►
From YouTube: Summit 2022: Kubevirt and Harverster
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
A
All
right,
everybody
we're
back
on
schedule.
Welcome
back
introducing
our
next
speaker.
This
is
nuno
de
carmel.
I'm
sorry,
that's
the
best.
My
american
trent
accent
will
allow
me
to
say
her
name.
I
hope
it
all
right,
and
I
will
turn
it
over
to
you,
sir,
and
this
is
for
kubert
and
her.
B
Yes,
oh
there
we
go
sorry
all
right,
so
hi
everyone
nice
to
meet
you
all,
I'm
nunodo
carmo.
I
work
for
suse,
rancher
and
I'll
be
talking
about
a
specific
topic
that
is
close
to
my
heart,
which
is
windows,
but
how
to
run
actually
windows
vm
on
arvister.
B
So
let
me
just
get
back
here
so
here
I
am
like
I
said
I.
I
worked
for
susie
rancher
since
few
months
now
and
as
a
tech
writer.
Actually
so
before
all
this
demo
will
be
very
user-centric.
I
don't
know
really
all
the
details.
B
However,
I
know
my
details
when
I
will
run
my
workload.
I'm
also
a
cncf
ambassador,
microsoft,
mvp
and
docker
captain
around
wsl,
mainly
so
cubert
is
quite
new
for
me.
I
discovered
it
thanks
to
actually
being
the
tech
writer
for
harvester
now
and
I
really
liked
it.
So
I
wanted
to
try
out
a
lot
of
the
lot.
Sorry,
a
lot
of
stuff
and
well
here
I
am,
you-
can
find
me
normally
on
the
rancher
user
slack
the
kubernetes
like
on
cubert
I'm
on
the
channel.
B
My
twitter
is
normally
where
I
have
the
most-
let's
say:
community
communication,
it's
at
unix
tech
and
I
have
a
blog,
it's
wsl.dev
again
around
wsl
all
the
time,
but
we
are
here
for
cubert,
it's
the
cuber
submits.
So
here
is
a
story.
So
that's
how
I
will
present
it,
like.
I
said:
I'm
a
user
at
the
end
of
the
day.
B
So
all
the
technical
talks
that
were
before
me
and
after
me
might
be
really
into
details
here,
it's
more
like
how
someone
like
me
can
already
use
kubert
to
run
virtualized
environments
on
top
of
kubernetes.
B
So
I
started
to
write
for
arvister
like
I
said,
so:
it's
not
a
sales
pitch
at
all
right,
I'm
again,
like
I'm
a
detect
writer,
so
I
started
to
write
for
harvester
and
I
find
out
that
it
was
really
really
easy
to
just
spin
off
like
a
vm.
We
will
do
that
just
after
this
is
also
very
democentric,
and
the
thing
is
like:
while
it
was
easy
at
let's
say,
launching
the
the
vm
itself,
then
I
went
through
some.
B
Let's
say
settings,
that's
guided
me
through
the
rabbit
hole
and
I
fell
into
one.
So
the
goal
was
simple:
it
was
to
run
windows
vm
on
convert,
okay
thanks
to
arvister
in
this
case.
First,
like
it's,
the
second
point.
So
writing.
While
writing
the
docs.
B
I
was
learning
about
it
and
I
learned
a
lot
about
like
what
kubert
was
doing
behind
the
scenes,
how
it
was
like
communicating
with
the
with
kubernetes,
with
the
objects
of
kubernetes
and
how
potentially
we
could
even
announce
it
thanks
to
the
virtualization
capabilities
that
kubert
is
actually
leveraging
also
on
the
sorry
at
the
os
level.
So
I'm
a
windows
guy
right.
I
know
some
linux
again,
my
wsl
part,
but
I'm
not
into
really
like
the
kernel
stuff
and,
let's
say
the
visualization
on
linux
as
we
speak.
