►
From YouTube: Architecting and delivering Java Applications in the Cloud Era | Jakarta Tech Talks | Mar 8, 2021
Description
Presented by Otavio Santana and Karina Varela
Nowadays, the best company is the one that delivers value quicker to its customers. They seek to spend less, and in consequence, they end up searching for cloud solutions. On the technical side, we have Java, which is the most mature and popular language of the past 25 years. Nevertheless, the most asked questions of the past years are: how do Java and cloud fit together? Does Java still fit the new technology trends or Is it time to deprecate Java to use newer languages such as NodeJS and GO?Join this session to understand how the Java language, standards, culture, and ecosystem around it evolved to keep the pace of cloud deployment. Learn how you can maintain your developers' productivity by architecting your Java application aligned with the most recent best practices and recommendations, besides, defining a good integration and delivery scenario for your application lifecycle.
Get Social!
Website: jakarta.ee
Subscribe: https://www.youtube.com/user/EclipseF...
Follow us on Twitter: https://twitter.com/JakartaEE
Like us on Facebook: https://www.facebook.com/JakartaEE/
Join us on LinkedIn: https://www.linkedin.com/groups/13597
Tag us: #JakartaEE
Want more great Eclipse Community webinars? Become a member of our Jakarta Tech Talks Meetup: https://www.meetup.com/jakartatechtalks
A
B
A
B
A
B
A
B
Guys,
hopefully,
karina
is
able
to
finish
the
presentation,
one
important
thing,
so
that's
the
interaction
first
presentation.
So
if
you
have
any
question
we
are
looking
to
the
chat
right
now.
So
any
question
any
comments.
Let's
let
us
know
because
once
you
put
that
comment,
we
will
read
this
message
instantly.
So
please
let.
A
A
A
B
To
be
here
with
you,
so
let's
talk
right
now
with
the
one
amazing
person
on
my
side.
So
by
the
way
happy
women's
day,
and
today
we
talk
with
karina
varela,
he
is
a
principal
technical
marketing
manager
at
red
hat.
So
yes,
he's
amazing
and
principal
in
every
kind
of
specific
topic
and
key
contributor.
B
B
A
A
B
Okay,
again,
that's
my
point
because
I
love
history
and
that's
by
the
way.
That's
why
I
decided
to
come
here
to
the
ria,
so
a
city
of
castle.
There
is
no
relationship,
the
direct
relationship
between
cloud
and
agile.
However,
what
can
I
say
is,
after
agile
the
cloud
become
more
and
more
popular.
Why?
Because
the
whole
goal
of
agile
methodology
is
to
deliver
it
fast
and
at
least
fast
as
possible
a
product.
B
So
that's
why
they
had
the
agile
manifesto
and
they
has
they
have
this
four
approach,
so
individual
and
interaction
over
processing
tools,
working
software
over
documentation,
customer
collaboration
over
contract
negotiation
and
responded
to
change
over
following
the
plan.
So
the
whole
point
here
is
basically
to
having,
on
your
mind,
that
why
do
we
have
agile
the
agile
methodology
was
born,
because
we
want
to
decrease
the
it
make
facet,
the
release
of
product
and
how
you
we
do
that
we
break
down
the
silos.
Everything
is
about
to
break
down
the
silos.
B
That's
why
the
whole
point
of
agile
is
make
your
our
client
closer
to
us.
So,
instead
of
put
the
clients
outside
and
wait
like
one
a
year
to
leave
a
version,
we
can
try
to
release
with
more
frequency,
so
the
client
can
give
us
more
feedbacks
right
and
finally,
we
have
the
opportunity
to
put
the
client,
the
client
close
to
the
developers
and
and
but
there
is
some
issue
about
that,
because
we
put
the
clients
close
to
the
developer.
B
However,
we
forget
about
the
operation
team,
and
that
is
not
enough
right,
so
I
have
the
products
I
have
the
source.
I
understood
the
product.
However,
my
product
is
not
on
the
production,
yet
that's
why
we
decided
to
go
deep
on
that
breakdown
silos
to,
and
then
we
created
the
devops
creature
right.
The
whole
point
is
to
make
the
developer
to
work
together
with
the
operation.
So,
instead
of
wait
a
long
time
to
deploy
application,
you
can
decrease
this
kind
of
time
to
do
it
fast
and
fast
and
fast
and.
B
A
B
A
Yeah,
so
we
are
trying
to
break
down
the
silos
across
different
type
of
different
pieces
of
the
the
organization,
and
when
we
came
up
with
containers,
we
could
break
up
the
silos
break
down
the
silo
between
dev
and
knobs.
So
we
got
this
show
together,
although
we
still
had
a
silo,
a
wall
of
confusion
right
in
the
middle
of
the
business
team
and
us
the
tech
people.
So
we
needed
a
way
to.
A
Build
this
bridge
in
between
these
silos
and
get
these
people
together
and
the
way
we
can
do
it
is
by
adopting
the
bees
devops
culture,
where
we
have
also
a
a
way
to
have
artifacts
that
both
business
people
and
tech
people
can
understand,
and
then
that
can
be
delivered
as
in
the
way
that
ops
people
can
understand.
So
that's
mostly
what's
up
with
the
beast
devops,
but
now
that
we
know
what
happened
on
the
project
management
pillar
also,
we
saw
what
happened
on
the
technology
pillar.
A
We
now
have
different
ways
to
deliver
software,
but
not
in
the
process
itself,
but
on
the
on
on
actually
where
we
are
implanting
deploying
that
software.
Where
are
we
deploying
it
and
then
comes
the
cloud,
so
you
said
that
this
everything
we
just
said
is
linked
to
the
beginning
of
the
cloud
era.
Is
that
right?
What
have
you.
B
Yeah,
so
right
now
we
have
no
idea
what
does
cloud
mean
to
us?
However,
the
whole
point
here
is
to
talk
about
cloud.
Is
somebody
else's
problem?
The
whole
point
is
okay.
I
am
a
software
developer,
I'm
a
software
company
and
I
don't
want
to
rush
about
some
details
of
operation.
What
do
I
do?
Basically,
I
paid
you
somebody
else
problem
I
paid
you.
