►
From YouTube: Continuous Deployment - The Last Frontier for Continuous Delivery - Fernando Ike , Zup Innovation
Description
Continuous Deployment - The Last Frontier for Continuous Delivery - Fernando Ike , Zup Innovation
What is the strategy for implementing new versions of a new version application? Continuous Deployment not only automates the deploy, it’s beyond when aligned for the product and services lifecycle. Continuous Deployment as part of Continuous Delivery becomes valuable to test and experiment in production that can’t test in not production environments. Choosing what’s the better strategy for each type organization, budget or departments means to know what’s the main deploy strategies, when possible to use any of them, pros and cons. It’s a talk to dive for the most deploy strategies
A
Hello,
I'm
a
friend
like
I'm
a
developer
advocate
at
the
zoo
today.
I
would
like
to
talk
about
continuous
deployment
as
something
a
little
bit
different
than
what
you
usually
see
around
the
internet,
simply
as
an
automation,
deployment
or
the
next
step
that
continues
deliver
some
organizations,
don't
realize
how
the
important
plugin
strategy
is.
They
only
recall
it
when
there
are
some
incidents
after
deploying
a
new
organizational
version
production
so
for
any
organization
that
wants
to
innovate
in
learning
fast,
it's
very
important
to
have
a
deployment
strategy.
A
Just
look
at
the
state
of
devops
report.
They
mention
a
metro
related
directly
to
deployment.
It's
a
ginger,
flavor,
really
failure
rate.
Sorry,
oh,
let
me
introduce
myself
again
but
tell
me
about
my
experience.
How
long
time
ago
I
was
seasoned
when
I
had
to
measure
services
like.
Oh
s,
sorry
send
me
apache
and
symbol
servers.
A
A
So
most
of
the
texts,
podcast
episodes
and
presentations
states
that
continuous
integration
continues,
or
even
continues
upon
as
equation
steps.
These
a
common
sense
about
that.
What
do
you
think
do
you
agree
with?
What's
this
picture
represent?
If
you
do
that,
it's
not
really
problem
it.
It's
if
your
thoughts
are
different.
Please
comment
in
the
chat,
your
social
media
and
reach
out
to
me,
obviously
yeah.
I
would
like
to
know:
what's
in
the
configure
integration
continuously
continuous
deployment
means
to
you
and
how
your
pipeline
happens.
A
So,
let's
back
to
concept
com
definition,
what's
the
intake
continuous
integration
for
the
reference,
I
use
a
margin
follow
site
definition.
The
continuous
integration
is
a
software
development
practice
where
the
member
of
our
team
integrated
their
worker
frequently.
Usually
each
person
integrates
at
a
list
daily,
leading
to
multiple
integrations
per
day.
Each
integration
is
verified
by
an
automation,
build
including
tests
to
detect
integration
errors
as
quick
as
possible.
A
Well,
let
me
use
the
definition
I
found
in
the
jazz.
Humble
site
is
a
co-author,
often
on
some
awesome
book
with
the
developer
called
continuous
delivery,
reliable
software
user
to
build
testing
deployment.
Automation,
control
delivers
is
the
ability
to
get
changes
of
other
types,
including
new
features,
configuration
changes,
bug
fixing
experiments
into
production,
earning
the
handoff
users
safely
quickly
in
a
sustainable
way.
A
A
This
is
why
continuously
for
many
people,
this
is
seen
as
the
devops
heart.
I
agree
with
that.
After
review
of
continuous
integration,
continuous
delivers
concepts
through
the
words
of
the
premiere
people
in
19
industry.
Let's
see
what's
the
chinos
department
is,
although
continuously
by
moscone
by
timothy.
I
prefer
to
use
it
as
jazz
home
definition.
A
He
wrote
a
text
continues
to
even
control
the
verses
versus
kojima's
deployment
in
the
text.
He
said,
continuous
definition
that
the
practice
of
the
release
every
good
view
to
users,
for
example,
a
continuous
delivery
tool
or
service
that
does
an
integration
for
comic
book
questions
in
or
measuring
a
question
is
good,
and
so
all
the
tests
unit
tests
integration
tests
any
other
test
that
is
important
for
the
company.
It's
ready
to
go
to
deploy
in
release.
A
This
picture
represents
my
vision.
What
the
hierarchy
of
continuous
integration
controls
the
planet
and
continues
the
planet
is
continuous
integration
continues.
The
planet
is
a
part
of
the
continuous
delivery
because
of
continuous
delivery
gives
us
a
power
for
business
to
decide
which
features
it
there
really
be
enabled
to
users
of
customers.
A
Almost
everyone
knows
some
of
the
families.
Strategies
like
a
blue
green
cannot
release,
but
there
are
many
other
strategies.