B
B
So,
like
every
story,
there's
a
preface
my
assumption
here
is
really
that
you
all
have
interesting
keywords
or
at
least
use
it
way
better
than
I
do.
You
might
want
to
run
or
see
how
to
run
some
windows
capabili,
let's
say
vms,
on,
convert,
optimized
or
not
for
now
or
you
just
really
care,
don't
care
and
you
are
a
friend
of
mine.
So
I
love
you
too.
B
What
will
be
covered
during
the
demo?
That's
important.
I
will
show
like
a
running
windows
vm
on
top
of
cubert.
With
already
the
settings
applied,
I
might
go
through
some
additional
settings
live
so
that
will
be
the
live
part.
Let's
say
where
I
might
add:
maybe
a
disk,
I
don't
know
we
will
see
it.
B
How
it
goes
then,
on
again
it
will
be
against
slides,
but
it
will
be
more
like
what
was
the
ruby
tool
for
me
was
really
the
kuber
top
level
api
objects
and
that's
where
I
totally
lost
myself
in
a
good
in
a
good
way,
but
it
was
like
wow.
We
can
do
all
that
already
and
it's
not
even
v1
yet
so
I'm
yeah
mind
blown
and,
of
course,
arvister
hci
will
be
part
of
the
the
demo
on
its
own.
If
you
are
interested,
I
have
the
links
on
the
end.
B
But
again
my
colleague
shang
yang,
made
the
q
verts
session
last
year,
okay
during
the
first
one
on
harvester
itself,
and
that's
where
you
will
get
really
more
details
like
technical
details.
Again,
I'm
the
writer
guy,
I'm
not
the
dead
guy.
So
what
will
not
be
covered?
However,
it's
I
will
not
show
an
installation
of
cubert.
I
will
not
show
an
installation
of
harvester.
I
put
links
here
and
I
won't
speak
about
the
feature
gates
of
cubert.
Neither
I
will
really
use
as
it
came
like
as
a
user.
B
B
B
We
will
actually,
however,
kind
of
deep
dive
into
the
cuber
top
level
for
one
aspect,
just
showing
you
like
the
trail
that
I
had
to
go
through
again
as
someone
that
was
interested,
but
it
was
quite
interesting,
then
again
we
will
be
back
to
the
vm
itself
and
and
see
what
actually
the
trail
led
me
to
perform
and
to
do
and
finally,
kubernetes
running
on
kubernetes.
Okay,
it's
more
like
just
saying,
like
we
are
leveraging
already
some
kind
of
powerful
technology
that
is
below
and
we
have
like
h8
clusters.
B
B
B
So
here
is
my
vm:
let
me
just
put
it
bigger,
so
you
can
see
better
all
right.
So
this
is
the
vm.
As
you
can
see,
I
took
the
the
windows
server
2022
iso
that
you
can
find
for
free,
like
it's
the
insider's
one,
but
you
can
find
it
for
free
for
test
purposes.
B
And
here,
as
you
can
see,
I
mean
that's
windows
right,
like
with
the
windows,
11
look,
but
that's
a
windows
vm,
and
here
we
can
already
see
like
there's
this
name:
okay,
cool
and
then
we
have
like
the
ip
from
from
the
cubert
masquerade,
actually
network
and
and
then
what
is
nice
here.
What
I
wanted
to
show
more
is
like,
thanks
to
all,
let's
say
the
capabilities
of
kubert.
B
We
actually
have
virtualization
enabled,
for
example,
on
this
vm.
As
you
can
see,
it's
like
four
cores.
Okay,
one
socket
for
cores.
It
takes
a
model
of
my
own,
let's
say
running
harvester
os,
so
it's
a
it's
just
a
laptop
by
the
way,
but
you
can
install
it
on
on
other
bare
metal.
