►
From YouTube: KubeVirt Community Meeting 2021-10-27
Description
Meeting notes: https://docs.google.com/document/d/1kyhpWlEPzZtQJSjJlAqhPcn3t0Mt_o0amhpuNPGs1Ls/edit#heading=h.d5rswh7pb7ew
A
Hello
everyone:
this
is
the
weekly
community
meeting
for
project
covert.
It
is
wednesday
october
27th
2021,
I'm
your
host
chris
caligari
welcome
everybody.
I'm
going
to
post
our
meeting
notes
to
chat.
If
you
could
log
into
that
google
document
and
record
your
attendance,
we
do
track
that
information.
A
I
would
really
appreciate
it
and
at
this
time
we
have
no
agenda
and
just
one
item
on
the
open
floor
to
talk
about
events
I'll
give
you
I'll
give
everyone
a
minute
or
two
to
fill
in
some
points
that
you'd
like
to
talk
about.
B
Hello,
this
is
michael
from.
I
have
a
little
bit
to
add
about
the
proposal
I
made
in
last
meeting
about
the
zen
hypervisor
support.
B
B
A
Thank
you,
michael.
That
was
also
a
great
architecture
that
you
brought
to
us
last
week.
We
really
appreciate
that
effort.
A
Okay,
I
got
it.
Thank
you
so
much
leonardo
cloud
in
it
go
ahead.
C
C
C
D
Yeah,
I
I
think
you
actually
found
the
solution.
So
here's
the
thing
about
cloud
nf,
it's
kind
of
os
or
guest
specific
when
it
actually
completes
so
there's
not
at
least
not
that
I'm
aware
of
a
way
to
generically
determine
when
a
guest
has
done
anything.
Unless
we
have
some
sort
of
hook
into
that
guest,
that's
reliable
and
consistent.
So
with
cloudnet,
we
don't
really
have
that.
But,
like
you
have
observed
you
can
you
can
create
a
readiness
probe
and
then
have
that
check
something
within
the
guests
to
determine
readiness.
D
I
think
that's
the
best
solution.
We
have
right
now.
You
know
I'm
definitely
open.
I'm
sure
we're
all
open
to
hearing
ideas
to
improve
this,
but
it's
kind
of
that's
a
difficult
thing
to
determine
generically
across
all
guests
when
cloudinet
has
actually
completed
without
some
sort
of
application,
specific
hook,
like
he's
done.
C
Yeah,
I
got
it
yeah
sorry,
it
has.
E
No,
I'm
just
an
idea,
I
don't
know
if
it's
possible,
I
mean
the
only
thing
we
we
have
is
the
qa
new
agent.
I
don't
know
if
that
can
be
extended
to
somehow
get
back
this
information
to
kill
birth.
D
Sure
so
there
was
a
kimu
agent
hook
that
knew
how
to
interpret
cloudinet
results
or
or
knew
about
cloud
net
execution.
Then
we
could
hook
into
that
in
a
generic
way.
That
would
that
would
work.
We'd
have
to
talk
to
qmu
about
that.
Getting
that
kind
of
hook
created.
D
D
C
Know
what
I'm
basically
doing,
I'm
checking
if
the
results
the
result.json
file
is
there
and
if
into
the
var
lab
cloud,
need
data
folder
I
was
checking
the
cloud
needs
the
source
code
itself.
That's
basically
how
it's,
how
how
it
updates
these
it's
the
status
as
done
so
I'm
basically
returning
and
and
in
found
http
status
code
if
that
file
exists.
C
D
C
Yeah,
because
my
problem
is
that
I
needed
to
verify
that
all
packages
and
all
systems
and
lao
systems,
these
services
they
were
running,
but
that
part
of
the
cloud
needs
is
a
known
blocking
stage
with
that
with
the
modules
that
final
stage
is
not
wrong,
since
it's
no
blocking
the
vmi
resource
states
as
ready,
where,
for
my
particular
application,
need
it's
not
a
ready
state.
So
that's
why
I
needed
some
kind
of
a
synchronization
between
cloudiness
and
kubernetes.
F
A
Okay
I'll
take
over
then
for
events
discussion.
We,
as
you
all,
probably
know,
kubecon
n.a,
just
finished
up.
We
david
boswell
and
I
did
a
a
presentation
introducing
kubert
all
things
open,
also
just
finished
up.
