►
Description
We are living in an era where enterprises are embracing the DevOps model of software delivery. Technologies such as Docker, Kubernetes and Jenkins are becoming de-facto standards in the industry for building, deploying and managing modern, cloud-native, microservices-based applications. Join this talk as we introduce you to a fresh take on achieving continuous integration and continuous delivery for modern, cloud native applications using the power of Jenkins, Docker and Kubernetes.
This webinar covers:
- Setup your own Jenkins-based continuous integration/continuous delivery system
- Develop, import and deploy new microservices with automated continuous integration/continuous delivery
- Use automated provisioning in testing, staging, ephemeral and production environments via Helm charts and GitOps
Join us for KubeCon + CloudNativeCon in Barcelona May 20 - 23, Shanghai June 24 - 26, and San Diego November 18 - 21! Learn more at https://kubecon.io. The conference features presentations from developers and end users of Kubernetes, Prometheus, Envoy and all of the other CNCF-hosted projects.
Join us for KubeCon + CloudNativeCon in San Diego November 18 - 21. Learn more at https://bit.ly/2XTN3ho. The conference features presentations from developers and end users of Kubernetes, Prometheus, Envoy and all of the other CNCF-hosted projects.
A
Good
morning,
hi
everybody
and
welcome.
Welcome
today
we're
going
to
be
had
today.
Cn
CF
webinar
is
about
continuous
integration
and
delivery
with
kubernetes
I
know
it's
something
that
I'm
deeply
passionate
about,
and
I'm
really
excited
for
this
session
and
our
guest
speaker.
My
name
is
Lachlan
Evensen
I'm,
a
principal
program
manager
at
Microsoft
and
a
CN
CF
ambassador
and
I'm
happy
to
be
here
today
hosting
this
wonderful
content.
I'd
like
to
thank
you
all
for
joining
and
being
here
today,
it's
great
to
have
such
fantastic
numbers.
A
I
know
it's
an
interesting
topic
in
the
open
source,
ecosystem
and
I
would
love
to
welcome
and
extend
a
warm
welcome
to
our
guest.
Today,
our
guest
speaker,
his
name
is
James
Strachan.
He
is
a
senior
architect
at
cloudBees
and
he
will
be
presenting
his
expertise
on
CI,
continuous
integration
and
continuous
delivery
on
kubernetes.
Just
a
little
background
about
James
James
is
responsible,
he's
been
in
the
open-source
community
as
a
developer
for
20
years.
Creating
many
great
projects,
including
groovy
apache,
kamal
dom4j
and
jackson.
A
He's
been
working
most
recently
with
Jenkins
on
kubernetes
and
was
previously
working
at
on
fabric
aid
at
Red
Hat.
So
today
he's
going
to
be
sharing
our
CI
and
CD
with
kubernetes
and
without
further
ado,
I'd
love
to
hand
it
over
to
James.
For
this
exciting
content.
Thank
you,
James.
You
take
it
away.
Actually,
before
I
before
I
forget
I
have
two
housekeeping
things:
I
got
a
little
overexcited,
so
it's
already
good.
A
So
during
the
webinar,
you
are
all
on
mute.
All
the
attendees
are
on
mute.
There
is
a
chat
window
down
the
bottom,
but
we
would
ask
that
all
your
questions
be
held
to
the
end
and
as
you
want
to
ask
them,
there
is
a
QA
button
down
on
the
bottom
panel.
You
go
ahead
and
click
that
QA
but
Q&A
button,
and
you
can
drop
your
questions
in
there.
A
There
will
be
time
allocated
at
the
end
for
answering
each
and
those
questions
as
we
get
through
to
there,
so
this
session
will
be
about
40
to
45
minutes.
Think
questions
at
the
end
and
I
will
be
facilitating
the
questions
in
the
Q&A
section.
So
we
ask
that
you
ask
them
there
and
also
this
session
is
being
recorded
and
there
will
be
a
link
set
out
after
the
presentation.
So
sorry
about
that
James
I
will
now
hand
it.
A
B
Thank
you
very
much.
Thank
you.
So
hello,
everyone,
let's
get
started
so
before
I
talk
about
Jenkins,
X
I,
just
want
to
briefly
talk
about
the
software
ecosystem
right
now.
So
a
lot
has
happened
in
the
last
5
years.
That's
really
changed
what
it
even
means
to
deploy,
develop,
deploy
test
and
orchestrates
oft
we're.
So
cuban
''tis
is
an
amazing
thing.
I
love
it.
I'm
hugely
passionate
about
communities.
It's
one
of
the
most
amazing
open-source
projects
are
ever
used
in
my
life.
B
It's
amazing
stuff,
so
cuban
entities
in
2018
cuban
eighties
is
basically
one
the
continual
distribution
game
in
2018,
I.
Think
I.
Think
today,
eks
on
Amazon,
these
announced
aks
is
meant
to
be
GA
like
this
week.
The
next
week
gke
has
been
around
for
a
while.
So
all
the
big
free
cloud
providers,
Amazon,
Google
and
Microsoft
all-
have
managed
kubernetes
as
a
service
cuban.
It
is
amazing
technology,
but
friends,
don't
let
friends
install
communities
right,
it's
kind
of
tricky
to
install
and
manage.
So
what
you
really
want
to
do
is
outsource
that
to
someone
else.
B
B
You
can
do
auto
scaling
in
it,
and
the
cloud
provider
manages
and
installs
communities
for
you,
which
is
truly
amazing
and
2018
is
the
year
that
pretty
much
everyone
in
the
Middle
where
business
and
software
business
as
a
doctor
Cuban,
if
you
so
docker,
is
adopted.
Communities,
Oracle
and
IBM
cloud
foundry
and
pivotal
support
Cuban
it
easy
PKS,
all
the
public
cloud,
binders
pretty
much
have
a
man
is
keeping
its
service
now
and
then
for
on-premise
as
law.
So
most
vendors
out
there
like
kept
you
and
Red
Hat,
various
others.
B
So
now
a
lot
of
developers
are
now
faced
with
the
challenge
of
learning
human
matey's
and
learning
how
to
use
it
and
then
how
to
do
things
like
cncd,
coupled
with
this
we're
all
increasingly
moving
towards
micro
services,
and
so
we
want
to
everyone
wants
to
be
a
high-performing
team.
So
we
all
want
to
go
fast
and
do
small
little
iterations
of
code
and
get
that
code
quickly
into
production
as
people
as
possible.
B
So
you
can
get
feedback
from
users
to
know,
what's
good
and
what's
bad
and
how
to
improve
so
we're
all
trying
to
iterate
and
continuously
improve.
So
we
all
basically
want
CI
MTV
on
communities.
This
is
where
we're
all
kind
of
aiming
for
our
aiming
towards.
We
want
automated
CNC
deep,
so
we
can
just
spin
about
code
and
stuff
just
happens,
and
today,
if
you're
trying
to
be
a
high-performing
team,
you're
now
faced
with
all
this
stuff
to
figure
out,
so
you
need
to
set
up
a
cloud
somewhere
or
pick
a
cloud.
B
You
need
to
set
up
communities
clusters.
You
need
to
set
up
environments
like
development
testing,
staging
production.
You
need
to
move
your
application
code
into
containers,
so
you
start
building
docker
images
now,
rather
than
you
need
to
figure
out
how
to
deploy
your
application
on
Cuban
ATS,
whether
you
generate
some
Yamal
or
whether
you
some
shots,
then
you
need
to
figure
out
how
to
do
the
actual,
continuous
delivery
thing,
releasing
the
images
and
Yama
on
and
then
how
to
do.
B
Using
you
know,
Best
of
Breed
DevOps
practices
like
immutable
infrastructure,
there's
also
a
new
term
in
your
DevOps
handbook
called
git
hops
now,
which
came
from
the
movie
folks
and
it's
basically
a
technique
of
doing
operations
through
using
gates,
in
other
words,
store
all
of
your
operational
configurations
like
what
applications
are
running
in
staging
in
production
and
what
their
configurations
are
and
so
forth,
but
I
think
it
and
use
gate
is
the
master
record
all
of
your
environmental
configurations
of
your
clusters.
Why
use
gate?
Well.
Gate
works
many
many
well
for
source
code.
B
So
let's
operate
software
through
source
code
so
that,
for
example,
rather
than
changing
something
you
huge
CTL
apply
in
production
and
then
nobody
else
on
the
team
knows
what
you
just
did
and
all
of
a
sudden
production
breaks-
and
nobody
really
knows
put
all
of
that
stuff
in
gates,
repository
and
then
use
github.