Somebody
else
know
how
I
pay
to
somebody
else
to
help
me
on
that.
That's
pretty
similar.
B
B
B
A
B
B
B
As
a
service,
let's
decrease
a
little
bit
of
the
complexity
and
pay
to
somebody
else
to
handle
that
problem
to
us.
So
let's
go
to
take
and
bake.
So
basically,
the
whole
point
is:
let's
go
to
the
marketplace:
buy
a
ticket
already
and
then
just
put
in
the
fire.
So
I
decreased
the
number
of
know-how
and
also
I
decreased
the
number
of
knowledge
to
make
the
pizza.
So
basically
you
need
to
put
in
the
microwave
and
that
is
infected
to
the
service.
B
A
B
Pass
is
the
point
where
you
decided
to
increase
even
more
right,
so
in
the
pizzas
it's
a
piece
of
delivery,
so
you
don't
need
to
handle
anything
else.
So
take
the
phone
call,
the
pizza
or
maybe
use
some
delivery.
Applications
such
as
ubereat
here
europe
and
us
and
ifood
in
brazil,
and
there
is
it-
you
just
need
to
prepare
the
table
or
perhaps
you
can
eat
pizza
while
you're
watching
us.
So.
A
B
For
example,
hiroko,
so
basically
you
as
developer,
don't
need
to
hand
the
whole
configuration
you
don't
need
to
handle
the
the
installation
of
the
database.
So
you
have
the
source
you
plug
your
source
there
and
the
whole
management
of
infrastructure
plus
the
service
that
your
application
needs
belongs
to
somebody
else,
such
as
rirocco,
for
example,.
A
B
Okay,
you
have
the
option
dine
out,
but
let's
think
about
the
future,
where
you
are
able
to
do
that
right,
right
karina,
so
you
can
go
to
restaurant,
you
eat
the
pizza,
and
that
is
it
so
don't
need
to
hate
to
rush
about
anything
else,
and
that's
amazing,
because
the
whole
complexity
belongs
to
somebody
else.
The
whole
problem
belongs
to
somebody
else.
I
don't
need
to
have
the
know-how.
I
don't
need
to
have
the
infrastructure.
I
don't
need
to
have
anything.
B
I
just
need
to
have
my
wallet
to
pay
to
this
kind
of
service,
there's
exactly
something
where
I
decide
to
go
to
dine
out.
So
I
don't
need
to
prepare
the
pizza.
I
I
don't
even
know
how
the
pizza
works.
I
don't
need
to
stay.
There
sit
down
order
it
and
that
is
it
so
and
pay
for
the
video
right.
So
that
is
it
so.
B
Why
exactly
what
you're
using
right
now
so,
if
you're
watching
us
on
youtube
or
probably
you
are
using
gmail
something
like
that,
so
basically
don't
need
to
to
to
know
how
this
thing
works
right.
So
just
it's
ready
to
you
and
remember
everything
has
a
trade-off.
So
if
you
go
to
sas,
it's
decreased,
the
it's
the
risk,
however,
probably
be
more
expensive.
This
way
so
need
to
handle.
If
you
go
to
the
last
one
where
you
need
to
do
everything
by
your
home,
probably
cheaper.
However,
the
raw
risk
is
on
you.
A
Okay,
I
see
and,
as
you
can
see
in
this
diagram,
that
we
created,
you
have
several
options
other
than
just
yassaso
pass.
So
these
are
the
ways
you
can
actually
deploy
your
application.
Today
you
can
host
just
the
infrastructure
or
have
someone
providing
the
platform
or
you
can
just
consume.
The
software
are
already
delivered,
although
we
have
some
other
flavors
that
are
coming
out
that
are
emerging
from
these
existing
ones.
One
example
is
the
cass
right
otavio,
which
means
container
as
a
service.
B
Yes,
yes,
you're
right,
so
basically,
if
you
go
to
the
nist,
the
seller,
this
you
right
now
have
the
definition
of
cloud
and
the
three
three
ones.
The
three
first
one
are
the
pillar
of
the
whole
approach
and
based
on
that
right
now
we
have
more
and
more
extensions
and
modern
derivation.
As
kind
of
said
before
we
have
a
container
at
the
service
of
cars
where
it
is
exact
in
the
middle
between
yes,
so
it's
in
pass
because
it
decrease
the
complexity.
B
B
A
B
A
B
You
are
exactly
if
you're
a
java
developer,
then
you
need
to
know
how
to
install
a
database
yeah.
You
should
shoot
it,
though
no,
but
usually
in
the
clouds
more
difficult.
So
because
I
need
to
think
about
password
need
to
find
how
to
update
the
user
and
the
password
and
so
on,
the
credential,
the
security
and
so
on.
The
whole
point
is:
when
you
have
accounts
cast.
A
Yes,
basically,
when
you
are
working
with
a
cast
provider,
you
will
have
the
orchestration
platform
where
you
can
deploy
the
containers
that
you
created
see.
So
this
part
will
be
automated
for
you.
So
on
the
other
hand,
what
have
you?
What
we
have
are
the
managed
application
services,
which
means
that
these
are
not
sas
because
they
are
not
the
applications
ready
to
use,
but
these
are
actually
services
to
help
the
developers
the
produce
faster.
A
Why?
Because
they
will
not
have
to
worry
about
the
database,
they
can
just
use
a
managed
database.
They
do
not
need
to
know
how
to
install
kafka
or
how
to
configure
kafka
topics,
because
they
can
just
hire
managed
event
streaming
platform
from
some
provider,
so
they,
then
they
can
just
deploy
the
application
and
connect
to
these
cloud
services.
That's
why
we,
I
call
them.
I
called
it
like
managed
application
services,
because
it's
not
sas,
but
it's
managed
by
someone
else
right
not
have
it
still
on
the
cloud
still
somebody
else's.
B
A
B
A
Exactly
so
like
what
have
you
said,
the
three
main
pillars,
let
me
put
my
here
so
the
framing
pillars
I
ass
pass
and
sauce
okay.
So
what
comes
with
cass
is
that
you
have
a
layer
that
is
in
between
yes
and
in
between
sas,
because
it
would
be
you
would
have
like
infrastructure
provisioned.
