►
Description
Join us for Kubernetes Forums Seoul, Sydney, Bengaluru and Delhi - learn more at kubecon.io
Don't miss KubeCon + CloudNativeCon 2020 events in Amsterdam March 30 - April 2, Shanghai July 28-30 and Boston November 17-20! Learn more at kubecon.io. The conference features presentations from developers and end users of Kubernetes, Prometheus, Envoy, and all of the other CNCF-hosted projects
Intro: SIG Cluster Lifecycle - Di Xu, Ant Financial & Alexander Kanevskiy, Intel
Sig Cluster-Lifecycle Intro The Cluster Lifecycle SIG is the Special Interest Group that focuses on cluster deployments and upgrades. Our SIG is trying to improve user experiences on bootstrapping a minimum viable Kubernetes cluster that conforms to best practices. With our primarily installer tool, kubeadm, a streamlined installing and upgrading process can be well managed. We've recently introduced a new Kubernetes object, called Cluster API to bring declarative, Kubernetes-style APIs to cluster creation, configuration, and management. In this introduction session, we will present our SIG's mission statements, review recent updates, and discuss our roadmaps. Some new lifecycle projects are introduced as well. You are very welcome to join in and contribute to our SIG.
https://sched.co/QGOE
A
B
C
B
Is
the
second
time
we
talk
about
the
c-class
lifecycle
in
Kumiko
and
China
and
I'm
dizzy
from
China
across
Chinese
location,
right
I,
moved
from
on
financial
I
am
a
top
50
code
contributor
in
the
whole
community
and
McGahan
delay
is
a
DVD
X.
So
if
you've
got
any
questions
or
problems,
don't
get
hurt,
add
to
contact
me.
We
can
have
our
slack.
A
And
arm
Alexander,
can
you
have
ski
I
I
work
for
Intel
I'm
based
in
Finland,
my
github
handle
is
cut
so
again.
Also,
if
you
have
questions
please
contact
I
am
one
of
the
reviewers
of
cube
idiom,
which
is
part
of
the
cluster
lifecycle
and
I,
have
also
interested
in
several
our
areas
like
node
and
few
hours.
A
All
right,
so
we
see
cluster
lifecycle
so
who
we
are
and
what
actually
we
are
doing.
The
sequester
lifecycle
is
one
of
the
biggest,
if
not
the
biggest,
seek
of
a
kubernetes
community
like
we
have
600
more
than
600
people
on
the
mailing
list.
We
have
more
than
2,000
people
on
we're,
like
means
to
our
slack
channel,
sequester
lifecycle,
and
if
you
look
at
all
our
our
slug
channels,
like
you
baby
airman
hours,
you
will
see
a
lot
more
people
than
web.
A
We
have
incredible
amount
of
companies
like
it's
literally,
a
dozens
of
companies
who
contribute.
We
have
contributors
all
around
the
world.
Probably
we
don't
have
anyone
from
Antarctica
yet,
but
who
knows?
Maybe
we
have
quite
significant
amount
of
commits
counts
of
pull
requests
per
cycle.
So
it's
like
an
order
of
thousands
for
our
projects
and
C
cluster
life
cycle
is
not
like
just
one
project.
We
have
a
collection,
it's
an
umbrella
for
like
more
than
15
projects
in
a
world
with
smaller
tools.
B
Was
our
concrete
statement
which
actually
has
many
pieces
into
it?
So
we
are
doing
this
because
we
don't
want
to
ended
up
in
that
situation.
The
only
way
your
uncle
Teddy's
it
so
complex.
You
might
spend
too
much
time
to
install
a
whole
cluster
or
spend
$1000
or
some
company
to
her
people
Stratocaster.
So
we
want
to
try
to
reduce
the
complexity.
D
B
And
actually
the
Cabana
means
try
to
to
simplify
the
prescribing
steps
that
you
try
to
put
strength,
Acosta
and
trying
to
intend
you
to
be
a
building
block,
that
other
tools
can
rely
or
reuse
some
modules.
That
could
be
envious,
so
how
people
escape
your
own
class,
the
providers
and
clay
many
companies
are
using
copy.
They
mean
to
to
develop
armed
enemy,
to
develop
their
own
with
travel
tools.
A
So
to
look
in
a
bit
graphical
way,
what
we
actually
do
so
with
cube
EDM,
one
of
our
major
and
like
one
of
the
first
building
blocks,
what
zip
cluster
lifecycle
created.
