►
From YouTube: Jakarta EE + MicroProfile + Kubernetes = The Productivity Dream #slideless | JakartaOne Livestream
Description
Adam Bien- Jakarta EE / Java EE / Java SE Developer
---
JakartaOne Livestream Cloud Native for Java (CN4J) is a one-day virtual conference for developers, engineers and technical business leaders with the focus of building enterprise Java on Kubernetes.
This virtual event is a mix of expert talks, demos, and thought-provoking sessions focused on enterprise applications implemented using open source vendor-neutral Jakarta EE and Eclipse MicroProfile specifications on Kubernetes.
A
Okay,
hi
everyone.
Can
you
hear
us
welcome
to
session
five?
My
name
is
Shabnam
al
and
I'm.
Your
host
for
this
session
hope
you
have
enjoyed
this
event
so
far,
we're
seeing
a
great
attendance
so
far
and
quite
happy
with
that.
So
thanks
everyone
for
joining
and
for
this
session
we
have
the
one
and
only
Adam
in
here
Adam.
Your
talk
has
a
very
interesting
topic.
So
I,
let
you
take
it
away.
Yeah.
B
Hello
and
welcome
back
to
Jakarta
live
stream.
This
time
is
CN
cn,
forge
a
meeting
but
ism
it's
very
similar
infrastructure
and
I
watch
the
talks
from
Sebastian
and
a
bit
from
rot,
and
they
already
covered
a
corpus.
So
I
will
try
not
to
avoid
Koko's
a
bit
in
my
talk
and
focus
on
the
on
the
jakarta
ii
service
or
run
times
okay,
so
my
name
is
Adam
beam,
I'm,
a
freelancer
consultant
and
yeah
Java
developer
actually
and
I
really
still
enjoy
Java.
Sometimes
I
also
enjoy
clouds.
B
It
really
depends
whether
it
makes
sense
or
not,
and
yeah
and
I
would
call
myself
for
the
last
three
weeks
since
really
three
weeks
senior
deaf.
Why
this
because
I
successfully
managed
to
edit
a
ml
file
submitted
to
kubernetes
and
it
actually
without
any
errors,
so
usually
I.
Everything
worked
well
except
the
last
part.
You
know
editing.
The
Amalia
always
made
something
wrong
and
I
had
to
know
to
analyze
the
go
arrows
and
yeah
so
today,
I
will
also
try
to
avoid
editing,
yeah
Mille,
but
I
will
show
you
some
Yambol
and
okay.
B
So
this
was
the
introduction,
so
I'm
senior
deaf.
Now
so,
if
you
like
something
new,
a
hex
life
has
online
interactive
workshops
like
this
one,
and
there
is
in
an
pod
cast
a
hex
fm.
So
I
could
be
interesting.
There
are
some
some
conversations
with
people
from
Jakarta
and
an
micro
profile
community,
even
with
Tanya
and
Mike
Milankovitch
from
eclipse
foundation.
B
So
if
you
like
at
Union-
and
this
is
probably
the
most
interesting
thing
for
you-
this
is
my
question
and
answer
shows
once
a
month
first
month
of
the
amount
of
the
of
the
month,
8
p.m.
CET
and
ya.
Ask
whatever
you
like,
but
please
don't
send
me
emails
with
taking
a
few
questions.
So
this
this
is
the
deal
so
first
cloud
native
and
micro
profile
as
a
probably
it
was
already
introduced
several
times
and
just
my
point
of
view
of
the
entire
thing
and
and
what
is
the
relation
between
Jakarta
in
micro
profile?
B
So
very
briefly,
for
me,
Jakarta
II
is
more
like
you
know
the
operating
system
it
was
designed
by
or
designed
yeah
the
APS
were
designed
by
different
vendors
working
together
and
is
implemented
by
several
runtimes
or
service
and
I.
Don't
expect
too
many
changes
in
the
upcoming
years?
Why?
Not?
Because,
because
it's
it
is
already
very
good
and
actually,
if
you,
if
you
will
follow
you,
know
my
reactions
to
five
years
ago,
if
Jakarta
ee8,
you
know
came
late
and
or
Jakarta
7
was
was,
was
delayed
a
bit.
B
So
I
was
not
that
you
know
angry
or
or
sad,
because
I
was
happy
with
Jakarta
e6
and
still
I'm
still
I
am
so
it's
not
like.
You
know,
I'm
waiting
for
the
next
big
thing,
so
so
the
base
layer
here
its
Jakarta
II,
so
it's
very
stable
layer
and
we
will
get
innovation,
but
not
I.
Don't
I
don't
expect
any
revolution,
so
I'm
micro
profile
also
stable,
but
it's
not
driven
by
by
application
service
vendors.
B
But
from
my
perspective
it
is
influenced
by
the
cloud
native
foundation,
mostly
by
CN
CF,
and
why
that
because,
if
you
look
at
that,
so
we
have
matrix
which
is
open,
matrix
or
or
open
telemetry
JWT
is
an
arab
seed.
Health
is
I
would
say
heavily
influenced
by
Cuban,
it
is
config
is
absolutely
Cuban.