You
would
have
networking
storage,
nd
or
container
orchestrator
platform,
but
you
would
not
have,
for
example,
the
the
easiness
to
that
that
prof
heroku
provides,
for
example,
or
platform
message.
A
Juan,
if
it's
that's
not
clear,
please
let
us
know
we'll
keep
on
trying
to
answer
your
questions
and
folks
who
are
watching.
Please
interact
with
us.
We
love
interaction.
Okay,
so
what
have
you
now?
We
know
how
we
can
deliver
different
deliver
our
application
in
different
platforms
across
the
cloud.
We
know
that
we
do
not
have
to
reinvent
the
real
when
we
have,
for
example,
services
that
already
provides
some
functionality
that
we
need
to
use
and
when
we
don't
have
the
know-how,
we
can
just
consume
it.
A
B
A
B
B
B
Oh,
this
one
is
my
favorite
book
because
clear
native
transformation,
the
authors
said
that
is
much
more
than
software
development
and
also
much
more
than
operation.
Basically,
it's
a
methodology
that
embrace
everything
about
soft
development.
For
example,
in
the
past
we
had
waterfall
right
now
we
are
working
with
agile
methodology
in
the
future.
B
A
There's
a
thing
I
want
to
comment
here
so
so
far
we
have
seen
different
definitions,
as
you
can
see,
it's
not
a
standard.
Yet
we
do
not
have
a
final
statement
that
says
what
is
cloud
native
and
now
we
are
going
to
talk
about
the
cloud
native
computing
foundation
statement
for
that,
and
that
is
one
of
the
most
recognized
ones
like
because
it's
from
the
cncf
right.
What
have
you.
B
B
A
B
A
B
Okay
and
yes,
there
are
several
books
that
look
to
cloud
native
as
a
chronological
definition.
For
example,
we
began
with
clouds,
and
then
we
had
the
cloud
friendly
and
then
cloud
ready
and
right
now
we
have
the
cloud
native
application.
Basically,
if
you
have
application
that
was
born
in
the
cloud
thinking
around
the
cloud.
Congratulations,
that's
a
cloud
native
application.
A
B
Yeah
right
now,
some
some
people
know
I'm
I'm
helping
the
jakarta
e
and
also
in
microprofile
and
jakarta.
He
is
has
the
approach
of
cloud
native
application
and
I
decided
to
write
several
books,
several
articles
to
have
my
whole
definition,
because
the
definition
already
yet
at
least
you
don't
have
the
standards
and
to
me.
So
it's
might
have,
might
change
some
time.
A
Perfect,
so
we
have
discussed
the
origins
of
cloud.
What
are
the
options
for
us
to
deploy
our
application
on
the
cloud
and
we
have
discussed
the
concept
of
cloud
native
applications
and
next
thing
we
hear
in
the
java
world.w.
It's
always
the
same
thing
java
is
too
heavy
for
the
clouds.
Me
me
me.
So
what
about
that?
What
about
that.
B
B
B
We
had
a
better
integration
with
containers
with
more
memory
and
we
have
finally
have
a
release
every
six
months.
So,
instead
of
wait
like
three
years,
we
finally
have
the
opportunity
to
every
six
months
have
a
release
where
we
can
enjoy
more
the
jvm
itself.
Okay,
so-
and
one
important
thing
here-
that
let's
say
is
the
improvement
is
about
the
container
itself
is
not
about
doctor.
So
if
you
use
any
kind
of
container
the
jvm,
will
you
improve
to
you
this
application?
Okay,
kernel
next
light,
please
it
has
a
whole
point.
B
So
besides
I
have
several
improvements
in
the
jvm.
We
also
have
the
jakarta
yi
right.
Jakarti
is
a
product
under
the
eclipse
foundation,
it's
worked
with
agile
methodology.
The
whole
point
is
to
release
every
year
a
new
version
of
jakarta
and
the
whole
goal
is
about
cloud
native
right.
So
that's
what
everybody
wants
right
now.
B
B
It's
also
under
eclipse
foundation
umbrella,
and
it's
really
has
a
release
every
fourth
time
that
you
in
the
year
and
for
sure
it's
faster
than
jakarta
and,
as
you
can
see
here,
eglist
foundation
has
several
apis
to
help
the
java
developer
to
to
make
an
application
the
cloud.
So
we
have
right
now.
The
newest
version
is
jakarta
e9.
B
B
A
A
As
you
can
see
today,
we
have
several
options
to
deliver
java
applications,
either
with
jakarta
or
with
micro
profile
spec,
and
you
you'll
see
new
frameworks
for
those
who
are
working
with
java
for
longer.
You
will
see
that
we
have
new
frameworks
like
halidon
micronauts,
apayata
quarkus,
and
they
are
all
adopting
the
new
microprofile
standard.
B
That's
wonderful
because,
based
on
these
standards,
we
are
possible
to
use
more
features.
I
mean
you
can
create
some
wrapper
on
the
this
kind
of
specification,
for
example,
take
for
example,
spring
quakers,
and
it
does
happen
because
we
have
a
strong
community
and
also
we
have
a
strong,
strong
specification
team
to
help
on
this
side.
A
So
basically,
this
allows
us
to
compile
the
code
directly
to
the
co
to
the
to
the
code.
That
is
understandable
by
the
machine
that
we
are
deploying
the
application.
So
when
we
do
that,
we
are
actually
doing
the
work
from
the
inter
interpreter
and
the
c1
algorithms
in
the
jvm
in
order
to
compile
this
before
we
execute
it.
So
it's
really
really
really
really
fast,
although,
although
there
is
always
an
aldo
right,
what
have
you,
although.
B
Yes,
inside
the
g
jvm
with
it
is
it's
like
we
try
to
define
a
race
between
a
motorcycle
and
a
plane
with
one
mire.
Probably
the
motorcycle
will
be
faster
than
the
plane,
but
that
doesn't
mean
that
you
can
cross
the
whole
ocean
with
motorcycle
just
because
the
first
mile,
the
motorcycle,
is
faster
than
a
plane.
So
we
need
to
think
about
this.
The
whole
point
of
serverless
is
the
application,
will
run
once
and
then
gone.
You
don't
need
to
do
run.