So
it
has
a
scope
like
us,
as
we
are
doing
the
unique
style
principle
of
define
and
what
tools
is
like
do
want
to
do
is
to
make
this
tool
to
do.
One
thing
we'll
one
thing
which
is
good
and
be
a
building
block
to
a
bigger
solution.
A
So
what
cube
medium
is
used
by
other
components
of
our
projects
and
when,
as
soon
as
we
start
to
evolve,
cube
ATM,
we
found
out
our
pieces.
Would
we
consider
it
to
be
like
reusable
blocks
across
different
cluster
deployment
and
configuration
solutions?
So
what
we
wind
up
of
creating
were
etc'
EGCG
ATM
like
well,
what
tool,
which
is
specifically
like
special
operator
with
help
which
helps
to
deploy
a
high
availability,
CD
cluster
for
kubernetes
deployments
and
when
by
cluster
add-ons,
we
have
inside
cube
ADM
where
two
main
add
ons.
A
So,
let's,
how
is
component
config
working
group
organizers
and
how
we
started
to
improve
our
components,
beside
our
scope
and
when
cluster
API
again
it's
a
piece
of
building
blocks.
What
operate
like
what
cloud
cloud
providers
or
different
vendor
companies
can
create
a
specific
implementation,
four-way
hardware
for
a
infrastructure,
but
in
all
it's
a
common
goal
to
have
a
unified
API
level,
how
to
manage
requesters
towards,
like
higher
level
M,
to
end
solutions.
And
if
you
look
at
this,
which
is
wherever
C
cluster
lifecycle
is
like.
Several
projects
heading
towards
one
goal.
Yeah.
B
Actually
include
many
cookies,
our
first
cluster
tools
that
I
use
when
I
get
come,
nannies
and
minnekopa
will
try
to
post
wrap
several
Ultima
changes
on
your
laptop.
You
know
servers
and
try
to
install
that
all
the
packages
to
the
virtual
machines,
so
you
have
a
standalone
connected
cluster
in
your
laptop
actually
meeting.
Chrome
is
not
working
very
well
in
China,
since
the
world
pull
all
the
registry
from
the
trdi
all
right.
It's
not
working
China
and
the
cookie
strays
are
kind
of.
A
B
B
Well,
the
user
appearance
is
very
simple
and
the
customer
is
really
the
most
secure
as
everybody
now.
Security
is
a
very
tough,
not
I,
can't
not
a
lie
right.
So
actually
we
are
going
to
improve
the
security
in
the
next
few
release.
So
you
are
getting
interested,
please
cut
involved
and
we
welcome
all
y'all
contributors
informants.
B
Actually,
the
commune
scope
is
very
limited.
Any
tiny
to
be
a
beauty
plug.
So
if
you
are
trying
to
the
community,
will
only
try
to
thanks
to
the
APS
server
when
seized
up
and-
and
you
are
going
to
get
a
piece
of
rabbit
right,
so
lots
of
certificates
and
tokens,
so
the
local
local
folder
actually
come
in.
We
won't
randomly
attach
into
your
chains
and
do
something
else,
so
it
is
reserved
to
those
could
be
and
be
in
your
production
or
your.
You
know,
cloud
environments.
B
Actually,
could
we
enemy
doesn't
care
about
how
cool
it
is
drawing?
Actually
you
can
draw
your
ability
in
continent
rice
or
using
by
Sydney
after
the
community
is.
Is
there
so?
The
next
thing
you
do
is
to
install
the
feeding,
bio,
yum,
repo
y'all
or
your
apt
apt
game,
and
the
company
will
do
all
the
resources
for
you
and
the
country
plane
is
on
y'all
or
not,
and
if
you
have
several
ethnicity,
it's
out
there.
B
B
Actually,
all
the
components
I
was
a
fascist,
Inka,
vietnamese's
composable.
That
means
you
don't
need
to
draw
all
the
wrong
cookie
and
mutually
to
all
the
sins
for
you.
You
can
just
use
Wow.
All
two
phases
because,
like
you
can
try
just
using
could
be
mean
to
generate
all
your
certificates
for
you
and
you,
then
you
use
other
tools
to
bootstrap
your
cluster
and
all
this
all
the
senses.
We
are
split
into
smaller
building
blocks.
So
can
your
reals
in
other
shell
scripts
is
okay
or
your
coding
or
programming
code
code?
So
it's
already
preferred.
