►
From YouTube: Introduction And Kubevirt History
Description
In the first session of the KubeVirt Summit, Fabian Deutsch (@dummdida) talks about the project's history.
A
Where
it
is
so,
people
will
run
into
problems.
That
is
what
we
were
expecting
and
we
started
thinking
about
it
in
a
small
group.
Some
of
us
will
remain
in
this
cube,
stayed
involved
in
this
cube
project
ever
since.
With
that
problem
right,
we
started
to
explore.
What
can
we
do?
We
started
with
tprs
and
let
me
see
if
somebody
can
explain
tpr
the
explanation
for
the
acronym
tpr,
who
knows
what
the
tpr
is.
Please
use
the
chat
to
leave
a
clue
what
it
is,
I'm
giving
three
two
one.
A
B
B
A
A
It's
problem
difference
for
users
right,
but
it's
extensible
right
and
one
of
our
main
decision
or
in
potential
design
decisions
from
the
beginning
on
was.
We
want
to
extend
kubernetes
in
the
right
way
right.
We
we
didn't
want
to
fork
it
I
mean
we
we
played
around
actually
to
see
right.
Can
we
can
we
modify
the
cubelet
to
run
vms,
which
later
I
mean
not
our
effort,
but
later
we
saw
that
with
cri
right.
A
Others
had
that
desire
as
well
and
cr
was
broken
out,
and
now
you
can
run
anything
right,
virtual
cubelet,
they're
doing
crazy
stuff
with
cri,
but
back
then
it
was
about.
You
know
forming
the
cubelet
to
see
if
we
can
run
vms,
but
then
tprs
came
around
the
corner
and
that
allows
us
to
extend
the
kubernetes
eye
in
a
way
that
there
was
a
dedicated
api
for
bms,
but
does
that
make
sense?
A
Isn't
a
vm
just
a
container
and
as
I
alluded
before,
there
are
key
differences
right
containers
are
stateless
right,
12
factor
applications
like
they
are
designed
differently.
They
they
assemble
applications
in
different
ways
than
virtual
machines
do,
and
due
to
this
inherent
difference
in
the
nature,
we
we
thought
there
is
merit
to
it
to
to
give
it
a
dedicated
api
so
to
give
dedicated
api
endpoints
right.
So,
with
the
virtual
machine,
the
virtual
machine
effectively
was
born
right
all
right.
A
Most
of
you
know
them
probably
right,
and
they
had
a
track
record
in
that
in
the
virtualization
area
and
we
picked
them,
and
that
was
not
an
easy
decision
right.
They
were
also
controversies
around
this
controversies
around
this.
Should
we
take
delivered
or
not,
or
should
we
take
him
or
not?
The
reason
for
going
with
liberty
was
simply
they
provided
us
a
stable
api
that
enough
or
all
the
functionality
we
need,
but
we
paid
the
price
that
it
was
a
little
bit.
A
A
It's
now
already
away
or
our
recommended
way
to
deploy
keyboard,
because
it
can
really
take
care
of
the
whole
application
lifecycle
management,
2018.
I
think
it
was
2018.
I
was
putting
together
the
sliding
and
try
to
remember
it.
We
were
sitting
at
kubecon
and
we
actually
had
and
thank
you
to
the
cncf
again.
We
weren't
part
of
the
cnc
effect
then,
but
the
cncf
gave
us
a
room
to
discuss
a
virtualization
in
the
context
of
kubernetes,
so
the
kubert
still
young
kubert
project,
the
vert
led
project
actually
kata.
A
It
was
back,
then
it
was
actually
hyper.
Sh
and
rancher
vm,
we
were
all
sitting
together
in
a
room
we
were
discussing.
You
know
what
what
opportunities
are
there?
What
is
our
overlap?
What
are
the
use
cases?
Some
things
became
clear
right.
For
example,
cubert
vert
led
and
ranger
vm
in
the
one
corner,
which
is
what
running
traditional
virtualization
workloads
and
qatar
were
high
players
age
back
then,
to
run
container
workloads
in
a
vm
isolated
way.
A
I
also
added
community
features
here,
because
throughout
the
years
we
were
really
we
had
to
ramp
up
right.
We
want
to
make
keyboard
successful.
We
really
need
to
allow
you
and
us
to
run
the
workloads
we
want.
That
means
we
need
to
close
some
feature
gaps
to
existing
virtualization
platforms
like
openstack
like
overt
and
they're,
more
like
proxvox
right.
We
expect
these
features
to
exist,
so
we
want
them
to
deliver
to
deliver
them
in
a
nice
cloud-native
way
inside
of
kubert.
A
So
a
lot
of
features
for
foundational
features
were
introduced
in
that
year.
The
community
grew
right.
We
we
saw
the
first
contributions
to
kubert
from
non-redhead
contributors
right.
We
started
as
a
small
team
in
redhead,
but
then
we're
happy
to
see
excellent
contributions
from
what
external
conclusion
the
keyboard
project
was
happy
to
see
that
controls
from
other
companies
joined
in
2019,
so
in
2019
we
actually
also
had
adopters
right
and
we
had
contributors,
and
we
were
like
working
nicely
on
kubernetes
and
we
wanted
to
get
a
better
exposure.