So
then,
you
can
see
via
pop
request
and
commits
who
changed
what
when
in
production,
and
then
you
can
do
importantly,
you
can
roll
things
back
when
things
go
bad
in
production
right.
So
detox
is
a
really
nice
approach.
B
Then
we
need
to
automate
everything
and
get
feedback.
So
there's
all
this
stuff
that
we
all
have
to
do
is
developers
which
is
kind
of
unnecessary
heavy
lifting
right.
We
all
just
want
to
get
all
that
stuff
done
and
then
do
what
we
really
and
to
be
doing
is
focusing
on
delivering
actual
business
value
to
our
customers,
which
is
kind
of
what
we
should
be
doing.
So
we
have
this
kind
of
problem.
How
do
we
deal
with
all
this
stuff?
And
this
is
what
here,
Jenkins
X
comes
in
Jackie's
ex?
Is
your?
B
Is
your
personal
Butler?
Your
seafaring
naval
Butler,
who
sails
the
Seas
of
continuous
delivery
to
help
chart
your
course
to
so
Jackie's
ex
is
gonna,
try
and
help
you
deliver
continuous
integration,
produce
delivery
and
kubernetes.
So
Jenkins
X
is
a
new
sub
project
to
propose
a
budget
project
of
the
Jenkins
foundation.
It's
all
completely
open
source
and
Jackie's
ex
is
really
aimed
at
focusing
focusing
crystal
clearly
on
automating
CI
and
cd4
kubernetes.
B
So
what
does?
How
does
think
is
X
help
first
thing
Jake
is
X
does,
is
it
combines
a
whole
bunch
of
best-of-breed
open
source
tools
from
the
kubernetes
ecosystem,
together
into
a
single,
simple
to
install
and
configure
system?
So
Jake
is
X,
installs
and
configures
a
whole
range
of
open-source
projects.
One
of
the
lovely
things
about
the
CN,
CF
and
kubernetes
is
there's
a
huge
range
of
really
awesome.
Open-Source
projects
that
we
all
should
adopt.
B
One
of
the
challenges
with
the
communities
ecosystem
is
keeping
track
of
water
or,
what's
all
the
good
stuff
from
what
should
you
be
using
to
implement
your
CI
te
so
the
whole
ecosystem
of
open
source
when
cubanÃa
DCA's
is
amazing
and
fantastic
and
cool,
and
there's
lots
and
lots
of
pieces
of
software
that
there's
lots
of
different
things?
Well
so,
for
example,
jinkies
X
comes
with
lots
of
different
things,
but
just
to
name
a
few.
B
We
use
helm
as
a
package
manager
for
packaging
up
your
yeah
mul
templates,
so
that
you
can
install
them
into
your
staging
in
your
production.
Environments
and
helm
also
helped
us
a
few
things,
but
it
it's
a
package
manager
like
helm
and
like
a
young,
NPM
and
there's
seven
things,
but
they
also
just
from
configuration
management.
So
it
helps
you
when
you
install
at
helm
package,
which
we
call
them
charts
helm,
takes
care
of
managing
any
environmental
configuration
you
need
to
override.
B
As
you
install
it,
then
you
can
just
upgrade
that
applications,
a
new
version
comes
out
and
it
keeps
your
configuration
with
the
new
version.
So
you
can
roll
forwards
and
backwards,
while
also
the
in
configuration
management.
We
use
a
tool
called
scaffold
from
Google
which
takes
care
of
abstracting
away
the
differences
between
how
you're
actually
going
to
do
the
docker
builds
on
the
Cuban
disease.
Cluster
I
will
get
into
it
too
much
right
now,
but
there's
a
whole
different.
B
There's
all
kinds
of
different
ways
of
building
docker
images
on
accumulators,
cluster,
using
the
clock,
a
daemon
directly
using
an
open-source
project
from
Google
can
occur.
That
avoids
talking
to
the
doctor
daemon
to
build
up,
and
just
all
things
like
Google
container
build.
There
are
various
other
cloud
services,
so
scaffold
takes
care
of
that
abstracting,
a
way
of
building
docker
images
and
doing
various
development
things
we
should
show
I'll
show
you
later
there
various
other
tools
like
nexus
for
handling
artifacts.
B
You
know
you
might
be
using
cuban,
it
is,
but
you
still
want
to
be
releasing
things
like,
if
you
don't
say
spraying
or
java,
unreleased
homes
and
jails
and
libraries
and
stuff
you
might
be
using
node
than
doing
npm
stuff,
you
might
be
releasing
various
other
artifacts.
There's
a
project
called
monocular,
which
is
kind
of
like
an
open-source
app
store
for
browsing
all
of
the
applications.
Your
team
has
built
and
jenkins
accessibility,
so
it's
kind
of
like
an
app
store.
B
We
can
look
at
all
the
different
apps
and
all
the
different
versions
to
help
teams
share
their
apps
with
other
teams.
So
you
can
kind
of
search
and
find
applications
that
you're
building
so
jagat
x,
automates
the
installation
and
configuration
of
all
that
staff.
So
that's
kind
of
cool,
but
the
other
thing
it
does,
which
is
super
important.
It
alternates
the
CI
NCD
for
your
actual
application,
so
the
actual
apps
you'll
be
developing
and
building
it
automates.
B
B
I'll
show
you
how
all
works
in
a
few
moments,
but
the
basic
idea
is:
we
use
git
repositories
for
all
the
environments
to
manage
promotion
and
configuration
management
of
the
environments
using
a
single
gate,
workflow
using
for
requests
and
so
forth,
so
that
your
your
team
or
possibly
multiple
teams
can
work
together
to
prove
commercials
and
so
forth,
and
finally,
jakey's
X
gives
you
lots
and
lots
of
feedback
so
that
all
sounds
kind
of
cool.
How
do
we?
How
do
we
use
cuban
eighties
Jenkins
together?
So
we
basically
need
the
Kuban.
B
It
is
cluster
and
Jenkins
X
to
be
installed
on
the
cluster.
Now
not
everyone
has
a
cluster
already.
So
there's
a
couple
of
different
approaches,
but
before
we
get
that
far,
the
first
thing
you
need
is
a
binary
called
GX.
So
if
you
already
use
cube
annuities,
you
know
about
cube,
cuddle
or
cube
CTL,
depending
on
how
you
say
it.
Q
control,
vs.,
cube,
CTL
binary
that
you
used
to
do
to
interrogate
cube
annuities.
Jenkins
X
has
its
own
binary
called
GX
for
Jenkins
X,
which
is
a
go
binary.
B
It's
statically
linked
binary
that
you
can
download
for
Windows,
Mac
and
Linux.
If
you
use
a
Mac,
you
can
use
brew
to
install
it
use
Linux,
you
can
use
curl
or
you
can
just
go
to
that
web
page
and
download
the
binary
and
just
put
it
on
your
path
so
install
JX,
because
you
will
use
that
command
line
quite
a
lot.
If
you
adopt
Jenkins
X
and
we
hope
you
do,
then
these
are
the
two
commands
to
install
Jenkins
X
and
set
it
all
up.
B
So
there's
two
options
here:
the
first
option
is
to
create
a
new
communities
cluster
and
the
install
Jenkins
X
on
top
of
it.
The
second
option
is,
if
you
already
have
a
cluster
that
you
want
to
use
you
just
type
Jegs
installer,
we
kind
of
prefer
the
first
option,
because
then
you
can
just
spin
up
a
new
cluster
play
around
with
it
and
then,
if
you
want
just
kind
of
tear
it
down
and
it's
kind
of
an
isolated
customer,
you
can
just
kind
of
play
with.
B
We
also
preferred
you
use
the
public
cloud
to
start
with
either
Amazon
or
Google
or
Azure
they're
all
supported.
You
just
type
that
one
command
in
Jake's
create
cluster
then
address
for
Amazon
gke
for
Google
aks
for
was
Europe.
We
have
a
bunch
of
other
providers
and
we're
trying
to
provide
as
many
clouds
as
we
can
and
we're
trying
to
eks
was
just
announced
they
pretty
much.
So
we
hope
so
we
rather
you
use
the
public
cloud
because
then
web
hooks
work
would
get
up
and
all
that
kind
of
stuff.
B
But
if
you
have
a
cluster
already
use
that
that's
all
fine
Jake's
install
works
as
well.
There's
a
URL
there
that
you
can
check
on
to
get
more
details
of
how
to
do
it
all
and
what
I'll
show
you
is
a
very
quick
demo
of
how
the
install
works.