A
A
Underscored
where
what
we
have
is
what
like,
as
soon
as
you
always
installed
as
soon
as
the
couplet
is
running,
you
run
with
a
cube,
ADM
and
cube
a
DML
provision.
You
control
node,
like
control,
control
components,
and
it
will
be
able
to
join
the
worker
node
to
your
control
plane
and
then
like
on
my
upper
part,
like
all
kind
of
your
don'ts,
all
kind
of
your
additional
preferences
like
dashboards
and
so
on.
It's
out
of
a
scope
of
cube
idiom.
A
So
one
tool
does
what
specific
tasks
doesn't
know
a
specific
area,
so
unless
it's
with
key
difference,
for
example,
for
the
tools
like
with
copes
for
hours,
so
those
tools
are
covering
the
whole
area,
so
we're
taking
care
of,
for
example,
a
creation
with
virtual
machines
on
where
Amazon
Cloud,
when
running
with
you
buddy
m2
provision,
something
and
when
configure
our
things
like
load,
balancers
and
etc,
etc.
For
your
cluster.
So
here's
the
key
differences
between
were
like
tool
concentrated
on
one
task
and
end-to-end
solution.
B
13
the
G
release
and
the
the
general
feedback
we
got
is
the
user
ability
is
very
good.
You
pretty
good
people
like
to
use
compared
me
to
Harvey,
describe
their
clusters
and
the
common
feedback
we
began
is
the
high
value
military.
So
this
is
the
leg
of
the
koobideh
me
doing
so
in
the
next
release.
We
just
improve
the
high
value
military.
A
D
A
And
if
you
want
to
know
details
about
how
its
implemented,
what
challenges
we
are
I
would
highly
recommend
you
to
actually
watch
replay
of
Yukon
Europe
session
done
by
Luba
Luba
here
and
Fabrizio
about
Cuba
I
am
deep
dive,
so
Fabrizio
was
directly
talking
about
what
like
technical
details,
what
what
steps
are
challenging
and
why
it
took
us
quite
some
time
to
do
high
availability
right,
yes,
and
one
more
thing.
He
was
probably
already
noticed
in
the
previous
things.
One
of
the
previous
slides
is.
We
have
nowadays
official
log
of
cubanÃa
motility.
A
A
Now
it's
give
a
DM
official
over
alright,
but
let's
switch
gears
a
bit
so
you'll
be
the
MS
fine.
It
helps
you
to
create,
like
single
cluster,
multiple
nodes,
multiple
worker,
multiple
master
nodes,
so
most
of
customers
actually
run
not
a
single
cluster.
We
usually
have
from
like
five
to
hundreds
or
even
some
customers
are
having,
like
thousands
of
the
clusters
running.
So
the
question
starts
to
have
okay,
I
created
one
cluster.
How
I
do
it
the
same
way
like
for
all
of
our
my
clusters?
A
How
I
can
do
was
same
over
were
different
cloud
providers,
so,
in
the
end
of
a
day,
I
don't
really
care
how
kubernetes
is
deployed.
What
I
expect
like
I
said-
let's
say,
application
developer
is
what
my
application
should
run
on.
Kubernetes,
regardless,
where
with
kubernetes
cluster
is
actually
hosted,
would
be
like
Amazon
or
or
some
other
provider,
or
on
my
own
premise
solution.
So
what
cluster
api
tries
to
solve
it?
So
requester
api
is
not
a
single
program.
It's
a
collection
of
vendor
specific
operators
and
this
vendor
specific
operators.
A
A
Here's
the
example
of
what
you
can
see
about
like
llamó
of
crv
custom
resource
definition
for
per
cluster
api.
So
if
you
are
familiar
with
kubernetes,
you
know
you
can
find
a
lot
of
similar
concepts.
What
cluster
API
have
compared
to
to
your
like
normal
application
developer,
so
you
have
port
right.
If
you
forget
about
like
vertical
port
after
scalar
port
is
something
immutable
in
cluster
API.
We
have
the
same
instance
code
machine,
so
machine
once
created.
It's
not
changed
anymore.
A
So
if
you
want
to
upgrade,
you
delete
one
machine
and
you
create
another
machine
with
newer
properties
when
one
layer
up
so
reports
you
have
replica
set,
we
have
machine
sets.
So
it's
exactly
set
of
machine
with
exactly
the
same
properties
and
we
have
a
conversion
controller
which
actually
creates
those
machines
based
on
a
common
template
spec
and
when.