Our
strategies
mentioned
here
considered
to
use
a
container
concentration
like
as
covenants
to
explain.
Most
of
them
most
of
the
deployment
strategies
are
possible,
applying
virtual
machines
or
better
metal
servers.
They
are
a
little
bit
more
complex
to
permit
but
completely
completely
doable.
A
Recreating
place
it's
a
basic
for
in
common
for
every
organization
teams
or
productive
when
they
start
from
zero
nowadays
in
the
high
performance
teams.
It's
things
old-fashioned.
Unfortunately,
low
performance
in
chinese
organizations
is
to
use
its
main
strategy
to
their
deploys
on
the
internet
documentation.
We
will
find
a
reference
to
recreate
replays.
A
They
are
basically
the
same
thing.
The
application
version
is
changing
in
the
same
server:
either
virtual
machines,
on-premises
servers,
I
script
or
configuration
manager
tool
like
a
instable
chef.
There
have
form
etc.
Stop
application
version,
one
copy
version,
two
starter
version:
two.
I
didn't
mention
details
about
the
process,
but
you
can
figure
out
that
more
steps
are
necessary
to
update
so,
let's
say
pros
and
cons
pros.
A
Cons
operation
can
be
complex
for
holdbacks.
These
updated
versions
could
be
painful
when
the
changes
in
the
databases
are
in
the
application.
Architecture
are
necessary.
Applications
are
available
during
version
transition.
It
means
deploy
after
off,
see
hours
like
a
night
or
freedom,
even
it's
a
nightmare.
A
These
are
a
reason,
though,
for
the
human
legend
for
of
never
deployment
of
friday
evenings.
Because
of
the
case
of
bugs
and
the
failures
we,
you
will
end
up
working
as
tomorrow
or
no
no
tomorrow
into
monday
off
sloping
off
open
hours.
Sorry,
like
spelling
fault
holding
the
plumbing,
is
called
brampton
brent.2,
basically
a
new
version
that
will
deploy
in
substitute
substitute
the
old
version
by
number
of
instances
per
time
or
percentage
here.
Deployment
holding
the
plumbing
example
instances
are
updated,
perth
update,
2
per
time
version
2
is
deployed
into
instance,
container
construction
checks.
A
A
Let's
see
pros
and
cons
for
your
rolling
deployment
brush.
Is
there
no
time
for
users?
It's
good
prevent
your
release.
If
your
own
instances
need
with
the
new
version,
aston
are
not
healthy.
For
example,
if
you
have
ten
instances
for
an
application,
if
you
want
off,
then
it's
a
health
container
extraction
will
return
to
last
very
healthy
version.
A
Cones
on
the
hand,
it
could
be
hard
to
knew
the
kind
of
version
that
has
failure
privilege
when
these
strategies
configured
to
create
small
bets
per
instance
or
presenter
backwards
compatibility,
firm
databases.
Apis
is
very
important
because
it
is
a
strategy,
keeping
multiple
versions
running
simultaneously.
A
Oh
blue,
green,
it's
a
simple
component
to
duplicate
your
environment
with
the
new
application
version.
Router
traffic
to
new
environment
after
the
old
new
environment
is
deleted.
It's
a
simple
and
cool
here
are
example,
with
quattro
instances
with
the
version.
One
updating
this
version
is
the
to
create
the
same
number
of
the
instance
through
version
two.
A
A
A
You
can
separate
the
environment
to
upgrade
the
hardware
or
type
of
of
vms
like
in
before
two
and
five,
if
you
use
aws
cons,
duplicate
computer
resourcing
cost,
if
your
environment
is
an
infrastruct
infrastructure
as
a
in
the
bare
metal
scenario,
it's
ricardo.
Oh
sorry,
and
yes,
my
apologies.
I've
committed
I'm
mistaking
the
phrase
in
the
parameters
and
how
it's
required
half
of
the
service
in
the
idle.
It's
a
little
hard,
a
little
hard
to
explain
to
auditors
or
executives
that
your
servers
are
not
running
an
application.
A
This
is
just
waiting
to
be
using
the
next
deploy.
I
know,
I'm
you
using
extremely
simple,
but
you
got
it
right,
be
careful
with
dependence
services
like
at
that
databases.
They
need
to
prepare
to
support
multiplication,
multiple
application
versions.
I
need
to
drink
some
water.
I
made
up
a
lot
of
mistakes
in
last,
two
slides.
A
Okay,
go
ahead
kind
of
releasing
inspire
and
then
the
miners
in
the
kinder
bird
detail
they
take.
Oh,
they
took
like
a
light,
a
live
counter
in
the
mine
to
identify
toxic
gases.
A
After
time
all
of
the
times
the
alarm
owns
some
owns
once
the
canary
was
already
dead.