It
is
and,
and
so
cloud
compatible
and
ok
rest
client
is
our
invention.
I
would
say
6f
or
rest
easy
that
this
was
how
this
started
and
what
the
rest
is.
Is
you
can
inject
an
interface
open?
B
Api
is
also
a
standard.
The
open
standards
previously
known
as
swagger
and
open
tracing
is
distributing
tracing
API
with
Jaeger
tracing
also
a
part
of
CN
CF
and
default
torrents
I
would
say
this
is
similar
to
Java.
Api
is
because
it
it
comes
with
bunch
of
annotations,
which
you
can
use,
for.
You
know
fallbacks
barkhad,
circuit,
breakers
and
stuff
and
stuff
like
that.
So
so
what
it
means
is
this
upper
part
iterates
faster
than
the
lower
part,
which
is
the
same.
You
know
an
operating
system.
It
has
a
slower
release
cycles.
B
Then
then,
the
software
which
runs
this
is
not
quite
or
or
I'd
say,
a
framework
which
runs
on
top
of
the
operating
system.
Right
so,
and
so
this
is.
This
is
my
point
of
view
and
I
think
this
is
a
good
thing
that
they
are
separated
because
ya
take
loosely
coupling
right
like
micro-services.
The
release
cycle
of
this
one
is
different
to
this
one,
so
it's
good
things
to
keep
them
separated.
My
point
of
view
is
user
point
of
view.
B
If
I
were
a
vendor,
I
had
different
point
of
view
for
pod
from
the
user
perspective,
it's
fine.
So
this
is
the
second
slide
with
some
content
on
it.
So
what
productivity
means
for
me
first
and
and
what
it
means
is
not
like
you
know,
I
need
it
for
sure,
because
if
you
do
a
Jakarta
or
micro
profile
a
lot,
you
will
set
up
your
environment
once
and
you
are
happy
usually
but
I'm,
not
only
working.
You
know
for
me
alone,
somewhere
I,
also
working
on
my
four
different
clients.
B
Customers
with
different
teams
and
I,
have
to
set
up
the
stuff
on
different
machines,
help
developers
and
and
some
developers
don't
like
Jakarta
II,
you
know
or
micro
profile,
so
they
will
try
to
find
something
which
doesn't
work.
So
what
I
would
like
to
see
you
know
is
as
far
as
easy
setup
as
possible,
so
how
easy
it
can
be.
So,
let's
take
a
look
on
white
fly.
I
can
show
you
how
what
I
do
with
white
fly
so
I
have
a
script,
reinstall
white
fly
and
what
it
basically
does.
B
It
pulls
the
the
recent
white
flies
zip.
This
is
the
fool.
The
largest
white
fly
deletes.
The
old
one
and
I
have
I
think
it's
worldwide
fly
home,
wild
fly,
home
white
fly
home
yeah,
and
this
is
it
points
to
the
to
the
recent
white
for
installation.
I
have
the
same
for
payara,
but
I
don't
like
to
install
pirated
a,
and
the
reason
is
because
I'm
using
screen
sharing
for
now
because
of
crowdcast
and
it
collides
with
paya
report.
B
So
I
changed
the
configuration
a
bit
and
if
I
would
do
this,
but
I
have
exactly
this
the
same
script
script
for
payara
and
for
liberty
and
before
project
I
do
this,
and
on
my
on
my
with
my
clients
or
poor
or
external
projects,
I
tell
developers,
code,
download
and
unzip
and
you
are
set
actually
so
I
hope.
It's
done,
this
is
the
setup
and
I
think
it
is,
can
be
any
easier
actually.
So
the
next
thing
is
time
to
hello
world.
B
So,
if
I
would
you
know
require
to
have
I,
don't
know
if
you
we
had
to
know
recompile
the
UNIX
kernel
in
order
to
deploy
a
java
application
or
Jakarta
application
I
think
we
wouldn't
be
very
successful.
So
for
me,
it's
important
to
know
how
fast
you
see
results
on
the
screen,
how
fast
you
can
deploy
and
I'm
talking
just
about
local
deployment
and
what
I
usually
do,
and
this
is
independent
of
clouds
or
not.
So
we
started
with
white
fly
so
setup,
Jakarta
eh
project,
let's
say
CN
cloud
native
white
fly.
B
So
if
I
do
this,
what
happens
behind
the
scenes?
Ac
n4j
project
was
created
with
conventions,
so
I
can
rely
that
if
I
do
maven
clean
install,
it
will
create
a
war
with
the
same
name
as
the
folder.
So
I
have
the
same
for
a
war
here,
then
I
can
rely
that
in
palm
there
is
the
artefact.
Id
is
the
same
as
here
and
if
I
open
my
IDE
and
people,
ask
me
whether
I'm,
using
NetBeans
or
not
I
use
NetBeans
all
the
time
and
for
shorter
screencasts
I
use
Visual
Studio
code.
B
And
hopefully
the
project
appears
here
so
it
tries
to
synchronize-
and
this
is
my
Jakarta
a
project
and
what's
also
important
to
me,
you
know,
don't
make
me
think
so
we
have
artifact.