A
We
have
machine
deployment
and
deployment
controller,
helps
to
do
appropriations
like
scaling
it
up
and
down
how
to
do
the
rolling
cup
grades
and
practically
like
similar
approach.
How
you
manage
your
application
now
it's
applicable
to
how
you
manage
your
infrastructure,
your
infrastructure
become
describe
it
in
a
way.
What,
like
you,
have
a
Yama
file,
you
have
it
and
get
it
completely
repeatable.
You
have
a
detox
for
your
whole
infrastructure
running,
but
what's
like,
in
short,
what
cluster
API
is
about
and
similar
to
pupa
I
am
deep
dive.
A
B
Oh
Alex
talk
about
is
from
a
high-class
level,
but
actually
we
have
a
small
problem
too.
So
how
many
reflects
do
you
remember
to
start
by
just
over
how
many
steps
do
you
want
to?
You
have
to
do
to
create
your
sub
host
data
again,
so
actually
we
are
doing
to
make
them
to
be
component
based
what
it
doesn't
mean.
It
means
all
your
components:
can
you
reuse
the
same
repo
to
build
up
your
own
services
and
actually
know
the
most
distinguished
feature?
All
communities
is
the
kinetic
api's.
So
what
about
the
the
component
context?
B
All
the
component,
complex,
are
config,
trees,
flex,
flex
with
flex,
and
the
flags
were
be
deprecated
in
the
next
few
releases.
So
you
need
to
modify
your
flex
and
it
is
incompatible
for
regard
and
mini
Stratos
to
manage
your
clusters
because
it's
not
versioned
it's
in
not
checked.
All
your
flags
are
attached
to
send
over
to
the
arguments
of
the
shell
scripts.
It
is
not
well
documented.
B
B
Because
configs
such
as
control
manager,
scheduler
and
you
can
query,
could
be
proxy
if
you
notice
the
components
that
start
that
stirred
up
by
the
could
be
me
have
already
used
some
of
the
component.
Config
sanitizer
could
be
proxy,
so
only
well
all
to
flex
and
needed
no
more
acts,
no
more
arguments,
Hritik
tested
relief
for
the
administrators,
so
so
keep
version
instead
or
stable
Yama
file
because
it
very
well
compatible
with
older
versions.
So
what
you're
like?
B
A
B
Fire
of
Jason
file
and
there's
a
scheduler
component
or
the
controller
manager
component
we're
just
a
render
the
Yama
file
justified
to
his
own
config
AP
attacks.
So
it's
very
convenient
convenient
for
internal
internal
controllers
and
internal
coats
to
to
share
the
same
same
settings
and
we
are
still
involving
the
component.
If,
yes,
so,
if
you
get
interested,
please
come
to
harm
us
and.
A
Just
to
complement
what
did
you
say
it
about
what
come
come
component
config?
The
reason
why
we
have
it
is
what
it's
not
only
simplification
of
the
flux,
it's
more
about
how
to
manage
your
cluster
more
dynamically.
So
it's
often
with
scenario
what
you
create
the
cluster.
We
created
a
specific
set
of
options
and
when,
like
few
days,
how
later
or
maybe
even
few
hours
later,
you
realize
what
oh
I
actually
need
to
change
what
flag
and
this
current
set
of
tools
it's
really
hard
to
to
change
with.
A
B
A
D
D
A
Labor
work
so,
for
example,
like
webcomponent
can
pick
it's
not
not
hard
thing
to
do.
It's
not
like
rocket
science,
it's
just
like
big
amount
of
work,
which
needs
to
be
done
across
all
the
components
of
kubernetes.
So,
speaking
of
all
the
contributions
we
already
mentioned,
we
have
hundreds
of
already
existing
contributors
and
we
see
cluster
lifecycle,
actually
quite
good
of
value
of
our
contributors.
So
every
cycle
we
like
sig
leads.
We
are
talking
about
whom
we
want
to
promote.
A
So
if
you
want
to
get
with
kubernetes
membership,
if
you
want
to
get
like
the
review
or
approval
status,
contribute
and
gain
your
past
work,
a
kubernetes
community,
and
unlike
many
of
our
six
we
see
cluster
life
cycle
is
not
a
us-centric.
We
have
multiple
contributors
in
the
europe
time
zone.
So,
if
you're
talking
about
here
about
China,
we
have
a
quite
good
time
overlap
where,
where
people
from
Europe
can
help
with
people
from
Asia
time
zone
and
when
handover
it
to
u.s.
time
zones.