A
B
A
B
We
have
also
something
that
is
a
little
bit
more
interesting,
which
is
like
the
bios
is
ufi
and
we
have
the
seeker
boots
activated
and
that
one
to
reach
that
part
again
we
will
see
just
after
was
quite
interesting
really
so
overall,
if
I
now
get
back
here,
so
this
is
harvester,
like
I
said,
is
already
installed.
I
won't
speak
much
about
it,
but
here
you
can
see
like
I
have
three
vms
one
is
running,
which
is
my
windows
server
for
cpus
8
giga
ram,
and
here
is
the
actually
the
pod
network.
B
So
you,
if
you
remember
it,
was
10
0
2
2,
like
it
was
again
another
network
on
top
of
it.
So
here,
if
I
just
go
to
the
config
as
a
user
again,
everything
here
will
translate
to
cubeverd
okay,
so
it
will
like
run
and
create
the
vm
for
me.
So
it
sets
a
cpu
in
memory.
It
gives
it
a
name
choose
the
namespace,
so
this
is
really
like
kubernetes
now,
then
we
can
have
some
volumes
also
added
to
it.
B
We
will
see
it
maybe
a
bit
later
if
the
time
permits,
but
we
have
the
the
cd-rom.
We
have
the
disk
and
everything
actually
here
is
abstracted,
so
it
translates
to
keyboard
capabilities
right
about
the
disks
and
we
have
even
container
disks
that
are
possible,
which
I
didn't
even
know
about
it,
but
that's
quite
cool.
Then
we
have
the
networking
and
again
I'm
not
selling
harvester.
I'm
just
saying
like
this
is
like
the
the
ui
for
qvert
below
right
and
here
so
like
I
said
we
have
masquerade
and
so
on.
B
So
the
point
here
that
is
important
for
me
to
show
now
is
that
I
have
also
like
a
yamo
and
let's
pass
everything
here,
and
the
points
that
were
important
is
like
thanks
to
this
configuration
also
sharing
on
yamo.
I
could
actually
add
the
features,
as
you
can
see.
There's
no
hyper-v
features
because
I
know
they
exist,
but
I
didn't
get
into
it
yet,
but
I
could
add
the
virtualization,
so
the
nested
visualization
into
it
and
a
little
bit
further
down.
B
I
could
add
the
seeker,
but
okay,
so
for
now
that
is
the
first
part,
let's
say
of
the
demo
at
the
end,
I'm
running
my
vm
here
right,
so
I
really
have
a
vm
running
windows.
Server
2022
on
top
of
cubert,
that's
cool,
but
that
wasn't
easy.
B
So
actually,
no,
let
me
let
me
get
back
here
so,
okay,
too
many
screens
and
let
me
jump
into
then
the
second
part.
So
once
I
installed
it
like
arvister,
I
created
my
first
vm,
okay,
nice,
but
now
I
wanted,
let's
say
the
virtualization,
so
let
me
go
back
to
cubert
and
then
on
kubert.
I
had
to
follow
some
trail
and
I
arrived
to
these
top-level
api
objects
right,
so
the
tip
the
the
first
one
that
I
needed
was
like
cpu
to
see
like
okay.
B
What
is
doing,
I
had
the
model
which
is
a
bit
yeah
here,
so
I
could
see
that
there's
a
lot
of
models
models
of
of
cpus
that
I
can
use
if
I'm
on
intel
and
also
on
arm
or
amd
so
again,
really
really
great.
B
Then
I
have
this
feature
here
because
I
wanted
I
I
saw
like
read
that
and
by
the
way,
thank
you
for
for
the
docs
rights,
because
sometimes
I
was
going
back
and
forth
on
on
openshift
and
read
that
documentation
and
they
were
showing
like
okay,
there's
this
vmx,
which
is
the
the
visualization
headers.
So
there's
the
the
vmx
that
is,
that
could
be
enabled
okay.
B
B
So
if
I
come
back
now
to
this
one,
I
have
the
cpu.