A
Perfect-
and
this
is
there
since
java
9,
I
think
the
option
to
compile
classes
in
with
ahead
of
time,
compilation
and
if
this
is
getting
pretty
hot
in
the
java
ecosystem,
because
it
allows
us
to
deploy
java
code
where,
for
these
use,
cases
that
we
mentioned
remember
that
there
is
no
silver
bullet.
So
for
the
appropriate
use
cases,
it
is
getting
really
adopted
when
you
are
going
to
the
cloud
and
you
need
resources
that
are
lightweight
and
then
start
faster
and
execute
really
fast.
So
it's
becoming
a
thing
right
out
of
you.
B
B
A
So
now
we
know
that
we
have
the
ecosystem
to
deliver
our
applications.
We
are
well
supported
by
the
java
standards
and
the
on
the
market,
and
we
have
several
options
on
the
cloud
and
then
we
start
to
come
up
with
several
services
because
we
have
smaller
services.
Therefore,
we
have
a
bigger
number
of
services
running
on
this
for
this
to
achieve
the
same
goal
and
we
need
to
make
them
interact
somehow,
so
we
need
to
either
use
orchestration
or
choreography
to
have
them
interacting.
A
I'm
sorry
so
if
you
think
about
ubereats,
for
example,
where
you
have
four
services
to
do
the
order
to
accept
the
receive
the
payments
and
do
the
execution
of
the
order
and
to
actually
cook
the
order
and
finally
ship
this
order,
if
you
think
that
you
have
four
services,
how
would
you
make
them
talk
to
each
other?
Would
you
make
like
end
point
to
end
point
communication?
A
Is
that
the
way
to
go?
Is
that
really
the
best
way
to
go
like
every
time?
Sometimes?
Well,
maybe,
but
imagine
what
would
happen
if
the
number
of
services
increased
here
so
think
about
it?
So,
let's
first
talk
about
choreography
because,
like
one
of
the
the
most
prominent
voices
in
the
microservices
community
is
sam
newman,
who
wrote
this
book
building
microservices
and
he
said
that
you
can
either
use
a
service
to
orchestrate
other
services.
A
So
these
orchestration
services
take
care
of
invoking
the
other
services
in
order
to
achieve
the
final
goal.
Okay,
so
this
is
the
orchestration
when
you
go
for
the
event
driven
when
you
go
to
the
event
driven
our
orchestration
choreography,
I'm
sorry!
When
you
go
to
the
org
to
the
choreography
strategy,
you
actually
create
services
that
are
independent
and
decoupled
from
each
other
and
instead
of
invoking,
it
actually
reacts
to
events
or
it
meets
events.
So
in
this
example,
for
example,
in
this
example,
each
service
would
be
responsible
for
notifying
that
something
happened.
A
So,
okay,
I
received
an
order.
That's
it
that's
an
event.
Then
the
other
services
would
react
to
these
events.
Oh
there's,
a
new
order.
I
need
to,
I
don't
know,
invoke
the
kitchen
in
order
to
cook
it.
Okay,
so
with
that
being
said,
we
can
think
about
the
use
case
itself.
So
let's
say
that
talking
about
choreography,
okay,
so,
okay,
let's
say
that
a
new
order
was
created
and
we
this
service
will
emit
a
new,
a
new
event,
new
order
created.
What
would
happen
is
instead
of
invoking
another
service.
A
This
service
can
actually
react
to
it,
so
it
was
the
only
one
that
was
created,
so
I
need
to
process
the
payments
next.
It
finished
the
payment
and
the
execution
service
will
see
a
new
payment
order.
Event
was
immediate.
I
need
to
start
the
execution
my
execution
is
done.
Finally,
the
shipping
service
could
react
to
this
event
in
order
to
start
the
shipment
karina.
But
how
do
they
communicate?
How
do
they
see
these
events?
Actually,
through
the
broker?
A
We
can
think
of
technologies
like
kafka,
who
are
really
popular
in
open
source
that
can
actually
serve
as
the
communication
layer.
So
it's
it
doesn't
have
any
business
intelligence,
it's
just
there
for
the
communication
just
to
handle
the
events
that
happen.
So
it
has
the
broker
and
whenever
a
service
needs
to
emit
an
event,
it
will
publish
an
event
to
a
topic,
for
example,
and
multiple
services
can
subscribe
to
that
topic
in
order
to
react
to
the
events.
Okay,
is
that
clear?
B
A
Okay,
so
now
we
should
talk
a
little
bit
about
orchestration.
Why?
Because
orchestration
was
heavily
used
before,
where
you
had,
we
had
a
large
and
heavy
orchestrator,
a
platform
that
could
that
had
hold
the
business
logic
and
it
knew
where
to
call
each
part
of
the
the
architecture
each
service
in
the
architecture.
A
So
how
does
that
evolved
to
something
that
we
can
actually
use
in
cloud
environments
right?
So
we
start
with
traditional
processes.
We
where
we
have
also
a
lot
of
manual
tasks,
so
we
have
still
talking
about
the
same
example
of
uber.
It's
a
really
simplified
example.
We
have
a
human
task
where
someone
needs
to
accept
the
request.
Based
on
this,
we
can
either
notify
the
customer
or
cook
pack
and
deliver
this
event
with
service
orchestration.
What
we
can
do
is
actually
change
some
of
these
tasks
to
automated
tasks.
That
can
invoke
like
this.
A
A
Next,
we
have
the
shift
octavio.
So
what
is
a
mixed
way
of
orchestrating
these
services
in
the
cloud
environment?
What
we
do
have
right
now
is
the
possibility
to
not
only
invoke
directly
but
actually
to
emit
events
through
business
processes.
So,
as
you
can
see
here,
this
can
be
a
process
that
is
started.
A
A
B
B
B
B
I
don't
go
deep,
I'm
not
able
to
go
deep
because
the
time
is
not
enough,
but
basically,
if
you
have
application
piece
use
repository
use
a
management
to
your
the
instead
of
food
library
inside
the
code,
the
source
code,
for
example,
maven
npm
gradle,
something
like
that
configuration.
Please
don't
put
your
password
credential
inside
your
repository,
the
whole
goal.
