►
From YouTube: KubeVirt Community Meeting 2019-04-03
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
B
Yes,
the
issue
is
that
newer
machine
types
don't
supports
floppy
discs
anymore,
so
we
decided
to
starting
to
reject
them
completely,
not
for
selected
machine
types
only
but
completely,
because
we
are
not
aware
of
any
real
use
good
use
case
for
it.
The
implementation
is
done.
You
user
guide
update
is
in
progress.
That's
it.
A
C
It
may
have
not
so
there's
that
we're
going
towards
0.3
release
merchant
last
bits
and
also
integration
to
hyper-converged
operator
of
this
network,
adult
operator
was
merged
so
operator
of
the
way.
So
now
hyper-converged
operator
go
as
one
of
the
operators
that
it
deploys.
There
will
be
the
network
at
also
major
and
also
on
my
side.
I
got
back
to
looking
at
what's
going
on
with
the
server
VCI,
so
one
of
the
ideas
like
you
may
have
recollect.
C
Oftentimes
couldn't
start
some
paid
services
like
docker
and
whatnot.
So
one
of
the
ideas
that
I
had
is
that
perhaps
the
problem
is
with
the
image,
because
the
errors
that
we
saw
there
were
really
weird,
like
sack
of
balls
from
from
container
D.
So
what
I'm
trying
right
now
is
to
pay
a
little
price
and
just
try
to
replace
this
image
of
unclear
origin
with
with
the
set
up
that
so-called
kind
tool
can
produce.
C
A
kind
tool
is,
as
far
as
I
understand,
it's
like
an
official
from
from
religious
community
that
allows
you
to
run
clusters
in
context
in
actual
containers,
so
it
may
deploy
you
like
come
out.
You
know
it
set
up.
Each
node
gets
its
own
container
and
they
also
use
this
tool
in
CI
needs.
So
this
is
my
hope
that,
since
they
are
actually
using
it
for
reproducible,
you
know
environments
in
other
places,
then
it
should
also
be
stable
in
our
environment.
C
A
C
It's
actually
very
nice.
It's
like
a
CLI
tool,
you
just
call
kind
create
cluster
and
to
get
a
single
node
in
the
container
you
can
pass
configuration
with
multiple
nodes
and
whatnot.
I
was
even
wondering
why
we
would
need
to
to
use
something
like
to
start
environments
like
we
do
right
now
with
the
with
the
qmo
and
not,
if
you
can
do
it
doing,
containers
and
skip
the
second
virtualization
layer,
but
I
guess.
C
The
only
reason
for
that
would
be
that
you
want
to
completely
isolate
from
from
environment
of
the
host,
meaning
like
have
your
own
kernel
and
whatnot
I'm,
not
sure
what
we've
always
needed
to
be
sure
that
we
never
use
the
hosts
colonel
at
least
yeah,
but
I
mean
this
can
be
an
explanation,
but
otherwise
it's
a
very
nice
tool.
I.
A
A
A
D
Hey
so
you're,
just
not,
if
you
can
hear
me
yeah
great,
so
I've
been
debugging
a
back
when
Qbert
failed
to
reconnect
to
the
NC.
This
is,
as
far
from
my
point
of
view,
it's
quite
serious
because
it
blocks
user
for
a
using
the
VNC
connection
when
trying
to
work
with
virtual
machine.
Well,
as
it
turns
out
this
buck
is
not
ours,
it
belongs
to
the
QM.
Oh,
and
does
they
already
know
about
that
and
they
already
fixed
it?
Why
I
am
mentioning
it
that
it
is?
D
It
is
fixed,
but
it
is
fixed
in
staging
and
scheduled
for
release
in
qmo
4.0,
which
will
be
someday
in
the
future,
and
the
buck
is
actually
pressing
us
now.
So
there
are
two
solutions:
I
can
figure
out.
First,
one
is
setting
the
a
basically
what
we
do
is
we
open
a
socket
and
then
we
connect
to
this
UNIX
socket
a
quick
workaround
that
will
work
I
already
tested.
D
It
is
set
up
the
permissions
to
be
read
only
this
effectively
block
the
QA
mode
for
deleting
the
socket
and
everything
works
as
as
we
expect,
and
second
workaround
is
actually
not
a
workaround,
but
a
proper
solution
which
is
worked
on
by
sebastian
from
networking
team
sebastian
is
designing
a
proper
proxy
for
VNC
and
console,
and
from
this
point
of
view,
I
believe
it's
important
to
stress
that
we
choose
the
solution.
That
is
either
not
block
on
the
unix
sockets.
That
means
use
proper
VNC
server.
For
this
part,
so.
E
D
D
E
F
E
I
added
it
yeah
Roman
again,
although
our
team
is
also
there,
I
just
wanted
to
give
a
heads
up
that
Artyom
and
Oreo.
Whoever
is
a
person
behind
it.
Nick
managed
to
manage
to
make
virtual
machines
running
on
PPC
and
I,
where
there,
in
the
middle
of
trying
to
integrate
it
right
now
we
have
a
lipid
container
merged
ready
for
PPC
612,
we'll
see
how
it
goes
on
from
there
Artyom.
You
have
different.
F
Unit
to
run
it
on
his
pc
machine,
but
I
wouldn't
succeed
to
run
it
on
our
pc
motion,
because
some
memory
allocation
stuff,
like
from
what
I
saw
the
PPC
PMO
when
you
run
it
on
the
PC
architecture,
tried
to
allocate
some
to
look
some
amount
of
memory
and,
from
some
reason
on
our
office.
It
just
fails
so
I
I
hope
I
will
continue
to
work
on
it
on
Sunday,
because
tomorrow
we
have
a
demo.