Id
is
C
and
C,
and
W
F
and
I
would
expect
to
see
the
same
in
the
IDE.
So
I
don't
like
in
order
to
put
your
name.
So
this
is
very
important
for
productivity.
Just
you
know,
it
should
always
look
the
same
and
you
should
just
focus
on
business
logic
and
not
fiddle
with
the
build
process.
B
So
having
that
sad,
as
are
all
my
projects
actually
comprising
both
and
actually
this
could
be
three
two
already
I'm,
just
using
three
three
zero.
For
probably
one
of
my
clans
was
had
not
the
recent
application
server,
but
was
the
payara
what
I
have
and
Wi-Fi
it
could
run
on
three
two
and
I,
using
mostly
JDK
11,
sometimes
JDK,
1,
8,
and
because
still
my
client,
some
are
running
on
one
8.
I
have
one
8,
but
I
have
I,
think
is
Java
version
should
be
11,
yeah
and
14
is
for
fun
right
now.
B
So
this
is
this
my
setup,
so
why
it's
important
for
product
TV
productivity,
a
short
poem,
because
what
I
do
from
time
to
time
are
also
code
reviews
and
if
the
poem
is
complicated,
it
is
really
hard
to
understand
why
it's
the
complicated
and
my
expectations
are.
You
know
something
I
don't
know
something
esoteric
is
going
on
in
the
project
and
what
I
try
to
find
out
first
is.
B
Ok
now
by
the
way,
if
you
have
any
questions,
feel
free
to
ask,
and
so
asking
chant
and
I
will
switch
to
chat
and
we
can
have
a
discussion.
Otherwise
it
will
be
just
a
boring.
You
know
life
streaming,
read-only
streaming,
okay,
so
we
have
it
back
to
the
slides
time
to
hello
world,
no
best-of-breed,
no
bike
shedding
what
I
mean
by
that
is
so
I
already
actually
said.
I
would
like
to
have
both
without
being
specific.
B
What
actually
it
means
so
I
still
don't
know
whether
I'm
going
to
use
rest,
easy
or
Jersey
or
hibernate
or
Eclipse
link
I,
actually
don't
care
and
if
I
had
to
decide
now
what
will
happen
in
my
world?
I
would
spend
at
least
two
or
three
meetings
to
discuss
why
we
need
you
know
hibernate
or
payara
or
whatever
I,
actually
don't
care.
It's
like
look,
let's
go
with
that
and
then
no
focus
on
on
the
run
times,
and
if
payara
doesn't
work,
you
can
go
to
white
fly
and
the
Wi-Fi
doesn't
go,
doesn't
work.
B
We
go
to
open,
Liberty
and
and
and
Tommy,
and
then,
if
everyone
is
not
happy,
then
try
Haledon
with
a
little
bit
migration.
Our
core
causes
a
little
bit
migrations.
There
were
lots
of
options,
but
we
don't
have
to
decide
upfront
and,
from
my
perspective,
it's
just
a
child.
So
we
don't
have
you
know
to
decide
everything
from
the
first
minute
we
can
just
let
it
go
with
Jakarta
with
the
api's
and
decide
later.
B
Now
we
have
that
so
to
project
I
just
have
hello
world,
so
it
doesn't
matter
for
now
for
today,
because
code
as
meta,
we
talked
about
productivity
first,
so
fast
builds,
and
this
is
really
true.
So
if
we
switch
now
and
just
build
it
again
by
the
way
you
don't
have
to
do,
the
clean
I
am
only
cleaning.
If
there
is
no
maven
committers
watching
if
they
are
watching,
they
always
say
you
don't
have
to
clean.
B
So
if
someone,
if
there
is-
you
know
maven
committee
and
the
audio
just
oversee
the
cleaning
is
not
needed.
So
what
we
have
so
it
built
in
1.7
seconds,
which
is
actually
very
competitive
and
it
is
actually
faster
than
a
quark
whose
compile
or
or
hella
don't
compile.
Why
that,
because
I
don't
have
to
pull
the
dependencies,
nothing
I
just
know
have
the
API.
So
it's
a
little
bit
faster
yeah
from
this
perspective.
B
Okay,
so
we
have
fast
built
key
to
success.
If
it
builds,
you
know
four
minutes.
No
one
would
like
to
you
have
theirs
or
fast
deployments,
so
what
it
means
is.
It
has
to
be
somehow
docker
compatible
so
and
fast
deployments.
First,
so,
let's
see
what
happens
so
I
think
have
we
started
whitefly
I,
don't
think
so
so
start
whitefly-
and
this
is
the
recent
white
fry
I,
think
I,
don't
know
two
three
weeks
old
and
now
it
started
not
very
slow.
Actually
pretty
fast
and
I
created
a
tool.
B
What
is
H
and
I
get
lots
of
questions
about
the
tool,
and
this
is
the
most
primitive
tool
you
can
properly
build,
and
why
I'm
mentioning
this
because
this,
what
it
does
it
just
watches
changes
in
my
file
system
on
every
change?