The
the
main
goal
here
is
a
developer
does
not
need
to
know
about
the
credential
on
production,
okay,
the
whole
point.
A
B
B
Yes,
we're
improving
that
side
so
right
now,
it's
more
overwriting
environment
configuration
instead
of
go
in
some
xml
file
configuration-
and
it's
is
my
perspective-
is
much
easier
because
you
can
really
find
everything
that
you
want
beside
the
operation
configuration
that
is,
there
are
these
books
here.
Those
are
amazing.
No,
no!
The
next
slide.
Yes,
clean
code
clinic
teacher,
ddd,
effective
java.
So
briefly,.
B
However,
that's
not
totally
true
okay,
so
there
are
several
books
that
include
the
new
surname
books
and
he
usually
in
the
beginning
of
the
book,
said
that
okay,
there's
no
truth,
that
you
must
use
microcephs
because
everybody
else
are
using
if
you
use
it.
So
this
book
here
is
by
a
friend
of
us
at
aga
and
he
wrote-
and
he
said
basically
just
because
netflix
and
amazon
sites
use
microservice.
A
B
A
A
B
This
is
important,
however,
there's
a
couple
article
that
said:
okay,
cloud
native
is
not
about
containers,
so
there's
a
culture
around
the
cloud
native.
The
whole
point
is
everybody.
Has
your
whole
definition
of
cloud
native
to
me?
Is
yes,
but
to
somebody
else
probably
not,
and
and
if
you
don't
have
the
standard
to
cloud
native,
that's
fine.
A
Perfect
so
as
of
today,
what
we
are
seeing
is
that
most
companies
are
migrating
for
containerized
applications,
because
it's
easier
for
ops,
for
example,
to
deploy
it
or
you
can
deploy
it
anywhere.
The
same
way
you
do
with
any
application.
So
that's
why
we're
going
to
now
talk
about
how
you
can
use
best
practices
for
java
and
containers.
So,
first
of
all,
you
should
know
that
docker
is
not
a
standard
okay.
So
there
is
the
open
containers
initiative
which
is
responsible
for
actually
defining.
A
What
should
an
image
has
to
define
the
standards
across
the
image
creation
in
the
way
that
you
run
your
image
containers
your
containers,
so
an
example
of
this
is
that,
for
example,
you
can
instead
of
using
docker,
you
can
use
hey.
Mr
sorry,
from
blank
bangladesh
you
can
use,
for
example,
builder
with
builder,
you
can
create
an
image
that
follows
the
standard
and
docker
images
also
follows
the
standard.
A
Does
it
I'll
tell
you,
I
think
it
does
right,
dlc,
okay,
so
whenever
you
create
an
image
that
it
follows,
dlci
standard,
for
example,
with
builder
you
can
create
it
run
it
in
a
container
that,
with
whichever
platforms
that
supports
the
oci
standard,
an
example
is
you
could
create
a
container
using
docker
or
using
pogman.
So
there
are,
there
are
several
other
options
for
us
to
get
started
in
the
container
world.
What
have
you.
A
A
A
Okay,
so
think
about
two
moments:
when
you
are
creating
an
a
container
okay,
there
is
the
moment
where
you
build
the
image,
so
you
are
creating
the
image
and
there
is
the
moment
when,
where
the
container
will
run
so
you
think
about
view
time
and
you
think
about
runtime,
okay.
So
first,
let's
talk
about
the
build
time.
When
you
are
talking
about
the
build
time,
you
need
to
know
that
you
should
also
care
about
the
single
concern
principle,
which
is
similar
to
the
s
in
solid
pattern.
A
But
in
this
context
each
container
should
solve
a
single
problem
and
solve
it.
Well,
if
you
need
to
do
something
else,
maybe
you
could
go
and
add,
for
example,
a
side,
a
sidecar
or
use
another
image.
So
remember
the
single
concern
principle.
Another
thing
is
the
self-containment
principle,
which
means
that
every
library
that
your
application
needs
to
run
everything
your
application
needs
in
order
to
get
started
should
be
within
this
image.
So
it's
self
contained.
Okay.
A
A
B
A
A
The
metrics
can
collect
the
logs,
so
you
can
use
and
plug
into
your
application,
for
example,
fluency
log
stash,
prometheus
or
grafana
in
order
to
get
to
know
what
is
going
on
in
your
pods,
okay
in
your
container
or
pods,
depending
if
you're
running
on
a
kubernetes
thinking
about
the
container.
So
the
other
thing
is
the
lifecycle
conformance
principle,
so
you
should
use
graceful
shutdown
whenever
you're
you
whenever
you
can
so
you
should
worry
about.
A
How
is
the
way
that
you're
going
to
what
is
going
to
happen
when
you
start
your
container
and
when
it
shut
downs?
Okay,
next
is
the
process
disposability
principle,
so
this
is
linked
to
the
immutability
principle,
because
when
you
think
about
this
principle,
the
process
disposability,
you
should
have
in
mind
that
the
container
is
volatile.
Okay,
it
can
be
destroyed
and
recreated
as
many
times
as
we
need.
So
your
contain,
you
should
have
in
mind
that
the
startup
time
and
the
shutdown
time
should
be
quick.
A
A
Finally,
the
last
one,
which
is
the
runtime
confinement
principle
where,
since
we
follow
the
self-containment
principle
in
build
time,
which
is
the
image
itself
contained,
then
these
applications
should
only
use
the
resources
that
we
provide
to
it.
Okay,
so
it's
containered
in
contain
confined
in
distance
otavio.
I
spoke
a
lot
already
next.
A
So
we
have
already
spoke
about
everything
like
where
we
can
deploy
java
standards
java
applications,
how
to
orchestrate
services,
yeah
right,
luis
right,
how
to
orchestrate
services
and
how
to
create
the
images
when
we
are
building
the
image
and
when
we
are
running
this
container.
But
what
about
the
delivery?
A
Traditionally,
we
did
this
since
forever
octavio
since
clouds
we
already
did
this
before
containers
see
icd,
it's
something
everyone
knows
about
it,
but
not
everyone
does
it
right,
because
you
should
know
that,
for
example,
when
you
are
practicing
ci
continuous
integration,
you
should
build
your
software,
build
your
database
and
run
your
tests
at
every
change.