Fertility
spreads
is
no
longer
used
for
the
ig
kind
of
releases,
the
way
to
off
version
to
become
available
to
be
tested
in
production
for
the
set
real
users
before
to
deploy
for
our
users,
an
example
in
the
color
release,
load,
balance,
splitted
part
of
traffic
to
instant
version
2
and
keeping
orders
users
in
the
version
one.
A
A
Let's
see
pros
and
cons
for
calendar
plus
your
long
time,
free
users,
testing
for
real
users,
multiple
versions
running
simultaneously
and
cons
compared
to
the
plumbing
strategies
we
have
here
already
talked
about
it.
Holdbacks
is
lower
if
an
any
issue
is
founded
application.
Indeed,
dependence
needs
to
good,
monitor,
alaka
aka
observability
during
the
twice
quickest
possible
english
just
for
a
recorder.
It's
a
production
environment
with
the
real
users.
Even
though
there
are
it's
a
little
amount
of
them
complementary,
her
budget,
it's
a
really
important
for
to
keep
a
good
relationship
among
teams.
A
He
beat
tests,
eb
testing.
Even
tensions
are
very
popular
for
marketing
and
production,
productive
professionals,
it's
more
common
to
use
a
segmented
user's
base,
sorry
to
business
policies,
one
of
the
interesting
of
marketing
or
product
person,
for
example,
company
wants
to
know
what
is
the
best
quality
for
payment,
but
red
or
green.
A
Really,
a
rule
is
created
for
routing
users,
with
the
cuckoo
valley,
red
for
a
version,
2.
lottery,
balance
of
perseverance
value
in
their
how
to
the
users,
with
the
hedge
evaluated
to
version
2
and
keeping
the
hester
use
of
using
the
version
1
instances,
these
tests
file
that
they
read
but
didn't
increase
the
heaven
so
just
as
every
ruler
router
to
all
traffic
how
to
version
one
instances.
A
A
A
Let's
see,
needs
how
this
works.
Like
a
blue,
green,
it's
a
creating
digical
environments
to
deploy
version
2
in
the
traffic
memory
from
version
1
to
version
2..
Thus
you
can
observe
the
behavior
of
the
version
to
with
with
the
real
user
traffic
if
something
had
not
gone.
Well,
just
drop
a
router
to
version
2..
If
everything
is
good,
just
disable
version
1
route
in
deleter
version,
1
instances,
oh
shadow
pros
and
cons,
pros,
reduce
the
risk
of
the
production
environment
and
can
use
in
prep
production
environment
should
test.
A
If
the
application
new
version
is
good
to
deploy
in
production
testing.
A
new
version
with
the
here
user
traffic
is
good
to
know
with
the
a
new
feature
came
back
to
the
performance
of
break
something
in
the
application
comes
if
the
application
has
integration
with
the
third
party
apis
that
changing
this
data,
such
as
customer
financial
history.
Maybe
it's
good
to
use
a
hover
file
as
ipa
simulator.
A
Oh
circles,
my
first
impression
of
the
circles
that
was
very
interesting
circles.
For
me:
it's
a
mix
of
the
a
b
testing
in
calorie
it's
impossible
to
to
segment
your
traffic
to
for
attaches
like
eb
or
splitter
traffic
base,
application
versions,
for
example,
version
for
europe,
another
for
asia
and
north
of
the
rest
of
the
world.
A
A
A
Her
budget
must
be
a
part
of
the
your
controls,
the
planning
process.
It's
important
to
know
the
limit
of
the
experiments
on
production.
If
the
organization
don't
consuming
every
budget,
it
could
be
mean
that
company
doesn't
have
experiments
and
don't
does
innovate,
make
sense
right
so
pros
and
cons
circles
pros
journal
time
for
users,
real
users
for
testing
capable
running
multiple
versions
simultaneously
quickly:
roll
back
multi
tenants
to
support
different
type
of
applications
and
environments.
A
A
So
in
finally,
about
best
practices
doesn't
matter
with
which
control
department
strategy
you
need
to
use
it
to
here
are
the
best
practices
for
all
of
them.
Backward
compatibility
is
very
important.
Continuous
integration
continues
delivering
continuous
deployment
and
are
very
important
for
high
performance
organizations.
A
One
should
always
have
a
back
a
whole
back
plane
if
something
bad
happens
in
the
important
in
other
play.
Observability
is
a
new
buzzword.
I
know,
but
it's
good
to
know
if
something
is
going
to
be
wrong
in
the
point
or
when
running
multiple
multiple
versions
of
application
in
production-
and
this
is
references,
think
google
cloud,
the
container
solution
for
writing-
amazing
docs
about
deployment,
others
references
are
here
too,
and
and
thanks
so
much
cdcom
and
you
for
watching
this
presentation.
My
presentation,
my
failure
to
spawn
correctly
but
again,
thank
you.