It
doesn't
even
clean
clean,
deploy
and
copies
the
war
and
I
can
do
this
because
I
know
the
convention
to
the
folders
from
the
application
service
and
I
have
over
Liberty
payara,
and
this
is
this
is
Tom.
This
is
Tommy.
This
is
open
Liberty.
B
This
is
a
white
fly
and
this
is
PI
era,
so
all
for
service
except
Quercus
or
Haledon,
because
I
that
they
don't
understand
worse
or
they
they
use
jars.
Okay,
so
we
have
it.
So
what
it
means
is
now
I
can
already
Allah
rely
on
the
availability,
so
this
8080
is
not
always
true.
For
instance,
open
Liberty
uses
different
port,
but
because
its
application
service,
most
of
the
application
service,
will
allow
you
to
use
resources
so
see
this
is
this
is
actually
standard.
B
This
is
semi,
standard,
I
would
say,
and
pink
is
just
so
we
see
that
it
works.
This
is
why
it's?
Why
and
matrix
does
not
work,
because
a
white
fly
is
interesting.
What
Wi-Fi
does
it
from
the
beginning
introduces
bark
hats,
so
what
it
means
the
matrix,
the
Prometheus
matrix
and
that
this
is
a
micro
profile,
come
from
a
different
port
which
is
actually
smart
and
and
paradise.
The
same
can
do
the
same.
B
You
can
expose
the
matrix
where,
on
the
admin
part,
why
is
this
smart,
because
the
admin
interface
or
the
admin
port
has
a
different
pot
number
sorry
as
different
a
thread
pool
then
the
8080
port,
and
why
it's?
Why
it's
important?
So
if
the
service
overloaded,
you
can
still
pull
the
monitoring
data.
Okay.
So
now
we
have
that
so
now
the
question
is
ok
cool,
so
local
deployment
seems
to
be
fast.
So
what
about
docker?
So,
let's
see
so
one
step
further
to
the
kubernetes
and
the
clouds
docker
so
GlassFish.
B
Today,
let's
go
with
white
fast
as
well.
White
fly
so
with
white
line.
So
what
I
will
have
to
do?
Is
mavin
clean,
install
or
just
forget
the
clean
and
then
dock
build
and
now
tea,
let's
say
Eclipse.
Hopefully,
I
can
do
this
and
Eclipse
and
then
CC&R
we
have
dot
so
and
now
it
builds
the
maven
and
it
very
quickly
build
the
images.
So
the
question
is
why
that
quickly?
B
So
if
I
go
to
docker
history,
what
I
will
find
is
the
following:
Eclipse
I
think
Sein
WF,
and
what
do
you
see
is
14
seconds
ago?
Would
I
did
6
hours
ago
so
today
morning,
I
just
updated.
The
whitefly
and
the
pyre,
a
pyre
I,
was
already
updated
by
the
Wi-Fi
image.
219
1.
So
it's
already
in
the
Docklands
repository.
Is
this
on
my
slide,
so
you
can
just
just
try
to
replicate
it
at
home
and
yeah
and
a
7
month
ago,
I
created
the
operating
system
is
probably
sent
to
us
7.
B
So
so
so
what
it
means
is,
or
two
weeks
ago,
was
JDK
11.
So
what
it
means
is
because
of
application
service-
and
this
is
also
what
Sebastian
mentioned
in
the
in
the
core
code
stock
and
this
layers
are
frozen.
Immutable
and
I
could
I
could
push
them
to
a
cloud
which
will
I
will
do
actually
in
a
second
and
then
and
then
pull
them
from
from
from
a
cloud
and
run
them.
So
let's
try
this
in
in
in
in
in
real
time.
B
So
what
I
can
actually
do
right
now
is
say:
docker
push,
Eclipse,
CN
or
yes,
see
and
RAF.
So
if
I
do
this,
it
pushes
it
says
denied,
which
is
a
really,
which
is
a
pity,
because
I
will
have
to
login
to
my
to
my
to
my
dacha
hub
in
the
morning.
It
worked
and
now
my
token
expired,
and
this
is
the
docker
hub.
But
what
you
already,
what
you
would
see
is
that
all
the
layers
are
already
existing
and
I
will
just
push
our
push
around
in
our
eight
mix.
So
this
is.
B
This
is
a
pity,
because
yeah
I
cannot
show
you
the
push
to
my
docker
hub
and
which
would
which
would
replicate
somehow
the
the
push
to
a
real,
no
docker
registry,
because
in
all
clouds
there
are
private
docker
registry.
We
push
the
docker
registry
or
you
can
push
to
add
to
your
own
private
docker
registry
and
then
put
it
from
the
cloud
okay.
B
A
B
B
What
you
see
my
latest
white
fly
was
not
pushed
at
all,
so
I
will
have
to
do
it
now
and
then
push
it
again.
So
this
is
the
initial
build.
So
what
you
see
is
the
white
try.
9
1
is
pushed
to
2
to
the
public
registry.
It
happens
1
and
at
once,
and
then
I
will
repeat
that.
So
you
will
see
you
know
what
happens
on
every
on
every
change.