This
demo
is
sped
up
a
little
bit.
We
type
JX
peer
cluster
gke.
It
asks
you
a
few
questions
about
machine
names
and
zones
and
whatnot
and
machine
types.
Then
it
spins
a
bit
giving
it
this
cluster.
B
It
sets
up
the
ingress,
so
you
can
talk
into
the
cluster
there
installs
the
helm
chart
for
Cuban
eighties,
its
foots
away
from
Jackie's
ex
and
then
the
final
thing
it
does
is
set
up
to
get
me
prostitutes
for
the
dip
for
the
get
up
to
peace
which
we'll
see
in
a
moment
so
the
whole
thing.
This
is
sped
up
a
little
bit
and
the
whole
thing
on
my
laptop
tensed
LG
kita,
to
take
two
to
three
minutes
from
scratch.
On
the
other
clouds,
it's
a
little
slower,
but
you
know
within
10-15
minutes.
B
You
should
be
all
set
up
and
ready
to
go
so
once
you've
typed
your
JX
create
Krusty
command.
You
should
be
all
set
to
now
enjoy
and
have
fun
with
Jenkins
X.
So
what
does
James
X
gives
you
so
typing,
but
one
command
in
what
you
get.
So
the
idea
is
once
you've
run
that
one
code,
man,
each
team,
gets
their
own
development
tools,
environment,
so
basically
Cuba
Nettie's,
name
space,
where
all
your
development
tools
run
and
your
builds
run
and
so
forth.
B
So
right
now
each
team
gets
their
own
master,
so
the
run
Jake
is
master,
so
you
might
use
Jenkins
already
in
your
job
and
you
might
have
like
a
big
Jenkins
server
with
a
gazillion
jobs
in
it
we
give
each
team
their
own
Jenkins
master.
So
if
you
decide
to
use
the
Jenkins
UI,
you
only
see
your
pipelines
and
not
everything
in
your
entire
organization.
Each
team
also
gets
an
elastic
pool
of
build
agents
for
running
your
builds.
So
you
don't
have
to
fight
all
that
this
kind
of
limited
build
queue
in
Jenkins.
B
Each
team
gets
there
an
elastic
pool
that
if
Kuban
eighties
has
enough
capacity,
all
your
bills
just
run
immediately.
Also
there's
things
like
Nexus
and
immaculate.
In
addition,
each
team
gets
their
own
staging
and
production
environment,
so
in
other
words,
each
team
can
then
create
apps
and
go
straight
to
the
staging
environment
automatically
without
having
to
have
a
meeting
with
other
teams
and
agree
that
you're
allowed
to
do
this
right.
Each
team
can
just
deploy
their
stuff
immediately
to
the
staging
environment
out-of-the-box.
B
We
do
automatic
promotion
to
staging
and
then
we
leave
promotion
to
production
as
a
manual
step.
You
can
then
create
as
many
environments
as
you
wish
and
call
them
whenever
you
want
and
an
environment
can
either
be
automatic
or
manual.
The
other.
The
box
defaults,
our
staging
and
production
and
staging
is
automatic
in
production
these
manual,
but
you
can.
We
hope
everyone
becomes
automatic
for
everything
eventually,
but
we've
left
it
as
this
kind
of
safe
default
and
then,
when
you're
happy
and
calm
you
can
automatically
promotes
production.
So
let
me
switch
gears
now.
B
Let
me
just
do
a
live
demo
of
chickens,
X
and
I'm
going
to
do
a
live
demo,
Tinky's
X
on
Google,
Cloud
and
I'm,
going
to
use
github
on
Google
Google
Cloud.
So
what
could
possibly
go
wrong
right?
A
live
demo
on
a
webinar
over
a
hotel,
Wi-Fi
is
good.
So
I've
got
the
JX
binary
on
my
path.
They
protect
JX.
It
gives
you
all
the
gazillions
of
commands
that
were
inside
the
Jegs
binary.
Let
me
clear
that,
because
it's
kind
of
noisy
so
the
first
one
I'll
show
you
is
gate.
B
J,
it's
get
in
for
getting
Berman
lists,
my
environments
that
I've
got
so
I've
got
the
development
environment.
I've
got
personal
Eddy
environment,
which
I'll
show
you
later
I
learned
the
stairs
you
in
the
prediction,
so
whatever
I
do
now
is
I'm
gonna
create
a
brand
new
app.
If
I
take
JX
get
up,
I,
don't
think
I've
any
apps
running
so
I
got
no
house
running
so
I've
got
a
completely
empty
Jenkins,
X
environment
I
mean
I'm
gonna
create
a
new
application.
B
We
support
a
whole
bunch
of
different
wizards
and
languages
for
creating
brand
new
projects.
You
can
also
import
an
existing
project
with
JX
import
I've
been
a
style
creating
a
brand
new
micro
service,
and
since
this
is
CN,
CF
and
so
much
in
the
Cuban
is
ecosystem
uses.
Golang
I'm
gonna
do
a
go
along
demoing,
a
weak
support,
Python
Perl,
PHP
Java,
obviously
stream
boot,
those
other
things,
but
us
I'll
stick
to
go
along
for
this
demo.
B
So
I'm
going
to
do
the
create
quick
start
command,
which
has
a
library
of
different
quick,
starts
to
use
as
your
initial
base
of
a
project
and
there's
a
whole
raft
of
these,
and
you
can
add
your
own
quick-
starts
into
this
menu
and
configure
the
quick
starts
on
the
13
basis,
so
that
you
could
just
pick
the
approved
languages
or
friends
or
in
your
organization.
So
I'm
gonna
pick
the
girl
Lang
example.
Oh,
actually
I
need
to
go
into
the
computer
run
directory.
So
it's
okay
I'll
stay
here.
B
Let
me
create
a
going
example
and
I'm
gonna
call
this
city
ends
the
rocks
there
we
go.
Actually
that's
add
you
might
see.
It
looks
like
C&C
frogs,
but
nevermind.
So
we've
got
a
new
example.
I'm
going
to
say,
use
my
github
username.
Yes,
that's
good.
Would
I
like
to
initialize
a
git
repository?
Yes,
so
it's
copied
these
files
onto
my
file
system.
So
it's
a
very
simple
app,
as
you
can
probably
see,
it's
a
main
go
and
a
couple
of
shell
scripts.
So
that's
the
quick
stir.
It's
created
from
the
library
of
quick
starts.
B
Now
it's
going
to
run
simpler,
buildpack,
which
is
going
to
add
a
bunch
of
extra
files
which
I'll
talk
about
in
them
to
be
able
to
do
see
on
TV.
Now
it's
gonna
push
this
git
repository
onto
github,
so
that
we
can
use
it
from
different
people
can
work
together
as
a
team,
so
I'm
going
to
use
my
personal
gate
of
user
and
I'm
going
to
use
that
name
as
a
git
repository.
B
So
now,
Jackie's
ex
is
going
to
do
a
git
push
I'm
going
to
create
a
project
from
get
up
and
do
a
git
push
and
that
all
worked.
So
now
we've
got
a
new
gate
repository
on
github
calm.
So
let
me
just
show
you
that
I'm,
not
cheating,
I,
say
I'm,
just
gonna
copy
and
paste
this
command
here.
This
command
is
kind
of
Handy.
B
This
is
going
to
output
to
my
terminal,
any
activity
in
terms
of
pipelines
and
bills
that
take
place
on
my
project
I'll
just
kind
of
keep
that
running
in
the
background,
so
we're
basically
watching
a
custom
resource
in
humanities
that
describes
all
the
pipeline's
happening
on
my
project,
so
just
to
show
you
that
I'm,
not
cheating,
I'm,
gonna
click
on
this
github
URL
show
you
the
project.
So
this
is
github.com
recently
acquired
by
Microsoft,
which
I
think
is
awesome.
News
I
couldn't
think
of
a
better
place,
for
it
hope
to
end
up
by
I.
B
Think
Microsoft
is
an
awesome
player,
get
up
to
be
and
I'm
super
excited
to
see
what
happens
with
vs,
currently
inside
a
tab
that
could
be
epic
and
do
it.
So
this
is
get
up
to
come.
This
is
a
point
I
just
created.
You
can
see
it's
illuminated
girl,
there's
the
girl
source
file,
but
I'll,
probably
edit,
fairly
soon
notices
a
bunch
of
extra
files
being
created.
There's
a
docker
file
which,
because
the
QuickStart
didn't
have
a
docker
file,
he's
added
a
docker
file.