I
have
the
features,
and
now
I
have
policies
which
is
either
disable
or
require,
but
I
have
also
a
name
so
did
that
one
was
a
bit
more,
let's
say
a
bit
more
complex
to
get,
and
especially
when
we
are
going
down
really
down
thanks
to
this
blog
post,
I
could
understand
it
will
be
better,
and
here
will
be
like
the
name
of
the
features
from
a
kernel
perspective,
because
again
remember
we
have
two
layers.
A
B
We
have
also
the
os
layer
because
it's
visualization
it
relies
on
kvm,
and
now
that
was
like,
when
the
windows
guy
was
really
like,
trying
to
get
all
the
information
necessary
to
finally
get
everything,
let's
say
created
and
creating
a
vm.
That
would
make
sense
that
we
could
use
right.
So
here
again
is
also
the
key
emu.
B
Cpus
list,
like
the
naming
and
once
I
got
to
the
cpu
feature
right.
I
wanted
then,
to
have
ufi
and
secure
boots,
for
example,
by
default.
When
I
use
harvester
it
comes
with
bios,
not
ufi,
and
there's
no
shaker
boots
at
all
there.
So
I
was
like
okay.
What
could
I
do?
And
here
that's
where
I
found
out
where
it
was
okay?
Let
me
jump
here
back
and
that's
where
I
found
out
like
the
firmware
we
could
use
like
the
bootloader
cool
and
then
on
the
features.
There's
this
smm.
B
I
don't
even
remember
what
it
is,
but
it's
necessary
for
the
ufi
to
be
enabled
and
actually
thanks
to
the
dots
again.
So
I
didn't
put
this
one
here,
but
if
I
go
to
features,
I
can
see
that
I
have
the
smm
here
and
oh
okay,
nice
I'm
on
okay,
I'm
doing
some
live
annotation,
and
then
we
have
also
firmware
that
should
be
here.
B
Okay
and
here
I
have
bootloader,
and
if
I
follow
bootloader
I
can
set
the
ufi
and,
if
I
get
to
ufi.
Finally,
I
can
tell
like
sakerboot
is
on
or
off,
and
here
thanks
to
this
documentation,
I
could
find
out
that
it
requires
smm.
So
now
I
knew
where
to
find
smm.
It
was
on
the
feature,
so
I
could
enable
it.
So.
Finally,
I
would
have
everything
here
like
created
and
my
vm
could
finally
be
running
like
a
ufi
boot
and
also
the
secure
boots.
That's
what
we
see
here.
B
B
B
B
I
just
put
it
on
on
bolts,
but
that's
a
very
nice
feature,
which
is
the
run
strategy
instead
of
running,
which
is
true
or
false,
actually
the
run
strategy
because
we
are
running
on
top
of
kubernetes
right
so
right
now,
whenever
I
shut
down
from
the
os
as
we
speak
right,
the
vm
will
actually
restart
because
it's
a
pod
at
the
end
of
the
day
and
then
the
deployment
the
demon
set
just
said
like
okay,
this
pod
is
not
no
more
working
okay,
so
I
have
to
restart
it
and
then
your
vm
restarts
and
that's
not
really
what
we
want
thanks
to
cubert.
B
Actually
now
we
have
like
the
run
strategy
and
we
can
shut
it
down
and
we
have
this
rerun
on
failure
only
so,
if
it
fails,
it
sends
a
flag,
I
guess,
and
then
it's
like
okay,
the
pod
now
is
restarted.
However,
if
I
shut
it
down
properly,
then
the
vm
will
just
stop,
which
is
just
really
really
cool
it.
B
It
but
for
me
it's
really
really
amazing,
so
so
yeah.
So
here
is
the
again
the
the
rabbit,
the
rabbit
trail,
the
rabbit
hole.
So
I
we
had
the
cpu
features.
We
had
the
firmware
and
we
have
the
smm
feature
also.
B
B
Oh
no,
okay,
sorry,
not
I'll
tap!