Ok,
so
I'm
glad
that
it
works.
So
we
have
fast
deployments.
B
So
we
saw
locally
it's
actually
incredibly
fast,
and
this
is
that
fast
that
people
expect
me.
You
only
have
to
know
to
rewatch
some
of
the
older
ax
TV.
They
asked
me
know
what
you
are
doing.
Are
you
replacing
a
bytecode
with
their
what
no
I'm
complete
the
complete
application
and
because
I
have
no
dependencies?
It
is
very
fast
by
the
way,
because
most
of
the
Jakarta
and
micro
profile
applications
don't
have
any
external
dependencies.
They
are
very
easy
to
migrate
to
hell
it
on
or-or-or
Korkis.
B
B
Now
learned
new
environments
like
a
quark
ooze
or
held
on
because
it's
very,
very
similar
to
the
old
or
to
do
hold
to
Jakarta
and
to
micro
profiles,
actually
the
same
with
some
and,
for
instance,
in
Haledon
or
quark,
was
there
as
there
are
no
EJ
bees,
but
you
can,
you
know,
replace
the
servlet
with
request
code
and
transactional.
You
get
nearly
the
same
experience
nearly
so
we
have
it
Len
once
never
migrate
and
use
the
platform
first.
So
this
is
also
very
important
like
no
bike
sharing.
B
So
what
I
do
in
the
the
workshop
is
the
fault
of
workshops
and
projects
in
workshops.
No
one
cares,
but
in
the
projects
I
say:
okay,
we
focus
first
on.
You
know
what
we
have
on
the
JDK,
then
application
server,
and
if
something
is
really
lacking,
then
let's
talk
about
it
later
and-
and
there
are
lots
of
projects
where
we
just
went
to
production
without
any
external
dependency.
B
Okay,
cool
that
this
are
the
references
and
now
and
live
workshop
about
web
components,
and
this
is
my
very
last
slide,
which
is
really
nice,
because
then
I
can
show
you
some
more
stuff.
So,
okay,
so
we
pushed
the
entire
server
by
mistake
to
the
cloud
and-
and
so
we
have
white
fly
here
and
let's
perform
a
change
to
show
you
the
effect.
So
we
are
here
fear
message
and
let's
say:
Jakarta
ee9
with
micro
profile,
3,
of
course.
B
So
and
now
let's
build
this
again,
so
we
have
it
and
then
we
can
push
it
again
and
now
it's
become
interesting
and
this
is
a
real
push
to
a
remote
repository.
This
is
my
own,
a
hex
public,
docker
repository
and
I.
Don't
know-
and
this
already
happened
so
the
last
ten
seconds-
I
really
hate
that
it
happens,
but
it
actually
I,
don't
know
it.
Probably
it
compares
the
hashes
or
whatever,
but
actually
this
was
the
only
first
thing,
so
it
was
significantly
faster.
Then
the
very
first
first
push
it
took.
B
Probably
I,
don't
know
20
seconds
or
30
seconds.
So
what
it
basically
means
is
they're
smaller
they're
war,
the
faster
the
push
okay.
This
was
the
push
to
the
remote
docker
registry.
Now
the
question
is
you
know
what
about
Cuban?
It
is
so
what
I
have
here?
I
actually
wanted
to
use.
Cuba
natives,
but
I,
don't
use
bare
cabinet
communities
a
lot
so
either
I
use.
This
is
a
OpenShift
or
I
used
to
know
Cuban.
B
B
It
also
takes
too
long.
You
have
to
wait
until
it
deletes
the
tests
yeah.
Now
it's
gone,
and
hopefully
the
test
is
deleted
so,
and
so
what
I
will
plan
to
do
is
to
deploy
to
Cuban?
It
is
so
I
won't
use,
as
the
open
ship
is
basically
kubernetes
with
some
more
edit
features
and
the
first
time
I
won't
use
this.
This
edit
features
so
usually
I
use,
s2i
images
or
whatever,
and
this
time
I
would
misuse
open
shift
as
be
a
Cuban.
It
is
platform
and
just
pull
from
the
public
cloud
now.
B
What
here
happens,
is
it
just
copies
and
replaces
one
name
so
and
should
open
visual
studio
code.
So
what
happened?
It
created
a
air
out
a
service
and
a
deployment
config,
and
so
what
I
end
config
map
is
absolutely
optional.
The
config
map,
but
I,
wanted
to
show
a
little
bit
more
and
the
image
stream
is
a
requirement
more
or
less
for
for
openshift,
because
this
abstraction
from
dhaka
so
on
on
cuban
it
is
you,
don't
have
to
use
the
the
image
stream,
but
what
I
don't
have
a
build?
B
Config
build
configuration,
so
I
just
don't
use
it
now.
But
if
you
are
interesting,
just
search
for
s2,
I,
Adam
bean
and
you
will
find
deployments
on
my
youtube
channel
with
in
the
OpenShift
way
so
so
I
have
it.
This
is
my
Amal
and
what
basically
happened
is
I,
just
replace
you
know,
occurrences
of
service
name
with
C
and
W
F.