B
The
bill
packs
I,
did
a
docker
file,
so
we
can
make
a
docker
image
for
this
code,
each
other,
the
Jenkins
file,
which
defines
the
cncd
pipelines
and
exactly
the
scaffold
yeah.
Also
we
can
use
scaffold
to
do
interactive
development,
I.
Think
I
did
some
home
charts
now
notice.
All
of
this,
we
don't
even
need
to
look
at
these
effects.
B
I'm
going
to
show
you
them,
we
don't
need
to
look,
it
does
matter,
but
if
you
really
really
care-
and
you
really
want
to
tweak-
say
the
docker
file-
you
can
just
edit
the
dockerfile
in
get-
do
a
pull
request.
Most
omastar
and
you've
now
got
a
new
docker
file,
deter
the
pipeline
or
the
chart.
But
the
idea
we're
trying
to
do
is
a
lot
of
teams
were
adopting
communities
and
the
lot
of
developers
done.
There
were
Cuban
eightydaze,
they
don't
know
what
helmet
is.
B
They
don't
have
to
write
blocker
files,
so
we're
trying
to
automate
all
of
this
stuff
so
that
developers
don't
need
to
know
or
care.
So
you
could
squint
and
see.
This
is
a
kind
of
a
Cerberus
way
of
working
with
keeping
because
you
don't
have
to
know
Cuban,
it
is.
Is
there
you
don't
have
to
know
about
all
this,
but
when,
like
serverless
we're,
not
hiding
dockerfile
jenkins
files
and
charts,
you
can
go
in
there
and
tweak
these
things
to
her
content.
If
you
wish
now,
you
might
have
noticed
a
bunch
of
activity,
that's
happened.
B
We've
already
released
zero,
zero
warrants.
What
happened
there?
Is
I
typed
one
command.
We
spun
up
a
brand
new,
quick
start
and
gate,
and
it's
already
created
and
release
versions
released.
Zeros
are
the
one
we
create
semantic
version,
so
we
start
at
zero
one.
There
we
go
incrementally
so
we've
created
a
new
release,
number
zooms
on
one
and
we've
tagged
the
git
repository
with
the
source
code
and
we
generated
some
nice
release.
Notes
we've
also
under
the
covers,
created
the
docket
image
tag
that
zooms
in
the
one
and
push
that
to
the
doctor
registry.
B
We've
also
created
the
hound
chart
for
zero
zero
one
and
push
that
into
the
Hound
registry
into
monocular.
So
anybody
can
now
deploy
that
application
right
now
and
any
Cuban
Izzy's
cluster,
just
using
helm
directly
without
using
any
of
the
jiggies
x,
give
up
stuff.
If
you
chose
to
do
so,
I,
don't
think
you
should
do
that.
I
think
you
should
stick
with
fingers
X,
but
if
you
wanted
to
do
that,
that's
all
pinnacle,
so
we've
automatically
created
a
new
release
and
notice.
We've
created
a
pull
request.
Now.
B
What
we're
trying
to
encourage
is
good,
good,
behavior,
we're
trying
to
encourage
everyone
to
become
a
high-performing
team.
So
as
soon
as
you
create
or
import
a
project,
we
want
to
go
straight
to
the
staging
environment
and
release
and
go
to
staging
so
that,
as
you
then
start
to
develop
this
thing
and
do
code
changes,
you
do
that
through
pull
requests
and
we
try
and
encourage
everyone
to
release
frequently
into
the
staging
environment
right
and
then
you
choose.
B
When
you
want
to
go
to
production
notice,
we've
got
a
pull
request:
I'm
going
to
click
the
pull
request,
this
protocol
I
emerged
already.
I
might
really
I
was
too
slow.
So
what
we
did
was
generated
pull
request
on,
and
this
pull
request
is
on
the
staging
environment.
So
this
is
a
pull
request
to
change
the
staging
environment
by
a
ragged
connect,
so
we've
proposed
through
a
pull
request,
a
curve
change
to
the
helm
and
chart
for
a
staging
environment.
Develop
got
this
diff
very
briefly.
B
So
if
you
want
to
configure
staging
and
production
slightly
differently,
you
can
do
all
of
that
in
the
staging
and
the
production
and
give
your
postures
and
managing
it
all
with
cncd.
In
the
same
way,
we
do
ax
right,
so
that's
kind
of
cool,
so
you'll
also
notice.
The
pull
request
has
succeeded.
So
let
me
just
recap
what
we
did.
We
did
one
command
to
create
a
new
project
that
create
a
brand
new
game,
repository
that
also
set
up
a
jenkins
cnce
pipeline,
which
then,
because
we
pushed
a
master,
released,
zero,
zero
one.
B
If
ever
we
want
to
go
to
zoom
is
over
to
it
will
automate
that
if
ever
you
want
to
jump
by
the
way
to
say
to
zero,
three
zero,
so
just
create
a
gate
tag
and
the
automatic
pipeline
will
read
the
get
tags
and
always
take
the
next
highest
version.
Number.
If
you
say
maven
or
you
use
node,
you
can
use
the
pond
XML
of
the
package
Jason
to
do
major
minor
version
upgrades
as
well
or
just
use
get
text.
So
we
one
command-line
creating
a
new
project
and
we
released
it.
B
We've
also
used
the
get
ops
and
a
pull
request
to
promote
this
into
the
staging
environment
and
if
I
click
on
this
URL
here
we
can
see
the
application
running
in
the
staging
environment.
They
go.
We're
now
running
the
application
in
the
staging
environment
after
that
promotion.
So
that's
kind
of
cool
right,
so
with
one
line
of
one
command
line,
we've
automated
this
tier
LCD
of
an
application
and
release
the
version.
So
now
let
me
show
you:
let
me
change
gears
a
little
bit.
B
Let
me
show
you
how
to
make
a
pull
request
into
a
curve
change.
So
let's
switch
a
different
kernel
and
I
call
this
CN
CF,
Fox
and
I'm
gonna
make
a
coaching.
So
let
me
pop
up
my
trusty
vs
I'm
gonna
make
a
little
code
change
and
you'll
notice.
The
girl
code
is
it's
super
awesome.
It's
a
super
awesome,
go
good,
so
I'm
gonna,
add
change
the
title.
I'm
gonna
say:
I'm
gonna,
add
some
awesome,
so
I've
changed
the
source
code.
B
What
I'm
actually
going
to
do
is
I'm
going
to
create
a
new
issue
in
github,
so
I
can
do
that
by
the
website.
Obviously
they
can
also
create
an
issue
from
JX,
so
I'll
create
your
shoe
saying
more
awesome.
So
I've
created
a
new
issue,
though,
and
there's
the
URL
on
github
and
now
I'm
gonna.
Do
a
git
commit
I'll
use
the
command
line
just
to
show
you
there's
no
weird
magic.
This
is
just
my
entire
day.
B
Man
I
could
be
just
vanilla
gate
from
the
command
line,
so
I'm
gonna
do
create
a
branch
I'll
call.
It
fixes
check
out
fixes
and
then
I'm
gonna
get
commit
mine
say
my
name
and
I'll
say
fix
more
awesome
and
this
fixes
issue
and
one
I'm
gonna
do
a
git
push
and
that's
pushed
ups.
We
get
them.
So
that's
just
vanilla
gates
like
no
magic,
though
we
add
Jackie's
ex
stuff,
just
vanilla,
Jing
get
stuff.
B
B
So
I'm
gonna
click
get
a
website
and
say
let's
create
a
pull
request
and
then
say
so
notice
now
that
any
second
now,
hopefully
the
web
browser
should
be
French,
probably
if
I
open
this
tunnel
by
the
way
you
should
see
what
is
a
pull
request
pipeline
is
now
but
trigonal
on
so
because
we've
created
a
pull
request.
The
CIN
CD
that
we
automatically
set
up
because
we
using
Jenkins
X,
is
now
running
a
CI
pipeline
on
every
pull
request.
B
So
we
do
releases
when
you
merge
to
the
master
branch
and
when
you
submit
a
request,
we
do
a
CI
pipeline
and
the
CI
pipeline
he's
doing
the
things
that
people
normally
do
with
CI
right
check
that
the
pull
request
does
he
compiled
as
he
plants
all
the
code
quality
checks.
There
was
everything,
look
good
and
acceptable,
so
it's
running
all
those
kind
of
bills.
We
also
by
default
in
chengdu
statistics,
do
something
else
as
well.
B
We
do
something
called
preview
environments,
so
whenever
you
create
a
pull
request
by
default,
you
can
opt
out
of
this,
but
by
default,
whenever
you
create
a
pull
request,
we
do
the
normal,
builds
checks
and
tests
and
check
everything
works.