I'm
my
bad!
I
know
it's
recording,
there's
nothing
critical
there.
Let
me
show
you
virtual
machines
all
right
so
here
I
can
open
a
console
with
webvnc.
B
B
B
Sorry
here
to
my
console
here
and
change
the
ip
so
harvester,
sorry
arson,
not
our
sir
cuber
did
something
really
nice
with
the
service
objects
and
it
simply
can
create
like
a
noteport
okay
from
the
ports.
That
is
on.
Let's
say
the
node
port,
so
external
port,
which
is
like
by
default.
It
was
shown
as
30
000.,
then
a
port
of
kubernetes
the
service
and
then
finally,
a
target
board
to
the
bot
to
the
vm,
which
is
the
3389.
B
So
for
me
it
was
perfect.
It's
like
oh
okay,
so
now,
as
you
can
see
here,
if
I
switch
back
to
this
one,
I
have
not
my
pod
ip
every
time,
but
I
have
my
node
ip
with
the
port
with
the
different
ports
and
I
can
connect
to
the
different
vms
once
they
are
running
and
it
will
actually
go
to
the
different
vms
if
they
are
running
or
not.
Okay,
in
this
case
here,
I
don't
care
about
this
ip.
No
more!
B
B
So,
actually,
how
much
time
do
I
have
more?
I'm
just
showing
the
kind
of
the
last
big
11
minute
11
minutes.
Okay,
can
I
try
something
nice
then
potentially
sure
all
right,
so
let
me
again
swap
back
here.
So
what
I
will
do
is
I
will
not
create
a
new
vm
in
this
case,
but
what
I
will
do
is
like
I
will
try
very
quickly
to
create
a
new
volume,
so
it's
using
longhorn
behind
the
scenes.
B
But
again
I
I
will
refer
to
the
to
the
to
the
demo
from
sheng
yang
last
year.
It
was
really
nice,
kubert
submits
disk,
and
I
want
yeah,
just
a
10
gig,
all
right
and
it's
new
and
I
will
create
so
what
arister
is
doing
is
like
creating
a
a
pv
behind
the
scenes.
Okay-
and
now
I
can
no,
this
one
is
created,
so
I
will
go
to
the
virtual
machine,
so
the
visual
machine
is
running.
A
B
B
A
Questions
from
chad,
one
question
was
from
andre:
does
harvester
support
redfish
boot
instead
of
pixie?
A
B
Don't
know
again,
I
will,
I
will
send
out
to
I
put
like
the
resources,
but
everything
that
is
about
really
harvester.
Let's
say
I
know
ipxc
it's
it's
used
redfish,
I
am
not
sure,
cannot
tell.
A
B
All
right,
no,
I
didn't
do
anything.
I
just
really
created
and
added
these
settings
and
if
there's
l2
l3
cache,
I'm
not,
I
didn't
do
anything
if
it's
enabled
then
windows,
server
2022
is
the
one
that
enabled
it
for
me,
but
I
didn't
touch
anything.
The
only
thing
that
I
did
at
some
point
was
to
choose
a
model
of
cpu
in
the
in
the
configuration
really
like
I
showed
like
in
the
in
the
game,
new
cpus,
but
other
than
that.
No,
I
didn't
do
anything.
A
Okay
and
alexander
indicated
that
during
your
demo,
you
didn't
unclick
the
restart
vm
and
that
might
have
been
the
source
of
the
error.
Oh.
B
Wow
yes,
smarter
than
me:
okay,
it's
already
rebooted,
but
let
me
see
if
the
volume
is
the
volume
is
not
bound
so.
B
A
B
B
The
goal
here
is
to
actually
use
the
plug
volume
capability
feature
gates
and
you
can
just
add
a
new
disk
to
your
running
vm.
You
don't
have
to
to
cancel
it,
but
yeah
right
now.
Democrats
are
not
happy.