So
if
you
replace
this
with
Duke,
you
can
deploy
a
Duke
service
and
this
basically
it
so
don't
make
me
think,
and
a
convention
of
a
configuration
with
some
sensible
defaults.
B
Now
it
pulls
the
dependency
and
I
actually
forgot
what
this
is
payara
or
Dwight
fly,
and
it
should
start
it
already
started.
So
this
was
a
very
fast
deployment.
Why?
Because
the
base
layer
was
already
pushed
to
open
shift.
So
what
happened
right
now?
It
just
pulled
from
the
public
docker
hub,
the
war
and
it
started.
I
think
I
still
get
some
trouble
and
you
probably
are
yeah
okay.
This
is
a
easy,
easy
easy
to
see.
Why
that
because
the
whitefly
exposes
the
readiness
and
liveness
probe
on
different
port.
B
B
So
I'll
see
delete
all
config
map
yeah,
let's
go
with
that.
L
app
and
see-
and
re
f
takes
too
long.
So
the
problems
the
clouds
is,
the
deletion
removal,
so
diplomat,
is
fast.
Removal
is
slow,
so
this
is
what
we
have
to
fix
in
the
next
iteration
of
Q
Benitez.
You
should
delete
the
cloud
faster
so
now
what
I
also
have
here,
the
first
time
config
Maps-
and
this
is
actually
easy.
So
what
happens
with
the
config
map
is
the
following.
B
B
B
Deployment
and
we
have
the
deployment
pot
and
deployment
created
the
actual
pot
and
in
the
logs
you
see
it
started
and
in
the
overview
you
should
see
it
is
deployed.
So
this
was
a
full
deployment
from
more
or
less
scratch
with
my
prepared
template,
which
I
use
all
the
time
actually-
and
this
is
a
full-blown
the
largest
available
white
fly
on
the
planet.
B
So
let's
see
what
it
works
right.
So
we
have
to
go
to
routes,
and
this
is
the
sea.
Nwf
route
show
details,
I
don't
have
the
certificate
for
that,
so
we
have
to
white
fly
and
then
we
had
what
we
had
was
sea,
nwf
resources
and
pink
so
and
Georgia
Katya
a
a9
with
micro
profile,
3
2.
So
now
what
I
said
is
what
we
could
do.
B
We
could
actually
inject
a
config,
so
whatever
about
updates
config
property
and
name
equals
M
air
hex
message,
streaks
string,
air
hex
just
call
this
this
way,
plus
yeah
hex
message.
So
my
machine
is
already
a
little
bit
hot.
So
there
is
a
openshift
mini
shift,
running
docker,
running
standalone
server
running
and
the
crowd
crowd
cast
is
also
running.
So
a
lots
of
stuff
running
here,
and
that
means
and
this
video
code
as
well.
So
now
we
did,
it
now
may
even
clean
install.
Do
it
again,
so
we
build
rebuilds.
B
The
entire
project
I
have
to
push
it
to
the
cloud
talk
push
and
by
the
way.
What
you
saw
unintentionally
in
my
intercession
is
the
what
happens
on
every
three
months.
If
there
is
no
new
whitefly
version
available,
so
I
pushed
just
the
new
version
to
the
cloud
it
was
220
max,
I
don't
have
to
reapportion
or
the
entire
centralized
image,
because
it
didn't
change
just
E.
So
it
is
so
now
I
have
to
push
it,
and
now
we
should
see
that
it
actually
exactly.
This
is
the
war
changed.
B
Everything
remains
the
same
and
then
because
we
have
open
shifts,
I
have
to
do
something,
OC
import
image,
and
this
is
C
n,
WF
so
and
just
the
import,
and
we
should
see
that
created
two
hours
ago,
image
created
40
seconds
ago.
This
should
trigger
actually
redeployment.
So
if
you
go
here
to
overview,
we
see
that
it
actually
redeploys,
because
there
is
an
image
change.
So
what
I
just
did
is
say,
look
there
is
new
image
and
it
redeployed
the
image
and
what
it
actually
means
is.
B
A
B
What
I,
what
I
did
today
or
yesterday
I
created
the
new
images
for
payara
and
whitefly
and
I
created
another
application
just
to
test
to
test
whether
it
actually
works
for
the
presentation
was
the
newer
images
and
JDK
11,
and
it
is
for
something
else,
and
it's:
okay,
now
I
show
you
the
newer
stuff
and
to
test
it.
I
pushed
the
application.
What
happens
is
openshift
comes
with
a
docker
registry
inside
and
it
already
contains
the
base
layers,
but
this
is
how
clouds
are
working.
B
So
if
this
were,
let's
say,
Microsoft,
Asia
cloud
or
AWS
cloud
this
exactly
the
same
would
happen.
You
would
push
to
the
private
registry
directly
to
a
things
called
Al
container
registry
or
ECR,
or
something
from
from
from
Amazon
and
and
then
once
and
then
just
war.
So
there
would
be
no
difference,
but
this
is
true
what
I
did
just
for
test
I,
initially
initially
created
the
the
other
applications
with
the
same
immutable
layers.