Then
we
build
a
preview
docker
image
for
that
pull
request
and
push
that
to
a
docker
registry.
Then
we
create
a
preview
environment
which
is
basically
a
separate
cuba,
Nettie's
namespace
for
that
pull
request.
B
Then
we
deploy
the
hell
chart
into
that
namespace,
so
we
basically
deploy
the
pull
request
it
in
kubernetes,
as
as
it
would
be
deployed
if
it
was
in
staging
a
production
before
we
merge.
Now
we
do
this
for
a
few
different
reasons.
The
first
thing
is
we're
now
checking
that
does
the
code
actually
still
work
in
keeping
it?
It's
like
it's
quite
easy
if
you're,
particularly
if
you
fiddled
with
environment,
rebels
or
docker
files
or
cuban,
it
is
yeah
mole
or
something
it's
very
easy
to
break
an
application,
keeping
it
it's
right.
B
So
the
first
thing
the
preview
does
is
check
those
stuff
still
working
kubernetes,
but
the
other
thing
is.
It
really
helps
us.
Do
testing
of
various
kinds
if
you're
building
a
web
console,
for
example,
sometimes
by
just
looking
at
the
diff
on
the
pull
request,
you're,
not
quite
sure
what
the
change
means
or
how
the
change
is
going
to
look
as
a
UI
in
this
case
notice
that
the
pull
requests
pipeline
is
is
completed
so
that
everything
is
green.
So
we
can
merge
this
we're
ready
to
merge
but
notice.
B
This
comment:
Jenkins,
X's,
all
rotate
automatically
comment
on
the
pull
request
to
say
there
is
a
preview
environment
created
this
pull
request
and
there's
a
URL
that
links
to
the
app
in
the
preview
environment.
So
if
I
click
on
that
URL,
that's
opening
a
new
environment,
a
preview
environment
for
the
pull
request,
which
has
more
awesome.
So
our
the
code
change
is
now
running
in
its
own
separate
environment,
isolated
where
everything
else
before
we've
merged
anything.
Now.
B
This
is
awesome
on
many
levels
that
we
use
preview
moments
all
the
time,
our
selves
to
building
insects,
particularly
with
you
eyes.
This
is
amazing
because
any
code
change
in
the
UI
it
means
you
can
just
look
at
the
UI
and
have
a
look
at
it
and
see
what
it
feels.
You
can
also
do
automatic
testing
like
using
selenium
or
some
kind
of
load,
testing
or
contrabass
testing
tool,
because
a
preview
environment
is
literally
just
a
Haunter.
You
can
make
a
preview.
B
You
can
define
what
that
preview
button
is
completely
up
to
you,
so
you
can
deploy
other
micro
services,
you
can
link
preview
environments
or
the
preview
environments,
you
can
add
testing
applications
and
various
other
jobs
and
so
forth.
So
peak
environments
are
super
super
useful
and
particularly
for
you
eyes,
another
thing
about
creeping
moments.
By
the
way
it
is
sometimes
you
can
make
a
change
on
multiple
microservices,
so
you
might
want
to
make
a
pull
request
on
a
back-end
service
to
add
a
new
api.
B
Then
you
might
want
to
change
a
front-end
to
use
that
new
api,
so
you've
got
to
pull
requests.
You
can
link
together
the
previous
of
different
pull
requests
so
that
the
pull
requests
on
the
UI
can
be
using
the
pull
request
at
the
creepy
environment.
For
a
different
back-end
right,
so
you
can
do
sophisticated
testing
of
changes
before
you've
actually
merged
anything.
B
So
this
is
awesome
for
hypothesis
driven
development
where
you
just
want
to
experiment,
try
something
out
across
multiple
no
spaces
and
repositories
and
whatnot
test
it
all
out
and
then,
when
you're
all
happy,
let's
go
crack
on,
rather
than
merging
everything
to
master
all
the
time
and
then
having
to
back
and
work
back
out
things,
conflicts
and
stuff.
So
that's
previa
moments.
Now
this
is
working
great.
So,
let's
hit
merge.
B
So
we
validated
the
pill
time
visible
request
it
automatically
validated.
He
could
build
it
automatically.
They
could
run
in
cuba
knitting's.
We
get
to
truck
it
and
look
at
it.
We're
happy
with
it.
So
we
click
merge.
So
now
we
click
merge
notice
that
the
pipeline
is
now
running
a
new
master
built.
So
now
we're
gonna.
Do
we're
gonna
release
zero
zero
to
this
application.
So
let
me
briefly
recap
what
we
just
saw
that
was
kind
of
a
lot
of
looking
at
consoles
and
web
applications.
B
We
wrote
one
command-line
j.x,
create
QuickStart
and
that
create
a
brand-new
project.
It
create
a
brand
new
get
repository
on
github.
It's
set
up
the
CI
CD
pipeline
for
you
automatically,
then
look
at
the
Jenkins
file
of
a
docker
file,
I'm
sure,
and
then
we've
automated
preview
environments
and
every
pull
request
and
then,
whenever
we
merged
anything
to
master,
we
automatically
do
proteins
delivery
into
the
staging
moment
and
then
any
point
in
time.
B
We
can
then
manually
promotes
production
using
the
JX,
promote
command
line,
so
automated
cncd
on
kubernetes
I'm
gonna
leave
that
release
going
because
it
will
take
maybe
a
minute
to
complete
and
we'll
come
back
and
look
at
it
when
it's
completed.
What
I
like
to
do,
because
I
won't
leave
plenty
of
time
for
questions.
Hopefully
there's
quite
a
few
I'd
like
to
show
you
one
other
thing.
B
Now
some
people
do
like
work-in-progress
pull
requests
and
user
pull
request
is
like
a
place
to
hack
and
code
before
they're
ready
to
promote
them,
but
often
developers
wanted
like
Hackford
locally
and
just
try
it
out
right
now,
as
developers
work
when
the
notorious
for
running
stuff
on
our
laptops,
which
are
completely
different
to
the
production
environment.
But
we
might
have
a
Mac
or
a
Windows
machine
and
we
might
run
the
stuff
on
a
Windows
machine
and
it
seems
to
work
okay,
but
then
it
works
different
than
kubernetes.
B
So
we
want
to
encourage
people
to
use
the
communities
all
the
time
for
development
right,
always
use
the
same
operating
system,
containers
and
tools
all
the
time,
because,
from
a
from
a
set
of
DevOps
kind
of
high-performing
team
thing
we're
reducing
waste
by
not
testing
apps
on
the
wrong
thing.
We're
always
testing
on
the
right
thing
right.
So
we're
not
worrying
if
things
go
weird
on
the
windows,
but
it's
because
production
doesn't
use
me
in
those
or
whenever
right.
B
So
we
have
a
thing
called
death
pods
which,
if
I
switch
into
a
BS
code
or
whether
let
me
hide
Nathan,
because
this
project,
we
also
have
a
vs
code
extension
for
Jenkins
X,
which
wanted
to
show
you
very
briefly
so
that
what
this
does
under
the
covers.
Jenkins
X
is
using
humanity's
resources,
custom,
kubernetes
resources
to
define
pipelines
and
environments
and
all
those
kind
of
things,
and
so
we
have
a
little
plugin
for
PS
code
that
can
watch
and
browse
all
of
your
pipelines
for
each
pipeline.
You
can
see
what
it's
doing.
B
You
can
stay
right,
click
and
say
gimme
the
build
load
because
I,
don't
know,
I
wonder
what
should
I
want
to
look
at
the
gory
detail
of
everything
that's
happening
in
there
in
the
pipeline
hydro
bill.
Usually,
if
a
bill
goes
wrong,
you're
gonna,
look
at
the
bill,
blogs
and
stuff.
So,
with
a
couple
of
clicks,
you
can
watch
what's
happening
in
CI,
NCD
and
watch
things
happen
live
so
that's
kind
of
cool
and
we
should
know
what
is
the
the
pipeline's
almost
finished.
B
So
it
should
go
from
the
green
fairly
soon
to
implement
the
pipeline's
way.
We
use
something
called
a
build
pod,
which
is
really
just
a
community's
part.
We
build
tools
inside
it,
so
if
you're
in
this
case
wheezing
go
so
there's
the
girl
tools
in
there,
but
there's
also
tools
like
gates
and
cube
cuddle
and
helm
and
JX
varies
a
little
scaffold
so
of
a
whole
range
of
tools
we
use
to
implement
the
CNC
pipelines.
