►
From YouTube: VMware Cloud PKS Lightboard - Smart Clusters
Description
Watch this lightboard video to learn about VMware Cloud PKS and Smart Clusters
A
Hello,
everybody,
my
name
is
Bosque
szabla
I'm,
a
Technical
Marketing
Manager
at
VMware,
focusing
on
cloud
native
applications,
and
today
we
are
going
to
take
a
look
at
V,
America,
P
chaos
and
smart
clusters.
We
M
o'clock
p
chaos
as
a
service
from
VMware
that
provides
managed
communities
cluster
on
a
public
cloud
provider
of
your
choice.
It
does
so
by
giving
you
a
simple
control
plane
that
you
can
use
to
manage
multiple
kubernetes
clusters
across
multiple
clouds.
A
Now
very
copy
chaos
makes
it
super
simple,
both
from
a
deaf
perspective
and
an
ops
perspective
to
get
a
communities
cluster
up
and
operational.
Now
from
a
developer
perspective,
you
know,
rema
copy
chaos
will
eliminate
the
need
to
determine
the
capacity
what
kind
of
worker
nodes
master
nodes.
How
many
do
I
need
what
instance
types
do
I
need
for
a
particular
cloud
to
support
my
cluster.
At
the
same
time,
from
an
operational
perspective,
it
also
takes
care
of
providing
the
back-end
infrastructure
needed
to
support
a
highly
available
communities
cluster.
A
It
also
takes
care
of
the
back-end
infrastructure
like
networking
needed.
I
know
how
am
I
ports
connected
together.
What
networks
do
I
need
to
wire
things
together
up
VMware
copy
chaos
takes
care
of
all
of
this,
so
in
short,
getting
a
kubernetes
cluster
of
it.
Vm
a
cloud
maker,
P
chaos
super
simple
and
you
typically
get
up.
You
know
an
active
kubernetes
cluster
are
running
within,
maybe
a
couple
of
minutes
of
triggering
the
cluster
create.
So
let's
see
how
this
works.
A
Now,
let's
say
you
have
a
dev
team,
so
that
needs
a
cuban
ids
cluster
and
on
day
0.
They
are
requesting
for
a
cluster
or
day
one,
and
they
talk
to
the
McLeod
PKS,
Maya,
API
or
CLI,
or
even
the
UI
and
the
they
are
basically
giving
two
things.
One
is
the
public
cloud
provider
that
they
want
the
cluster
to
reside
in
to
the
region
within
the
public
cloud
provider.
A
A
It
will
create
a
avi
PC
or
it
will
use
a
specific
number
of
ec2
instances,
bind
it
to
this
V
PC
to
ensure
you
get
pod
networks
and
master
node
and
working
on
networks
up
and
ready,
and
then
within
that
particular
region.
Vm
record
PKS
will
take
advantage
of
multiple
availability
zones.
So
let's
say
this
is
a
z1
e
z2
and
a
z3,
and
it
will
create
a
master
node
for
that
cluster
in
each
and
every
zone.
A
It
will
deploy
these
three
different
masters
in
agent
availability
zone
to
make
sure
that
the
cluster
is
always
highly
available.
Now
all
these
three
masters
are
running
in
active
active
mode.
At
this
point,
once
the
initial
set
of
master
nodes
for
the
cluster
have
been
created,
remember
cloud
PKS
will
then
provide
back
the
k8
api
to
talk
to
so
this
way
the
tape
teams
can
start
talking
directly
to
the
cluster,
using
a
CLI
like
cube,
control
or
even
through
the
UI,
that's
provided
by
IBM
ARP
chaos
or
to
the
kubernetes
dashboard.
A
A
Now,
as
more
and
more
spots
start
coming
in
of
VMware
cloud,
PKS
will
ensure
that
their
corresponding
worker
nodes
created
to
ensure
that
there
is
no
pod
waiting
to
be
scheduled
for
compute
capacity.
So
it's
automatically
scaling
the
cluster
out
as
workloads
start
coming
in
and
more
capacity
is
needed
now
it
will
do
so
even
to
scale
down.
So
let's
say
there
was
a
batch
operation
that
ran
and
it
created
a
lot
of
pods
within
this
cluster,
and
now
the
batch
operation
is
done
and
the
pods
are
being
deleted.
A
What
we
Umbra
copy
chaos
is
doing
at
that
point.
Is
it's
constantly
monitoring
the
number
of
pods
scheduled
versus
the
number
of
worker
nodes
in
that
cluster
and
we'll
start
to
decommission
any
worker
nodes
that
are
not
needed
any
pods
on
that
worker
nodes
will
then
be
rescheduled
to
an
existing
worker
node.
So
that
way,
the
cluster
size
is
always
optimal,
based
on
the
workload
that
the
cluster
is
running.
So
this
is
kind
of
day
one
auto-scaling.