B
If
this
were
a
uber
jar
or
fetch
our
it
would
matter,
I
would
have
to
redeploy
always
the
entire
infrastructure,
but
the
cool
story
is
with
all
application,
service
and
Haledon
and
Quercus
as
well.
They
clearly
separate
the
business
logic
from
the
infrastructure.
What
is
very,
very
good
for
productivity.
What
I
forgot
to
mention
in
my
slides,
as
well
as
what
you
shouldn't
forget,
is
security
scanning,
so
many
companies
are
very
obsessive
with
security
scanning
of
deployables,
and
in
this
particular
case
you
only
have
to
rescan
the
the
worse
or
des
jahres.
B
B
B
What
you
also
saw
is
something
different
and
forgot
to
mention
is
the
following:
were
the
aliveness
and
readiness
probes,
so
they
already
shipping
with
actually
all
service
out
of
the
box,
and
if
we
how
to
show
you
this,
the
easiest
is,
if
I
change
to
my
pot,
which
is
kind
of
a
docker
container.
This
is
my
part
and
I
go
to
terminal
and
say
curl,
because
this
is
not
export.
It's
not
exposed
where
they're
out
I'm
running
now,
with
with
white
fly,
so
localhost
8080
slash
health.
B
So,
as
you
can
see,
this
is
the
health
check
and
I
can
also
go
and
say:
life
is
life,
so
I
actually
reused.
You
already
existing.
You
know
default
health
check
because
because
a
micro
profile
extra
can
rely
on
the
existence
of
the
same
in
no
Horry
so
regardless
which
applications
if
I'm
using
I
can
always
use
the
same
template.
B
So,
for
instance,
if
I
would
like
to
use,
though
there
was
about
fly,
white
fly
and
now
let's
talk
about
fishes
like
payara,
so
if
I
will
do
something
else
with
payara,
it's
the
recent
one
and
hopefully
I
pushed
the
payara
and
just
do
exactly
the
same
now
so
now
I'm
building
the
entire
application
against
payara.
So
everything
is
the
same.
Actually,
the
last
part
is
different.
Then
I,
push
and
yeah
I
can
push
okay
push
to
the
cloud,
and
you
see
there
are
more
layers
and
hopefully
payara
is
pushed
yeah.
B
It
was
pushed
as
you
can
see,
they're
still
7k,
so
it's
roughly
the
same
and
why?
Because
I
tested
payara
the
recent
one
5.2
or
one
pushed
to
the
cloud,
and
now
it
only
recognizes
the
changes
and
it
just
you
know,
push
the
world
now.
The
question
is
how
the
history
looks
like,
and
this
was
not
eclipsed
rather
than
a
hex
hex
yeah
hex.
B
B
Deleting
takes
too
long
delete
everything
and
then
reinstall
everything.
Why?
Because
I
would
have
to
change
the
part
to
test
a
problem.
Payara
doesn't
know
about
this,
so
8080
so
should
be
better
and
now
after
it's
deleted,
which
takes
forever
creation
is
a
lot
faster
and
diplomat
is
faster.
So
what
it
means
you
know
how
to
evaluate
a
evaluate
cloud
native
runtimes
creation
should
be
faster
than
deletion.
Then
it's
perfect,
so
OC
create
this
one.
So
it's
done,
and
now
we
should
see
what
happens
here.
B
So
we
have
C
n
W
the
first
deployment,
because
I
think
this
is
an
error
because
it
remembers
the
old
stuff,
but
let's
see
applications,
pots,
running
terminal
locks
and
PI
RS
up
and
running.
So
this
is
the
same
image
on
PI
R
right
now,
I
skip
over
liberty
and
told
me
it
will
be
exactly
the
same.
So
now
I
deployed
to
Cuba
natives
as
openshift
without
any
OpenShift
magic,
and
so
now,
let's
see
yeah.
It
is
already
blue.
It
means
health
checks
were
checked
and
go
to
the
route,
applications,
routes
and
cnw.
B
This
is
now
payara
and
see.
N
WF
actually
should
be
P,
but
so
this
is
now
from
payara
with
the
same
config
map.
So
this
is
the
same
output
and
because
it's
payara,
it
just
uses
8080
per
default.
I
think
health
should
be
available
here,
which
is-
and
this
is
the
default
life
and
which
is
the
a
little
bit
different.
What
I
just
check
it
should
be.
No
200,
not
I,
don't
check
inside
here
and
and
metrics
should
be
also
exposed
here,
metrics
which
they
are
and
open
api
is
also
exposed.
B
B
You
know
somewhere
else
and
they
were
project
was
funky
frameworks
which
most
of
the
friend
has
already
died
or
or
changed
completely
so
that
they
had
to
know
to
refactor
or
migrate
the
project
back
and
forth,
but
with
Java
EE
and
now
Jakarta
you
can
rely
on
unstable
API
and
even
in
the
case
you
know
in
Vinnie
case.
Hopefully
it
will
happen.
You
know
they're
renaming
to
Jakarta
a
namespace.