B
We
have
a
thing
called
a
dev
card,
which
is
basically
we
uses
the
exact
same
part
for
pipelines,
but
for
developers
to
use
so,
in
other
words
the
exact
same
version
of
Q,
cuddle
and
helm
and
scaffold
and
docker,
and
all
those
command
line
tools
we
use
to
do
cncd.
All
those
tools
are
available
to
a
developer.
So
if
I
open
the
the
toolbar
menu
in
vs
code,
I
can
type
this
open
their
pod
terminal
command.
B
Now
what
this
is
going
to
do
is
open
a
new
terminal
that
is
actually
going
to
run
in
Cuban
eighties,
we'll
just
give
it
a
few
minutes
to
start
up,
I
think
my
cluster.
Usually,
these
things
start
up
really
really
quick,
oh
I'm,
using
the
wrong
demo.
That's
why
it's
ticking
well,
I.
Did
this
demo
about
an
hour
ago
and
I
was
using
a
completely
different
their
pod.
So
that
is
why
it's
taking
a
while
to
spin
it.
So
this
might
take
a
few
seconds
for
this.
Dev
want
to
spin
up.
B
So
what
happens
with
the
death
pod
is
that
their
body
is
creating
a
brand
new
pod
on
Cuban
eighties
and
the
dev
pot
starts
out,
and
then
this
shell
opens
in
that
that
pod.
So
you
basically
using
a
command
line
shell
and
it
looks
and
feels
like
you're
on
your
laptop
but
really
you're,
running
everything
inside
cube
entities
in
the
cluster
and
it
all
works
beautifully.
B
What
we
also
do
is
we
synchronize
the
source
code
from
your
desktop
IDE
into
the
dev
pot,
so
that
we
can
do
run
tests
or
builds
inside
community
using
the
cube
aneesa's
cluster
to
do
bills?
I
talk
about
you,
but
if,
if
I
do
a
docker
build
on
my
laptop
it
takes
hours
like
my
Wi-Fi
is
kind
of
slow,
it
takes
forever
to
download
docker
images.
Pulling
and
pushing
docker
images
on
your
laptop
is
really
super.
B
Slow
building,
knocking
images
in
the
public
lands,
I,
Cuban
FTC
super
super
fast,
so
as
a
place
to
develop
the
cloud
is
amazing,
and
so
that
therefore
helps
us
automate
synchronizing
curve
into
the
dead
pod,
and
then
then
we
can
like
do
bills
and
so
forth
inside
it
well.
This
is
taking
a
little
while
I
might
actually
switch
projects
to
another.
A
node-based
application
was
just
doing
a
demo
earlier
on.
So
let
me
switch
that
that
was
taking
a
while
I
think
another
covers
Google
is
now
auto
scaling
by
cubing.
B
This
he's
close
to
really
takes
a
minute
or
two
for
a
new,
and
you
know
to
start
up
so
well,
that's
something!
Let
me
do
a
demo
over
here.
This
is
another
application.
It's
basically
a
similar
application
using
the
GS.
So
same
kind
of
thing,
no
jest
thing:
I've
got
my
pipeline
here
that
I
can
watch.
What's
something
oh
and
the
builds
finished
by
the
way
just
before
I
show
the
DEF
bones
really
really
quickly,
because
someone
who's
time
for
questions,
let
me
just
go
back
and
show
you
one
more
thing.
B
So
we've
just
released
Sierra's
over
to
using
the
pull
request.
So
here's
the
pulley
press
was
in
two
requests.
There
we
go,
there's
a
pull
request
that
releases
o2.
So
let
me
just
show
you
the
diff,
so
the
diff
to
promote
from
zero
zero
one.
Two
zero
co2
is
literally
just
changing
one
string
for
the
version
number
right.
So
that's
nice
and
simple.
You
can
just
look
at
that
for
requests
and
go
I.
Get
it
I'm.
B
Changing
this
one
version,
it's
nice
and
simple,
and
the
other
thing
I
want
to
quickly
show
you
was
if
I
look
at
the
issues.
If
I
look
at
the
rim
turret
and
release
release
notice,
we've
now
got
two
releases:
0
0
2
and
0
0
1
and
oh,
maybe
it
didn't
work.
I
didn't
link
to
the
issue.
Don't
worry
about
that!
Let
me
move
on,
but
yes,
we've
got
versions
Arizona
to
my
quick,
switching
back
to
the
DEP
one
that
sees
their
pods
are
working
for
no
GS.
Hopefully
that
works.
That's
looking
better,
yes,
whew!
B
Ok,
I
will
no
GF
step
bodies
up
and
running,
so
I
can
type
LS
al.
So
I
am
now
running
commands
inside
the
docker
container
inside
communities
right,
so
everything
I
do
inside
this
command.
Shell
is
using
the
same
exact
same
software
that
the
CI
CD
pipelines
are
using
and
it's
using
the
same
kubernetes
cluster
and
containers
and
stuff
that
you've
been
using
production,
I'm,
gonna
type,
a
command
called
watch
which,
if
you
notice
that
built
at
docker
image,
that
was
pretty
quiet,
quick
right
that
built
a
docker
image
based
on
the
local
source
code.
B
That's
in
the
file
system
and
it's
one
of
the
cover
cities.
You
seen
scaffold
to
watch
your
source
code,
build
a
docker
image,
deploy
using
helm
into
a
personal
namespace
and
then
he's
watching
the
cursor
changes.
If
I
go
into
another
terminal
and
type
Jake's
get.
This
will
wrap
horribly
because
I've
got
a
really
big
font,
but
there's
the
URL
of
the
preview
application,
so
I
can
I
can
open
that
in
the
browser.
So
this
is
the
this
should
start
soon.
B
Hopefully
what
I
can
do
is
I
can
run
a
colon
of
this
application
and
hopefully
hopefully
this
pod
comes
up
soon.
The
demo
gods
might
be
allowed
to
strike.
Things
are
going
so
well.
Hopefully
this
comes
up.
I
was
going
to
show
you
was
so
let
me
just
show
you
the
demo
first
and
see
if
ever
this
pod
comes
up
its
up.
Yes,
you
always
just
Google
and
it
only
seems
to
work
eventually,
it's
just
sometimes
you
have
to
wait
a
moment.
B
So
whatever
that
is
I'm
gonna
make
a
code
change
so
I'm
in
a
desktop
IDE,
but
the
source
code
is
on
my
laptop,
so
I'm,
going
to
edit
the
source
code.
I'm
gonna,
add
some
awesome
awesome.
So
I've
edited
the
source
code
for
this.
No
GSR
notice
that
the
dev
pod
has
rebuild
a
docket
image
and
it's
gonna
redeploy
the
Hansa
pretty
quickly.
Usually
this
is
much
much
quicker.
B
It's
lot
of
it.
Usually
that
normally
takes
like
a
second
or
two:
it's
going
a
little
slow
right
now
and
notice
that
this
KO
loop
that
I
got
is
curling
the
end
point
it's
now
undeployed
the
old
part,
and
it's
ruining
the
new
part
and
they've
got
awesome,
so
I,
basically
just
changed
the
source
code.
B
Let
me
go
one
more
time
to
see
if
it's
any
faster,
very
awesome,
so
I've
chased
the
source
code
what's
happening
under
the
covers
is
that
sauce
has
been
synchronized
interactively
into
the
deaf
pod,
but
their
pods
gonna
rebuild
that
darker
image
me
deploy
it
and
very
soon.
I
should
see
the
new
pods
coming
up,
so
it
takes
between
2
and
like
10
seconds
to
get
a
live
check.
B
Now,
right
now,
we're
doing
this
with
brute
force,
we're
literally
making
a
brand
new
docker
image
each
time
and
redeploying
that
image
from
scratch
in
given
a
DS,
we
should
be
able
to
speed.
This
makes
much
much
faster
to
do
things
like,
for
example,
node
and
watch
for
new
new
node
files.
We
could
copy
the
binary
into
the
pod
itself
and
we
start
the
continual
something.
So
we
could
do
various
things
to
speed
this
up
a
little
bit,
but
already
we
can
do
pretty
quick.
B
B
So
the
IDE
would
then
be
inside
that
pod,
which
would
be
kind
of
even
cooler.
Then
there's
no
synchronization,
but
doing
this
lets
has
developed
inside
the
cluster
super
quickly.
Before
we
ready
to
commit,
then
we
can
just
do
a
git
commit
when
we're
happy
and
that
will
then
they
will
pull
request
and
then
we
can
do
CIN
TV
so
really
quickly.