B
Want
to
hold
our
bi-weekly
big
meeting
and
the
is
China
very
friendly
time.
So,
if
you
get
interested
just
tell
me
and
it
give
me
some
feedback,
we
are
still
collecting
the
appropriate
that
time
slot
for
Chinese
users
so
and
I
will
be
the
chair
of
the
Chinese
Chinese
sig
sig
bag
weekly
meeting,
and
we.
B
C
A
Already
in
GA
states
of
general
ability
so
means
like
stable
component,
but
we
have
a
lot
of
components
which
we
are
still
in,
like
the
definition
state.
So
like
a
good
example,
is
cluster
add-ons
like
what
is
clustered
on
so
what
we
really
need
to
have,
what
we
can
do
it
late
at
what
can
be
considered
as
application?
How
we
actually
do
cluster
Adhan
management
like
when.
C
A
Is
year,
then,
we
know
what
we
want.
We,
we
have
some
proof
of
concept
code,
which
was
distilled
from
a
previous
experiences,
but
again
to
make
it
real
to
make
it
real
GA
level,
stable
level.
It
will
require
a
lot
of
work.
A
cluster
API
we
reached
finally
like
where
first
alpha
stage,
where
every
understand
tech,
okay,
what
we
are
doing,
the
first
thing,
but
together
to
GA
level,
to
get
it
like
as
a
stable
API.
A
It
will
take
a
lot
of
work
and
if
you
look
at
our
small
components
like
component
config,
it's
a
whole
lot
of
work
to
do
and
for
provision
ours
again
like
something
is
already
stable,
because
many
people
contribute
that,
and
some
of
them
is
in
a
very
beginning.
So
you
can
find
a
project
based
on
your
interests.
You
can
find
the
project
based
on
your
skills.
You
can
grab
like
set
of
easy
tasks
to
help
a
lot
of
complex
problems
where
you
want
to
get
like
deep
dive
and
more
wood
in
in
rows.
B
So
if
you
are
trying
to
know
all
what
we
have
done
before,
you
can
just
visit
youtube
to
see
all
our
recordings
there,
and
we
also
have
a
lot
of
select
channels.
You
can
just
the
peanuts
or
ask
some
questions
in
the
slack
Channel
and
just
to
check
out
the
documents
list
here.
It
were
just
know
better
about
what
we
are
doing
right
now.
A
D
A
D
D
A
D
D
C
B
Actually,
we
are
doing
the
combined
since
1.15.
Actually
I
have
done
a
lot
of
work
in
new
schedule
component
config.
We
are
trying
to
try
out
transit
ride
to
move
all
the
flags
of
our
component
into
a
schema,
APR
types.
So
actually
there
is
a
lot
of
testing
testing
works
and
incompatible
works
so
for
the
holodecks
to
the
field.
Release
so
it'll
be
measuring
well
well
to
release.
So
if
you
get
interested
just
to
come
to
harvest
to
improve
more
stable.
D
Yeah
aya
kazakh
resident
here
so
when
you
put
describe
a
cluster,
so
customer
wanted
ready
to
use
class
and
not
just
the
of
you
knows
the
Masters
knows
we
want
the
storage
back-end
to
be
radios.
We
should
have
the
storage
class,
the
one
the
for
the
backend.
We
should
have
the
networking
while
defined
isolated
from
other
clusters.
So
with
current
solution,
the
goal
is
to
put
stripe
a
ready
to
use
cluster
or,
just
you
know,
a
scratch.
B
A
But
if
what
working
yes
I
understand
the
concern,
so
what
your
bit
am
responsible
for
for
just
bringing
a
basic
cluster,
but
what
you
are
looking
for
is
probably
something
would
like
work.
Higher-Level
m2m
tools
are
looking
like,
like
mini
cube.
Does
what
for
a
local
cluster
were
cops
or
cube
spray?
Does
it
for
like
multi
machine
clusters?
So
look
at
with
high-level
tools
which
is
built
on
top
of
AP
is
what
sequester
lifecycle
is
maintaining,
and
that's
probably
what
you're
looking
for.
D
B
A
And,
for
example,
I
saw
I
saw
with
different
tools,
which
was
built
on
top
of
a
cube
spray,
so
we
could
spray
this
collection
of
playbook,
so
you
have
a
rolls
for
most
of
the
phases,
but
nothing
prevents
you
to
reuse.
Those
rules
into
your
like
higher
level
deployment
tool
with
plug
adhere
specifics.