Okay
and
every
change
you
do
so
you
should
open.
You
should
commit
frequently
your
that's
something.
B
A
Yes
and
then
you
create
the
unit
test
to
guarantee
that
it's
broken
right.
That's
why
you
create
a
bug.
Oh
I
gotcha
yeah,
so
we
still
need
to
tell
people
to
create
unit
tests
and
all
that
should
pass.
So
if
you're
doing
this
you're
doing
ci
right
and
then
you
think
am
I
doing
cd,
but
there
are
two
types
of
cds
that
have
not,
and
it's
not
cd
and
dvd.
It's
like
continuous
delivery
and
continuous
deployment.
B
Okay,
so
cicd,
as
I
mentioned
before,
the
whole
point,
is
about
deliver
the
applications
fast
and
break
down
the
silos
and
continue
the
delivery
content
deployment.
The
whole
goal
here
is
once
you
run
your
ci
and
you
you
open.
If
you
think
about
github,
for
example,
you
you
open
up
your
you
need
to
run
the
whole
test
right.
So
you
you
run
the
world
test.
B
A
So
that's
the
biggest
difference
in
between,
for
example,
continuous
delivery
and
deployment.
When
you
do
cd,
you
get
you
do
everything,
but
you
do
not
deploy
it.
You
will
have
your
package
ready
for
deployment
either.
It
fits
a
jar
if
it's
a
a
native
file,
if
it's
a
war
file,
it
will
be
there,
but
it's
not
beautiful.
It's
not
going
to
be
deployed.
When
you
have
continuous
deployment,
then
yes,
you
have
the
ci
cd
with
con
ci
and
continuous
deployment.
A
So
have
that
in
mind,
although
what
have
you
we
are
talking
about
cloud
cloud
and
cloud,
how
has
how
the
how
has
this
shift
shift
to
the
cloud
environment?
Well,
what
we
see
is
that
before
we
used
to
have
those
servers,
let's
say
jenkins,
for
example,
that
would
do
this
task
of
knowing
like
how
to
build
the
pipeline
and
etc.
A
We
would
have
the
workers
and
yeah
that's
how
we
did
it.
So
today,
in
cloud
native
environments,
we
actually
need
cloud
native
ci,
cd
tools.
So
that's
why
we,
there
are
new
tools
on
the
market
like
tekton,
for
example,
that
is
a
serverless
ci
cd
server.
You
can
just
define
whatever
the
pipeline
is
a
in
kubernetes
or
in
openshift,
for
example,
and
it
would
trigger
each
task
within
a
its
own
container
and
it
would
do
everything
without
a
master
server,
for
example
like
we
used
to
have
in
the
in
the
jenkins
world.
A
So
as
we
move
to
a
cl
hybrid
cloud
environment.
Because
of
all
that
scenario
that
we
explained
at
the
beginning,
we
see
that
the
engine
should
be
able
to
grab
the
source
code,
packaged
in
it
package
the
application
in
a
container
and
deliver
to
whichever
cloud
environment
it
is
either
it's
physical
virtual,
apply
a
private
or
a
public
cloud.
Okay,
so
that
is
all
that
happened
within
this
time
octavio.
A
A
No,
it's
okay!
So
guys.
Let
me
show
you
here
and
I'm
sorry
about
my
voice
today.
I
just
woke
up
with
this
yeah.
It
is
what
it
is,
but
here
we
go
so
remember.
We
are
talking
about
a
microprofile
specification,
so
here
I'm
used.
I
have
a
application
that
is
built
on
top
of
carcass
and
this
application
is
an
application
that
can
automate
travel
requests.
Okay,
so
we
have,
for
example,
travels,
and
whenever
you
have,
you
want
to
travel
to
a
place
we
need.
We
have
an
application
that
can
not
pay
you.
A
If
you
need
a
visa
and
if
your
visa
is
approved,
you
will
actually
get
a
hotel
and
a
book
flight,
a
hotel
and
a
flight
booked
for
you.
Okay,
finally,
remember
when
I
told
you
that
we
can
have
event
driven
processes.
This
is
what
we
have
right
now.
So
this
is
all
my
application.
Is.
It
has
here
a
automatic
check
based
on
decision
tables,
I'm
using
libreoffice
here.
So
please
don't
mind
about
how
it
looks.
I
really
don't
dislike
this,
but
you
can
see
that
these
are
my
business
rules.
A
Okay,
so
whenever
I
have
a
destination
country
that
is
u.s,
but
the
person
is
polish,
then
a
visa
is
required.
So
this
is
my
business
rule.
It's
pretty
easy
to
understand.
I
don't
have
to
look
at
any
code,
and
that
is
what
is
being
invoked
right
here.
Okay,
so
this
file
that
I
opened
is
this
one
right
here.
A
As
you
see,
when
I
apply
for
a
visa,
I
have
a
human
task
and
the
event
is
going
to
be
fired.
Who
is
going
to
react
to
these
events?
The
application
that
is
going
to
react
to
this
event
is
the
visa
application.
Okay,
this
visa
application
has
another
process,
and
this
process
right
here
is
responsible
for
checking
if
the
visa
is
approved
or
rejected.
A
It
also
contains
business
rules,
okay,
but
these
ones
are
implemented
with
drl,
and
so
basically,
this
process
will
re
is
a
service
and
the
other
is
another
microservice
both
running
on
carcass.
Okay,
these
processes
are
now
so
the
tooling
that
is
actually
providing
us.
These
process
capabilities
and
rules,
capabilities,
we've
even
driven
possibilities
here
is
cogito,
so
these
cogito
applications
are
actually
deployed
in
openshift
running
right
now.
A
On
the
cloud
and
in
openshift
I
have
deployed
this
through
an
operator,
so
I
have
an
application
that
uses
a
that
uses
business
process,
capabilities
with
cogito
that
uses
a
key
value
store
with
infinite
span,
and
that
is
going
to
communicate
with
each
other
using
stringsie,
which
is
the
open
source
version
for
kafka
by
redhead.
So
you
can.
We
were
going
to
communicate
using
stream
z.
So
I
have
these
three
operators
for
cogito.