That
was
a
very
quick
walkthrough
of
Jenkins
X
and
we
just
show
pop
a
few
links
up
while
we
chat
so
there's
a
few
links.
B
So
basically,
Jenkins
X
is
about
automating,
CIN,
CD
and
humanities.
We
want
to
help,
make
developers
super
productive
and
because
high-performing
teams
and
automating
as
much
of
the
development
in
CI
NC
see
the
experience
as
we
can,
using
all
of
the
amazing
open
source
technologies
in
the
CN
CF
and
given
a
DS
ecosystem.
If
you're
at
all
interested,
please
go
to
Jenkins
yo
and
check
it
out.
There's
lots
of
Docs
and
demos
up
there
even
after
this
webinar
is
complete.
If
you've
got
any
questions
at
all
or
any
feedback,
you
want
to
give
us.
B
A
Fantastic,
thank
you
James.
We
do
have
a
bunch
of
question.
So
I
just
wanted
to
remind
everybody
that
if
you
have
questions,
I
see
some
of
them
in
the
chat.
Please
go
over
to
the
Q&A
button
and
we
ask
them,
and
now
I've
tried
my
best
to
capture
questions
that
are
in
the
chat
that
are
not
in
the
Q&A
and
now
we're
going
to
go
through
and
I
will
facilitate
asking
these
questions
to
James
and
we'll
see
how
many
we
get
through.
A
B
So
a
great
question
I
should
have
made
that
a
bit
more
clear,
Jenkins
X
is
designed
to
work
on
any
communities
cluster,
so
anything
at
all,
wherever
it
is
whether
it's
on-premise
when
it's
on
your
laptop,
we
you
can
use
mini
cube
it.
You
can
work
literally
any
given
Izzi
cluster.
What
we
are
trying
to
do
is
we're
trying
to
configure
Jenkins
x-ray
understands
the
different
kinds
of
committees
clusters,
because
there
are
some
things
that
can
be
a
bit
wonky.
B
For
example,
we
need
we
need
various
different
things
like
DNS,
we're
currently
using
insecure,
docker
registries,
because
we
run
a
docker
registry
inside
Cuba
Nittis
and
doing
that
on
premise
is
often
kind
of
tricky
to
talk
to
it
without
HTTPS
and
all
those
kind
of
things.
So
there's
a
few
things
we
need,
but
we
we
have
run
Jenkins
X
on.
You
know
bare
metal
servers
that
all
works,
fine,
all
the
public
clouds,
it
works,
fine
it
should.
It
should,
in
theory,
work
on
any
communities
cluster.
B
You
can
find
we're
working
with
people
at
your
pivotal
and
Oracle
and
IBM
to
work
on
all
of
their
clusters.
The
aim
is
to
work
on
any
communities
cluster
as
anyone
who's
used,
cube
annuities
in
anger.
It
probably
knows
it
can
be
challenging
to
take
an
application
run
on
any
cube
aeneas
cluster,
because
there
can
be
differences
around
things
like
DNS
and
low
bounces
and
IPs
and
whatnot.
We
are
looking
to
try.
B
We
want
to
make
it
as
easy
as
possible
for
people
to
use
check-ins
X
on
any
cluster,
so
we
are
looking
to
work
with
something
like
maybe
the
hectare,
so
no
boy
project
or
something.
So
we
can
help
verify
that
the
cluster
is
set
up
for
Jenkins
next,
to
avoid
people
installing
checkers,
excellent
things,
kind
of
braking
or
whatever
to
help
people
understand
what
needs
to
happen
in
their
cluster,
but
before
they
install
it.
B
So
if
you
have
an
average
you
cost
of
it,
maybe
your
your
company
manages
or
creates
for
you
we
kind
of
want
to
help
make
that
super
simple
to
just
a
dot.
We
have
a
whole
bunch
of
plugins
by
the
winching
context
to
customize
the
install
experience
for
different
routes.
So
if
we
found
out,
for
example,
PKS
the
other
uses
a
slightly
different
place,
so
we
we
can
use
get
ups
and
how
to
configure
Jenkins
expert
different
providers
so
we're
all
set
up
to
deal
with.
B
You
know
all
these
small
differences
that
happen
in
different
kinds
of
courses.
So
please,
if
you
have
any
issues
with
your
cluster,
just
raise
an
issue
and
we
will
do
our
best
to
make
it
work
as
quick
as
we
can.
But
the
quick
answer
is
yes,
our
aim
is
to
work
on
every
Cuban
excess
cost
of
whatever
it
is
great.
A
B
A
really
good
question,
so
I
didn't
explain
this
at
all,
but
I
probably
should
have
Jackie's
ex
is,
is
kubernetes
native,
so
it
uses
humanity's
resources
for
everything.
So
let
me
switch
to
a
screen
and
show
you
so,
for
example,
in
environments
that
we
use
this
tunnel.
If
you
want
to
look
at
the
environment
sink
in
Jenkins
X,
you
can
use
cube,
CTL
and
get
environments.
B
If
I
can
spell
it
well,
you
can
use
n
for
short,
and
these
are
the
custom
resources
in
Cuban
entities,
so
we're
using
humanity's
resources
to
model
everything
in
Jenkins
extra
environments,
pipelines
releases,
they're,
all
just
custom
resources
in
communities
sense,
so
Jenkins
X,
the
JX
command
Jake's
getting
this
is
using
the
exact
same
authentication
and
certs
and
everything
the
cute
cuddle
is
using
so
cube.
Ctl
cube
cuddle
is
using
the
five
built
in
in
your
home
directory.
B
I
will
show
you
mine,
just
it's,
but
there's
a
file
in
the
home
directory
cube
config,
which
is
where
your
configuration
for
using
cube
CTL
links
right.
So
if
you
have
any
cloud
provider,
our
communities,
cluster,
there's
some
there's,
usually
a
CLI
or
something
to
log.
In
to
say,
Google
or
Amazon
0,
and
what
that
does
some
of
the
covers
is
that
sets
of
your
cute
comfy
Jenkins
X
is
using
that
exact
same
mechanism.
B
So
if
you
can
use
cube
cuddle
to
talk
to
a
cluster
Jenkins
X
we'll
use
that
to
configure
your
cluster
now,
if
you're,
using
JX
cluster
gke.
What
that
does
to
the
covers
is
that
uses
the
g-cloud
command
line
for
Google
or
Zed
for
Azure
and
soon
the
eks
one
for
Amazon
that
uses
those
command
line
tools
to
connect
to
the
cloud
so
we're
using
the
cloud
providers,
tools
to
set
up
the
clouds
and
connect
to
them,
and
after
that,
it's
just
vanilla.
B
Cuban
eighties
rest,
api,
it's
vanilla,
humanities
stuff,
so
all
of
the
api
is
were
using
the
jake's.
The
mud
line
are
just
cuban
eighties
rest
api,
so
everything
is
using
cuban
eighties
role-based
Authenticator
and
then
all
that
kind
of
stuff.
So
you
know,
if
you
really
wanted,
you
could
prevent
some
people
from
seeing
your
environments
or
pipelines
or
something
yes,
great
question,
though
any
other
questions.
B
So
one
thing
that's
common
in
the
joke
is
ecosystem
is
to
kind
of
store
all
sorts
of
stuff
in
kind
of
inside
jenkins,
inside
the
filesystem
jenkins.
What
we
kind
of
recommend
people
with
drinkin's
x
to
do
is
any
kind
of
artifact.
What
every
day
is
it
could
be
a
test
report.
It
could
be
a
an
artifact
to
a
piece
of
Yama,
a
piece
of
X
another
log
file.
B
If
things
are
more
for
reporting,
like
I,
don't
know
like
test
results
and
we're
working
on
a
plugin
to
don't
protest
result
insert
test
results
into
something
like
elasticsearch,
so
they
easy
to
query.
We
have
a
whole
bunch
of
tooling
right
now
for
exporting
metadata
from
Jenkins
X
into
things
like
elastic
search,
so
that
you
can
query
like
how
long
does
it
take
to
get
issues
fixed
or
what's
your
release,
frequency
or
what's
your
mean
time
to
recovery
and
all
those
kind
of
things?
B
A
B
So
a
lot
of
people
use
modern
repos.
What
so
we
have
something
called
build.
Packs,
I
didn't
really
talk
too
much
about
number
build
packs
are
just
little
packs
of
files
of
like
reusable
docker
files
and
Jenkins
files
and
hound,
charts
and
stuff,
and
these
bill
packs
having
them
the
docker
file
and
the
jaegers
phones
have
to
do
the
CI
CD
piece.