A
There
were
still
some
questions
right
because
it
might
not
look
like
a
spot
on
feature
that
is
relevant
for
the
cloud
native
world,
but
after
a
few
discussions
and
with
the
key
points
that
legacy
workloads
simply
don't
go
away
and
we
need
a
transition
path
right,
they
will
evolve
over
time.
But
what
do
we
do?
In
the
meantime?
A
Their
kubert
can
really
be
a
helpful,
helpful
tool
and
we
have
actually
seen
more
use
cases
where
even
that
new
vms
can
help,
for
example,
to
isolate
nested
clusters.
You
know
running
kubernetes
on
top
of
kubernetes
using
keyboard.
Lootz
has
shown
it
initially
great
work.
The
community
increased,
more
features
came
into
cuberd.
A
A
I
can
disclose
it
right
because
I
hope
that
every
all
of
you
by
running
your
recent
version,
cuberd
where
this
is
fixed,
so
users
and
operations
both
are
on
the
community
side
right.
It's
what
we
see
a
broader
adoption.
We
saw
a
broader
adoption
of
keyboard
and
more
contributions
again
for
scale
right
contributions
to
scale
cuberd
and
such
and
instead
of
features
what
we
saw
started
to
do
or
what
we
see.
What
naturally
came
was
actually
stabilization
right.
So
we
there
were
more
contributions
to
stabilize
cupid
right
to
fix
more.
A
A
There
was
a
ramp
up
phase
and
I
would
say
we
are
there
right
in
that
slope
the
slope
to
stabilize
for
a
bit
before.
We
can
then
continue
to
grow,
not
necessarily
grow
but
to
well
grown
one
way
or
the
other,
but
to
to
be
really
a
platform
that
you
trust
your
production
workloads
that
you
want
to
use
for
your
production
workloads
on
that
side
in
2020.
We've
also
seen
a
lot
of
work
on
and
we
have
sessions
about
it
and
we
get
to
in
a
second.
A
We
saw
contributions
of
how
to
enable
it.
You
know
to
migrate
your
existing
workloads
over
to
keyboard
and
note
on
the
questions
I'll
get
back
to
them
once
I'm
through
my
slides,
which
is
done
in
a
few
minutes
important.
Is
that
we're
not
just
there?
But
all
of
us
are
here
right,
I
think
from
a
developer's
perspective,
we've
really
done
a
lot
and
we're
stabilizing
a
set,
but
as
important
as
it
is
to
write
software,
it's
even
more
important
to
see
that
it's
getting
used
and
yeah
we're
seeing
this.
A
We
we
as
community
to
all
of
us,
are
part
of
that
community.
So
it's
great
to
see
all
of
you
here
and
I
would
like
to
welcome
you
again
and
allow
you
to
enjoy
the
sessions
which
are
following
up
today
and
tomorrow.
Right
and
I
want
to
extend
the
thanks
to
all
the
volunteers
who
helped
to
really
get
get
all
the
things
together
to
the
speakers
who
submitted
the
abstracts
and
will
be
holding
the
the
sessions
and
the
session
hosts,
which
are
standing
by
to
help
run.
A
This
conference
feel
free
to
tweet
about
it
hashtag
keyword
summit,
how
you
can
find
more
cuberet
material
on
cubelio
or
github
in
the
cube.org.
Last
but
not
least,
my
name
is
fabian
deutsch,
I'm
working
in
kubert
since
2016
and
yeah,
I'm
just
a
cuber
community
member.
Thank
you
very
much
and
now
I'm
getting
to
the
questions.
A
Hello
from
india,
hello
from
germany.
Actually
it's
no
snowy
out
here
who
other
than
reddit
openshift
is
providing
enterprise
ready
cuber
today.
So
there
are
other
companies
right
who
are
providing
keyword.
I
first
this
is
no!
You
know
this
is
a
community
conference
and
I
can
only
see
or
tell
and
share
what
I
I've
been
seeing
right.
So
we
know
that
kubernetes
is
providing,
including
kubert
one
of
the
products.
We
know
that
there
is
gardner
support
for
cuberd.
A
B
A
Absolutely
thank
you
for
keeping
me
honest.
It
wasn't
I
mean
yes,
absolutely
I
mean
harvester
was
great
to
see.
Oh,
it's
actually,
that's
the
funny
anecdote
right.
I
mean
harvester
as
a
successor
of
ranger
vm,
it's
nice
to
see
that
we
converge
on
an
underlying
technology
which
is
keyboard
and
that
closes
back
to
an
earlier
slide
right
from
2017,
no
intention
to
hide
anything
here.
Yeah,
obviously
harvester.
A
A
A
A
All
right
with
this,
I
thank
you
very
much
for
joining
this
very
first
session
and
again,
please
take
a
look
at
our
agenda
that
met
so
many
talks
and
again,
thank
you
very
much
to
all
the
speakers
coming
up
part
of
this
inno
organizational
kubert
summit
and
with
this
thank
you
and
I
hand
it
over
to
pep.
Thank
you
very
much.