A
You
can
see
that
I
have
the
two
applications
right
now:
okay
and
it's
running
so
remember,
we
spoke
about
cast
and
pass.
As
you
see,
this
is
very
technical
when
you
have
a
so.
The
bpmn
extension
I
am
using
in
vs
code
is
the
cogito
one
is
the
business
automation.
Let
me
show
you
here:
okay,
extensions,
yes,
so
beginning.
A
So
you
can
either
use
this
one,
my
red
hat,
okay,
this
is
already
a
version
that
is
compiled
and
tested
to
be
supported
in
the
official
product,
but
it's
free
for
use,
and
there
is
also
the
cogital
one
that
will
have
more
frequent
releases.
Both
today
are
on
the
same
status,
but
you
can
use
the
one
you
you
prefer
any
with
it.
You
can
either
use
open
like
vpmn
and
dmn
files
and
test
scenarios.
A
Okay,
with
this,
I
have
my
application
deployed
my
two
applications
and
if
I
want
to
check
it
up,
I
could
just
go
to
the
routes
which
are
which
were
created.
No
problem
have
anyway
abandon
it.
Yes,
anytime,
just
let
me
know.
Okay,
if
you
have
more
questions-
and
I
have
here
three
routes,
why
do
I
have
three
routes?
I
have
a
route
to
my
travel
application
to
my
visa
application
and
a
data
index
which
allows
me
to
search
through
all
the
data
in
there.
A
A
Okay,
so
here,
as
you
can
see,
I
already
have
a
new
request:
okay
for
carino
varela,
oppa,
okay
to
new
york,
the
visa
is
required,
so
this
was
defined
based
on
this
those
automatic
rules-
and
it
is
not
yet
approved.
So,
as
you
can
see,
I
have
yet
no
photo
and
yet
no
flight
because
on
the
process
we
should.
A
A
Because
on
the
process,
we
are
still
on
the
part
where
we
need
here
the
approval
for
the
visa
okay,
I
have
applied
for
the
visa
right
now,
so
let
me
apply
for
the
visa
task
apply
for
this
application.
When
I
do
this,
the
other
service.
So
now
I
have
sent
this
and
we
have
here
an
event-based
gateway
for
those
who
are
advanced
users
of
bpmn.
A
Well,
if,
whenever
I
get
a
rejection
for
a
visa
or
the
visa
proof,
I
will
go
to
the
path
that
happens
first,
okay
to
the
event
that
happens
first,
so,
let's
see
here,
I
have
the
passport
number
and
the
duration
12
days
and
I'll
submit
my
application
on
the
code
on
the
cogito
visas
application.
I
got
a
new
request
that
is
karina,
invite
here,
u.s,
okay,
my
visa
was
approved
once
my
visa
was
approved,
it
was
approved.
A
So
this
event
right
here,
was
emitted
and
published
to
the
catholic
chocolate
and
my
dogs
are
crazy
right
here,
because
we
are
emitting
events
for
everywhere,
so
I
published
karina.
But
how
is
the
configuration
right
now
here?
As
you
can
see,
I'm
publishing
messages
to
the
visa
approved
topic.
Okay,
once
it's
approved.
This
message
right
here
is
listening
to
the
visa
approved
topic,
so
I
emitted
from
this
process
to
say
I'm
finished
and
this
other
process
reacted.
So,
okay,
nice.
I
can
now
move
on
with
the
approval.
A
If
we
go
back
here
to
the
travel
application,
you
will
see
that
now
my
hotel
and
flights
are
booked
kind
of.
But
how
can
I
query
these
applications,
so
you
can
use,
for
example,
the
data
index
with
refql,
and
you
can
search
either
by
the
data
that
you
have
s
variables
in
there.
You
can
query
it,
so
I'm
searching
all
the
travels.
A
It's
pretty
easy
to
read
right.
So
all
the
travels
where
the
traveler
last
name
has
var
of
varela.
So
I
can
see
here.
I
have
two
travels:
okay
for
varela.
Another
way
we
can
query
if
we
can.
If
we
just
want
to
go
for
a
traditional
array,
we
can
go
and
query,
for
example,
for
the
process
instances
and
for
the
id
I
could
have,
for
example,
the
let's
see
variables
and
then
I
would
have
all
the
variables
for
each
process.
A
Okay,
so
it's
pretty
easy
to
deal
and
handle
when
you
work
with
like
that
karina.
But
how
are
you
coding
the
application
like?
Where
is
the
the
rest
endpoints?
Where
is
everything?
Basically,
that's
the
magic
quakers
allows
us
to
have
defense.
Startup
time
allows
us
to
have
the
the
lightweight
environment
to
run
on
the
cloud,
and
we
can
add
the
process
management
capabilities
on
top
of
it
using
cogito.
So
cogito
is
responsible
for
parsing.
These
codes
pre-compiling
it
before
it.
The
application
starts
and
have
this
and
and
generate
the
rest
endpoints
for
you.
A
So
if
you
open
the
target
in
the
target
directory,
you
will
see
all
the
endpoints
generated
in
there.
Okay,
so
otavio.
This
is
an
example
of
how
we
can
deliver
business
applications,
for
example
in
the
cloud
environment
following
good
practices,
because
I
already
showed
it
down
here,
because
I
was
afraid
of
time,
but
it
I
we
could
show,
for
example,
the
best
practices
during
the
build
time
during
the
container
run
time.
Oh
yeah.
B
That's
amazing
because,
for
example,
you
can
create
and
include
the
the
visit
rules
easily
right,
for
example,
right
now
also
mention
about
travel
right
now.
You
need
to
have
a
test
about
kovite.
So
probably
it's
much
easier.
They
put
several
if
on
that
condition
right,
so
I
need
to
put
just
one
more
event:
there
connect
and
put
the
largely
beside.
A
We
need
to
understand
that
as
architects
and
developers,
we
should
have
the
tools
in
our
toolbox
to
solve
each
problem
with
the
in
the
right
way.
So
if
you
need
to
solve
a
business
problem,
go
and
use
a
business
automation
tool
if
you
need
to
solve
inventing
problems
going
using
the
streaming
platforms.