We
don't
yet
have
a
build
pack
for
a
mono
repo
that
can
go
through
a
mono
repo
and
determine
which
things
have
actually
changed
at
which
should
you
release
or
not?
B
B
We
mirror
occurred
into
these
individual
people's
now
I'm,
not
necessarily
saying
this
is
an
awesome
thing
to
all
do,
but
if
he
did
use
that
tool,
then
the
mono
repo
gets
replicated
in
two
separate
repos
and
what
that
means
is
we
can
then
just
use
normal
git
commits
to
know
when
we
should
release
something.
The
problem
with
modern
repos
is
you
need
some
magic
tool,
so
imagine
you've
got
20
micro
services
in
a
mono
repo.
B
Every
commit
is
going
to
run
a
release
and
the
challenge
with
mon
repos
is
knowing
which
pieces
you
really
need
to
release
in
that
mono
repo.
So
you
need
to
use
a
basil
or
some
kind
of
magic
thing
that
is
going
to
look
at
the
diff
in
getting
go,
which
things
am
I
really
gonna
release.
If
you
can
put
that
in
the
jenkins
file-
and
you
could
just
use
it,
but
we
don't
have
one
of
those
things
yet,
please
contribute
one.
B
If
you
have
some
tool,
that
does
that,
so
we
kind
of
further
to
be
separate,
git
repos.
So
if
you
want,
you
could
mirror
your
mono
repo
into
separate
repos,
so
that
if
you
do
change
one
of
the
micro
services,
only
that
one
micro
service
will
get
a
new
lease
and
be
promoted.
So
that's
our
best
untrim
on
the
repos
right
now
today.
We're
weird
we're
not
awesome
at
mother
repos,
but
we
could
certainly
improve
on
all
the
time.
To
be
honest,
I
prefer
separately
pearls
from
micro
services.
B
It
feels
a
bit
more
micro,
so
easy.
It's
just
easier
for
people
to
see
when
individual
micro
services
really
do
get
change
again.
I'm
on
the
repo
with
20
things
and
there's
lots
of
commits
it's
just
easier
for
humans
to
just
see
what
changes,
but
we'll
certainly
help
you.
You
lose
one
of
your
phone's
been
taking
sex.
If
you
wanna,
join
slack
good
question.
A
B
Should
have
I
should
have
talked
about
yeah
I
should
have
taught
whether
we
have
a
whole
range
of
get
plugins,
so
the
aim
is
to
support
every
get
provided,
but
anybody
could
possibly
wish
to
use
eventually
today,
right
now
today
we
support
github
and
github
Enterprise
awesomely
and
big
cookie.
Big
bitbucket
cloud
is
working
great
and
we
have.
We
have
integrations
for
gitlab
get
and
bookies
server,
but
they're
in
different
stages
of
maturity.
I
think,
probably
the
next
one
that
will
be
rock
solid
is
bitbucket
server.
B
So
the
to
the
box
solutions
that
work
for
on-premise,
git
repositories,
I'll,
probably
get
reproduce.
We
get
a
bunch
of
writers
and
Bukkit
server.
We
are
working
on
gate
lab
and
get
ear.
Get
ear
is
super
close
there's
just
get
ears
like
a
go
based
open
source,
git
repository
server,
but
I
personally
am
a
huge
fan
of
just
because
it's
already
go
and
go
I
said
cool
and
it's
kind
of
tiny.
It's
like
a
really
tiny
double
container.
It
starts
with
really
fast
and
lean
mean
and
feels
very
much
like
it.
B
Oh
I
love
github
by
the
way
I
get
to
be
submitting
and
I'm
super
happy
that
Mike's
at
the
bottom.
But
yes,
I
really
is
to
support.
You
know,
get
lab
and
everything
people
are
using.
It's
just
taken
this
a
while
to
add
all
the
deep
integration
for
all
the
gear
providers,
because
it's
not
just
about
cloning,
the
git
repository.
What
we
need
is
pull
requests
and
the
feedback.
You
know
when
a
pull
request
triggers.
B
We
need
the
feedback
from
the
pipeline
to
go
back
into
the
git
repository
so
that
we
know
the
poor
aggressor
green
and
we
can
manage
the
stuff.
So
it's
thinking
there's
a
little
while
to
get
full
support
for
all
the
good
finders,
but
it's
definitely
on
the
roadmap.
We're
trying
to
go
as
quick
as
we
can
to
someone
more,
but
yeah,
github,
Enterprise
and
github
is
working
brilliantly.
Bitbucket
cloud
is
working
brilliantly,
bitbucket
servers,
close
I,
think
get
ear
is
really
really
close
enough
for
one
bug
and
then
get
labs.
A
B
A
brilliant
question:
I
love
it.
So
there's
a
lot
of
demand
for
multiple
clusters.
This
is
kind
of
a
common
thing
in
cuban
is
people
kind
of
say
how
many
clusters
sure
they
have,
and
there
isn't
a
really
good
answer
of
how
many
processors
now
he's
kind
of
one
of
those
things
that,
for
example,
you
could
have
a
cluster
per
environment,
so
one
course
of
a
development
one
for
testing
or
the
staging.
B
If
you've
got
ten
teams
in
your
organization,
you
could
have
ten
clusters
for
development
templates
with
staging
intention,
or
you
might
want
to
group
some
things
together
and
share
a
customer
one
about
so
there's
there's
there's
a
lot
of
different
ideas
of
what's
the
right
number
just
to
have,
but
absolutely
Jackie's
ex
can
work
on
multiple
clusters.
The
one
thing
we
don't
yet
do
is
we
don't
automatically
set
up
multiple
clusters
for
you,
so
you
literally
have
to
set
up
a
stating
cluster
prediction
cluster
and
you
kind
of
have
to
do
that
bit
by
hand.
B
So
we
hope
to
have
a
demo
soon
edge
of
spinning
up,
you
know,
say
three
clusters
for
their
staging
production
on
any
provider,
and
so
you
could
use
an
amazon
for
development
cuban
it
a
gke
for
testing
and
maybe
eks
production
or
whatever,
or
you
might
want
to
production
clusters,
one
on
Amazon
than
one
than
AWS
or
whatever.
So
the
quick
question
is
yes,
we
absolutely
support
multiple
clusters.
We
just
need
to
do
a
bit
more
work
on
the
tooling
to
automate
setting
it
up.
B
I
mean
the
trick
is
just
setting
out,
though
the
beauty
of
get
ops
is
detox,
isolates
the
development
pipelines
from
the
production
and
staging
environments
right.
So
one
of
the
issues
with
CD
is
sometimes
you
don't
want
your
development
environment
to
have
like
cluster
admin
on
production
right,
yeah
security
folks
would
probably
go
be
quite
scared
if
you're,
like
Jenkins
server,
had
cluster
admin
and
production.
B
So
what
the
lovely
thing
about
Multi
cluster
we've
get
ups
is
get
ups
means
the
development
cluster
is
just
literally
submitting
pull
requests
and
the
only
thing
that
runs
a
job
to
apply
those
changes
runs
in
the
staging
environment
or
the
peak
environment.
So
you
don't
have
to
expose.
You
know
the
production
environment
to
the
public
with
secrets
and
whatnot.
You
can
keep
the
job.
That
applies
the
latest
Hampshire's
in
that
cluster,
so
it
works
beautifully
from
a
security
standpoint.
So
quick
answer
is
yes.
B
B
Around
they've
got
a
new
project,
Gil
gimble
as
a
name,
that's
trying
to
help
abstract
the
differences
of
their
multiple
clusters,
because
once
you
have
say
I
don't
know
ten
clusters,
it
can
get
a
little
bit
kind
of
complicated,
switching
between
them
all
and
all
that
kind
of
stuff,
so
I
think
in
there
in
the
ecosystem,
we'll
have
more
more
tools
to
help
work
with
many
clusters
at
the
same
time,
so
we're
hoping
to
use
Hector's
gimbal
to
help
on
that
front.
But
yes,
the
countries.
A
Okay,
that
that's
we're
at
time,
I
wanted
to
thank
James
for
his
time.
That
was
incredibly
informative
for
me,
and
a
lot
of
the
feedback
in
questions
in
the
chat
have
been
incredibly
positive.
So
thank
you
all
for
joining
I'm.
Sorry,
we
couldn't
get
through
all
the
questions.
There
are
21
open
questions
still
I
encourage
you.
James
has
shared
different
ways
in
the
community
that
you
can
provide
feedback
and
ask
those
questions,
but
I
want.