So
then
it's
more
or
less
no
I
would
say
yeah
kind
of
migration,
but
it's
actually
just
no
search
and
replace.
There's
not.
B
You
know
a
huge,
a
huge
problem,
so
I
think
I'm
open
for
questions,
so
I
could
note,
implement
some
trivial
stuff,
but
I
think
the
other
speakers
did
a
great
job.
So
I
don't
like
in
order
to
show
you
how
great
a
micro
profile
is,
because
it
is
already
great
and
so
I
would
rather
open
for
questions.
So
now,
I.
B
A
B
So
for
JVM
it
took
quite
long
to
get
GC
right
and
very
fast,
probably
the
same
for
the
cloud
providers
was
deleting
stuff,
yeah
and
and
and
having
said
that,
this
is
OpenShift.
3:6
was
way
faster
with
deleting
than
3:9,
so
it
seems
like
no,
the
deletion
takes
long
and
longer
so
for
it
to
be
yes,
yeah
Johnny,
Aires,
I,
know
John
from
from
from
somewhere
from
Java
ones.
There's
our
old
friend,
actually
so
no
questions.
So
this
is.
B
But
our
Jakarta,
a
war,
is
over
two
hundred
Meg's
any
tips
for
slimming
it
down
yeah
sure.
First,
what
I
do
is
I
scan
the
war
as
search
for
everything
for
anything
which
is
not
provided
or
test.
This
is
usually
runtime
and
all
the
runtime
dependencies
equity,
as
if
this
is
a
big
project
and
200
max
is
really
big.
So
then
I
would
create
an
excel
sheet.
If
this
is
a
no
for
management
and
then
ask
myself,
you
know
what
is
the
absolute,
no
edit
value
of
this
dependency?
B
What
problem
does
it
solve
and
usually
you
know,
Jakarta
Commons,
link
and
Jakarta
is
used
mostly
for
nicer
to
string
or
left
padding
or
string.empty
or
something.
So
this
is
absolutely
optional.
So
I
would
just
remove
these,
but
Apache
pooi
is
hard
to
remove,
but
what
you
can
do
with
most
service,
it's
a
little
bit
harder
was
whitefly
actually
easy
with
headed
on
Quercus
and
also
easy
was
payara.
You
can
ship,
you
know
the
DD
base.
B
Use
micro
profile,
reactive
messaging
in
production
and
what
we
with
transforming
kafka
messages
to
WebSocket
messages.
It
works
great,
so
a
general
question
whether
it
is
major
enough.
Nothing
is
made
sure
enough.
We
found
problem
problems
everywhere
in
one
project,
even
on
IBM
old
host,
so
it
was
actually
impossible
to
find
an
arrow,
but
we
found
an
arrow,
so
you
won't
have
to
test,
and
if
these
test
passes,
I
will
just
use
it.
B
Don't
see
the
questions
I
have
to
scroll
are
scrolling
is
and
go
from
from
front
out.
Is
there
a
real
practical
Advent
advantage
having
multiple
implementation
in
comparison
to
eye-to-eye?
What
what
happened
is
a
real
practical
advantage
having
multiple
implementation
in
comparison
to
a
season
themed
framework?
Please
answer
the
suitable
projects
and
project
managers.
I.
Think
of
course,
because
if
I
enter
a
project,
I
don't
have,
you
know
to
say,
use
XYZ
or
something
else,
I
say
just
don't:
negotiate
with
redhead,
payara,
IBM
or
whoever
you
know,
I,
don't
care.
B
B
If
you
have
one
how
you
call
it
themed
framework,
you
would
have
to
negotiate
with
the
theme
framework
and,
having
said
that,
if
the
project
is
really
critical,
I
don't
think
you
will
run
without
commercial
support,
so
we
probably
could,
but
either
you
will
have
to
dedicate
you
know
persons
to
to
manage
the
dependencies
in
the
dddd,
runtime
and
or
easier
to
buy
support
and
have
no
security
patches,
for
instance,
very
good.
Thus
your
many
shift
installation
already
have
the
application
server.
You
are
using
downloaded
and
not
the
application
server.
B
The
base,
docker
images
were
automatically
downloaded
because
I
tested
today
morning
whether
it
will
run
on
JDK
11,
and
it
did
so
this
this
happened.
So
the
next
was
yet
we
did
it
and
jax-ur
as
multi-port
multi-part
from
data
upload
is
not
potable
between
Jessie
and
there
is.
This
is
true.
What
approach
do
you
use
in
your
project
to
make
it
potable
I
just
use
either
stz
or
our
Jersey
approach
and
I?
B
Commend
it
or
is
really
easy
to
find,
because
I
will
search
for
imports
and
you
know
to
port
it
from
Jersey
to
rest,
easy
is
not
a
big
deal,
so
it's
very,
very
easy.
So
I
don't
think
is
now
the
question
is:
what
happens
if
you
have
to
port?
How
long
will
it
take
and
I
would
say
not
that
long,
and
this
is
also
a
isolated
case,
and
if
your
project
is
really
really
important
and
you
have
lots
of
uploads
what
you
could
do.