A
So
if
you
have,
for
example
like
what
have
you
said,
a
rule
that
you
have
too
many
changes,
frequent
changes
or
if
you
have
business
users
that
want
to
change
frequently,
let's
say
discounts
in
a
store,
for
example,
or
if
you
have
long
business
processes
like
loan
approvals,
or
I
don't
know
any
type
of
business
processes
that
you
can
run
in
a
company.
Those
are
also
suitable
for
cloud
environments,
and
this
is
all
java
based
running
with
carcass
containerized
and
running
on
openshift,
which
is
a
flavor
of
kubernetes.
A
B
A
B
A
Being
presented,
I
just
want
to
go
what
have
you
through
it
all?
So,
to
recap,
quick
recap:
we
learned
and
if
you
don't
brief
for,
if
you
don't
remember
any
of
these
topics,
you
can
go
back
and
watch
the
recording
right
otavio,
because
we
discussed
origins
of
cloud.
What
are
the
options
for
you
to
deploy,
deploy
your
application
in
cloud
environments?
Do
you
want
to
use
the
ads?
Do
you
want
to
use
sas?
Do
you
want
to
use
cast?
So
what
are
the
options?
How
where
do
I
deploy?
A
B
B
On
our
stages,
yes,
exactly
I
mean
when
you
talk
about
okay,
that's
not
the
right
definition
of
cloud
native.
We
need
to
go
deep
and
self-definition
to
have
your
own
definition
right.
So
it's
a
challenge,
but
I
t
is
an
area
that
you
need
to
read
a
couple
times
every
time,
and
sometimes
you
learned
like
one
year
ago
it
became
deprecated
this
year.
A
So,
and
we
also
saw
the
what
different
options
for
you
to
containerize
your
application,
what
are
the
best
practices
when,
during
build
time
of
the
image
and
during
runtime
of
the
container,
we
saw
about
the
shift
on
the
cicd
platforms,
what
they
used
to
be
and
how
they
are
right
now,
and
we
also
saw
how
we
can
actually
deliver
this
cloud.
Enviro
cloud
applications
within
applica
platforms
like
openshift,
so
I'll
tell
you
with
that.
A
A
Yeah
gerald
asked
if
I
could
repeat
the
name
of
the
tool
for
ci
cd,
not
jenkins,
so
tactile
is
also
open
source.
Basically,
everything
I
work
with
is
open
source
because
red
hat
is
full
open
source,
so
it's
a
open
source
and
it's
backed
by
red
hat,
because
this
is
actually
used
on
open
shift
pipelines.
A
B
Thank
you
everybody.
So
here
you
go
as
you
can
see
here,
that
a
qr
code,
where
you
can
get
the
whole
reference.
The
whole
book
that
you
mentioned
today,
the
the
github
story
with
the
code,
so
that
is
it.
Hopefully
we
enjoy
and
see
you
soon
as
possible
and
yes,
stay
safe
at
home,
and
hopefully
we
are
able
to
do
this
kind
of
thing,
more
close,
like
a
conference
or
any
kind
of
meetup.
A
A
B
A
A
I'm
sorry
yeah,
my
throat
is
really
bad
right
now,
so
I'm
not
sure
I'm
not
an
expert
on
vertex,
but
what
I
do
know
is
that
kafka
vertex
being
used
more
as
a
reactive,
runtime
environment.
For
you
to
to
code,
your
apps
and
kafka
is
more
of
the
the
the
layer
where
you
you
do.
The
communication
within
it
yeah.
B
B
B
A
B
Out
of
it
is
the
application
service
to
handle
that
you
can
use,
for
example,
broker
you
can
use,
for
example,
I
don't
know
some
amazon
sqs
any
kind
of
queue,
and,
yes,
you
can
use
vertex
as
application
layer
api
to
handle
that
with
reactive
api.
So
yes,
okay,
those
are
different,
so
you
can
combine
and
mainly
because
they
are
not
competitors
so
there
they
can
work
together.
B
B
Therefore,
you
have
a
vendor,
locking
if
you'd
like
to
choose
between
another
application,
you're
not
able
to
so
we
have
a
vendor,
locking
and
another
hand.
We
have
jakarta
e.
That
is
a
specification
that
viewed
by
a
community,
we
have
a
huge
amount
of
communities
and
companies
behind
the
jakarta
e.
B
However,
we
need
to
make
sure
that
everybody
has
a
voice,
and
this
process
has
time
to
make
happen
right,
because
you
need
to
have
the
time
to
to
make
sure
that
everybody
was
listening
to
take
any
kind
of
decision,
and
this
take
time
to
happen
and
those
are
not
enemies,
because
if
you
like,
take
a
look
and
spring
is
spring,
is
a
wrapper
of
java
e
or
jakarta
e.
So
if
you
take
a
look,
for
example,
spring
data
jpa
is
basically
a
wrapper
of
jakarta.
B
A
B
Yes,
I
guess
the
question
is:
what
happened?
Does
the
spring
move
to
jakarta?
Yes,
they
did
so
they're
they're.
Looking
for
for
that
move
and
as
as
far
no
people
to
the
company
behind
spring
are
helping
a
lot.
The
jakarta
e
community,
for
example
in
jakarta,
no
sicko,
we
have
oliver
in
the
mailis
and
oliver
is
the
guy
behind
spring
data
and
to
me
spring
data
is
the
more
mature
solution.
When
you
talk
about
java
and
persistence
technology,
for
example,
no
secret
database.
B
A
The
question
is
called
support:
micro
profile,
it
like
it
actually
runs
on.
It
can
be
added
to
a
micro
profile,
runtime
like
quarkus,
or
you
could
add
it
to
a
spring
boot,
for
example.
So
it
really
depends
on
where
you
want
to
run
it
either
with
carcass
or
with
springboards.
It's
your
choice.
Okay,.
B
Yeah
corrito
is
amazing,
I'm
just
kidding
you
know,
because
quakers
do
not
support
complete
td
microprofile,
because
the
cdi
point
and
there's
no
bad
because
we're
talking
about
innovation,
so
kwaku
was
an
evolution
in
the
area
where
they
decided
to
don't
use
reflection
and
right
now,
they're
looking
to
standard,
that's
why
they.
They
are
working
right
now
in
cdi
light
to
make
that
possible.