
deploy
vertical
and
give
the
javascript
file
name
the
the
interface
name,
so
I
mean
the
interface
to
the
javascript.
So
so
so
that's
how
essentially
it
handles
the
this,
but
we'll
take
a
quick
look.
Then,
in
my
java
worker
vertical
so
worker
vertical
essentially
also
do
the
same
thing.
It
extends
abstract,
vertical
and
then
over
here,
then
it
does
what
it
is
going
to
need
to
do
in
this
case.
A
I'm
also
reading
the
configuration
and
all
of
these
things,
and
basically,
in
this
particular
case
too,
as
an
example
I'm
not
doing
so
much,
but
in
reality
you
want
to
implement
whatever
business
functionality
you
have
so
okay,
so
this
is
kind
of
the
example
and
now
let
me
quickly
then
get
more
into
the
back.
To
my
let's
see,
I'm
so
sorry,
let
me
kind
of
go
back
here
and
talking
about
the
building
part
I
mean
and
which
is
okay.
A
A
So
of
course
there
are
all
these
steps
too,
and
I've
also
done
it
on
google
cloud
as
well,
in
addition
to
ibm
cloud.
So
it's
pretty
much.
You
know
when
you
get
to
kubernetes
it's
nice
because
it
runs
on
all
clouds.
So
it's
essentially
a
good
control.
That's
how
you
do
it,
but
first
of
all
serverless
deployment
without
you
know
going
into
all
the
details,
we're
talking
about
scalable
applications
in
a
modern,
dynamic
environments.
A
These
are
the
cloud:
environments
can
be
public,
private
and
hybrid
and
then
serverless
deployment.
The
application
techniques
would
be
containers.
Microservices,
serverless
functions,
immutable
infrastructure.
All
these
things
for
serverless,
that's
the
whole
point
and
vertex
is
kind
of
set
itself
up
really
nicely.
So,
okay,
so
in
order
to
kind
of
build
and
deploy,
all
we
need
to
do
will
be
it's
actually
in
this
case,
I'm
still
using
docker.
A
So
I-
and
there
are
other
containers
you
can
use
to,
but
essentially
do
a
darker,
build
and,
and
then
over
here
too,
and
I
would
obviously
need
my
docker
hub
too
and
basically
I
build
and
then
have
my
particular
in
this
case
a
hello
world,
and
I
basically
build
my
image
and
basically
push
it
to
my
docker
hub.
Then
all
we
need
will
need
to
do
is
actually
with
the
k
native
and
you
would
be
able
to
then
do
a
coupe
control
and
apply
a
file
name
which
is
like
my
service
yemo.
A
So,
okay,
I'm
kind
of
sorry
and
let
me
kind
of
take
a
look.
I
think
I
need
to
bring
this
up.
Okay,
okay,
I
just
see
some.
I
just
want
to
make
sure
that
I'm
not
like
you
know
running
all
of
kind
of
no
time
so
okay,
so
this
is
just
how
you
can
kind
of
quickly
do
it.
So
it's
kind
of
nice
too,
because
I
did
it
on
google
cloud
and
and
then
it's
essentially
the
same
step.
You
know
for
a
very
simple
type
of
scenario.
A
So,
okay,
let
me
then
kind
of
oops,
okay,
so
okay,
so
I
know
that
I
only
probably
have
like
a
minute
or
so
left.
So
let
me
quickly
quickly
kind
of
go
for
this
workshop
next
monday
at
8
a.m
and
pacific
time.
If
you
want
to
interested
scan
this
ibm
cloud
chicago,
there
are
multiple
ways:
there
are
san
francisco
sydney
team
that
will
be
hosting
me.
So
that's
a
monday
we'll
get
more
into
details.
A
You'll
get
a
48
hours,
use
of
a
kubernetes,
open
shift
on
ibm
cloud
and
then
okay,
sorry
and
then
there's
also.
I
have
also
a
live
stream
on
twitch
too
every
wednesday
it's
around
like
wednesday,
12
30
central
time,
and
it's
this
is
the
link
to
join
the
ibm
developer
on
twitch
and
I'll.
Be
there
every
week,
rain
or
shine,
I
will
be
there.
A
I
also
want
to
invite
you
to
join
my
discord
server
if
you're
interested
in
continuing
the
conversation
I
mean
as
such,
I
feel
kind
of
like
a
whole
bunch
of
stuff,
I'm
talking
kind
of
really
really
fast,
without
even
thinking
in
such
a
short
time.
So
there's
actually
plenty
of
conversation.
We
can
talk
about
it
the
whole
year.
Everything
and
you
want
to
join.
A
Please
join
me
on
my
discord
if
you're
interested
and
here
and
then
I
want
to
thank
you-
and
it
looks
like
my
time-
is
fire
up,
and
so
this
is
the
link
to
get
more
information
about
developer.
The
ibm
developer
for
reactive
systems
and
in
ibm
doesn't
actively
maintain
the
vertex,
but
it's
our
redhead
brothers
and
sisters
over
at
red
hat-
that's
doing
it
so,
but
then
we
work
closely
with
them
and,
of
course,
corkis
who
are
participating
in
that,
and
this
is
our
ibm
cloud,
free
tier.
A
If
you
would
like
to
get,
you
know
a
free
free
copy
of
it,
so
yeah
so,
okay
cool,
and
I
think
I
believe
I
should
be
getting
some
questions
and
I
will
try
to
answer
I'm
sorry.
Let
me
let
me
kind
of
go
back
now.
Okay,
so
I
got
one
question:
can
vertex
work
without
netting?
Yes,
it
it
doesn't
have
to
be
with
nutty
too.
How
does
it
compare
to
something
like
kafka,
kafka,
okay,
so
kafka?
A
A
But
again
we
will
talk
more
and
in
fact
I
all
I
was
showing
was
only
by
the
way
it's
only
k
native
serving
so
that's
the
part
that
handles
deploying
a
lot
of
things
I
didn't
even
touch
would
be
like
the
version
and
things
like
that.
There's
also
a
k-native
eventing
part,
which
I
would
also
like
to
be
talking
about,
but
well
you
brought
up
operators,
so
I
thought
I'd
bring
this
up
as
well.
So
okay
does
vertex
support
operator,
oh
okay,
so
I
already,
I
think
I
already
answer
it.
A
A
Let
me
kind
of
make
sure
okay,
I
think
it
looks
like
I've
answered
all
of
the
questions
right
now,
although
I
have
to
admit
I'm
not,
you
know
kind
of
very
detailed,
but
let's
follow
me
on
my
discord
or
over
here
too
then
yeah.
Thank
you.
Thank
you.
Emily.
Thank
you.
So
much
and
thank
you,
everybody.
I
know
it's
kind
of
a
bit
of
a
rush
and
but
I
hope
you
can
join
me
at
my
workshop
next
monday.