Stu
gott
and
myself
did
a
presentation
running
cougar
on
arm
64
and
raspberry
pi
4b.
A
That
was
both
both
conferences
were
really
well
received.
A
We
had
a
lot
of
attendance,
a
lot
of
questions,
and
so
it's
always
good
to
see
no
rest
devconf
czech
republic
is
is
now
calling
for
papers.
The
deadline
is
october,
31st,
so
sunday
get
your
papers
in.
If
you
want
to
speak,
alicia
frozie
has
volunteered
to
submit
a
paper
for
us.
I
believe
that
it's
complete.
A
All
right,
thank
you
so
much
we
got
one
in
if
anybody
else
wants
to
submit
a
paper
go
ahead.
Just
let
me
know,
though,
so
I
can
create
a
an
issue
in
the
community
repo
and
and
if
there's
anything
that
you
need,
like
demo,
material
presentation
slide
deck.
A
We
have
all
of
that
available
bad
news
with
super
computing.
Con
2021.
We
totally
missed
the
call
for
papers,
it
was
the
call
for
papers
was
in
january
and
the
conference
is
coming
up
in
november.
A
This
is
this
is
one
that
I'm
passionate
about,
because
all
the
big
scientific
organizations
are
present,
like
cern
and
lawrence
livermore
institute
nasa,
and
I
would
really
like
to
pitch
cooper
to
those
groups.
A
So
I
will
try
to
do
super
computing.
Con
2022
come
come
the
call
for
papers
time
in
january.
Sorry,
everyone.
A
A
That's
about
all
I
have
for
events,
the
ncf
incubation
not
a
whole
lot
of
progress.
The
last
the
last
update
I
got
was
nvidia
has
been
interviewed
for
to
get
their
opinion
on
working
with
the
project
and
see
we
got
a
video.
We
got
apple
computers
actually
volunteered
to
to
be
interviewed,
which
is
interesting
because
the
the
cncf
technical
oversight
committee
representative
for
coobert
is
from
apple
so
not
sure
how
that
that
conflict
is
going
to
work
itself
out.
A
I
haven't
heard
anything
other
than
another
apple
developer
wanted
to
do
the
interview,
and
then
I've
been
speaking
with
howard
zang
from
arm
michael.
I
believe
that
you
work
with
howard,
and
I
think
I
have
him
talked
into
being
the
third
organization
to
interview.
A
So
hopefully
that
works
out
well
and
we
can
proceed
on
to
the
next
stage
of
due
diligence.
G
Oh
yeah,
we
so
there
is
finally
some
progress
on
getting
away
from
kubrick
clanco,
which
causes
huge
issues
for
a
lot
of
consumers
of
the
library,
because
we
depend
on
certain
kubernetes
versions,
but
the
consumer
projects
have
other
kubernetes
versions
which
then
conflict.
G
It
should
already
be
possible
for
people
to
directly
use
client
chen
now
on
our
api
happy
if
anyone
wants
to
try
it
out
already
and
give
some
feedback,
but
the
next
step
is,
then
that
you
don't
have
to
look
at
any
of
this
at
all
and
you
will
find
just
the
api
in
github.com
keyword,
slash
api
and
you
directly
just
add
it
as
a
vendor
dependency
and
run
client
again
on
it
without
anything
else,
and
then
hopefully
you
only
have
it
in
one
api
dependency
to
kubernetes,
slash
api,
and
this
one
is
normally
absolutely
easy
to
adjust
to
your
needs
yeah.
G
A
I,
like
your
mailing
list,
topic
about
hell
and
then
you
have
a
coincidental
666
pull
request
number.
A
A
F
F
A
F
You
have
those
those.
G
G
D
G
G
D
G
F
0X0
must
be
specified
for
now.
Yeah.
G
G
G
D
F
G
A
Let's
tag
this
with
needs,
more
information.
A
What
is
it
needs
more
info.
G
A
A
K
We
don't
copy
it,
we
just
we
copied.
G
A
A
A
G
D
B
D
A
Okay
is
vert,
does
cooper,
vm
support
sound
purge.
G
G
B
G
G
A
So
speaking
of
vert
cto
things
got
strange
during
the
all
things
open
demo.
I
had
a
catacota
scenario
open
for
a
short
time,
maybe
10
minutes
and
when
it
come
time
to
curl
the
vert
cto
binary
down
from
github.
We
are
getting
403
permission
denied
from
github.
A
Was
remote
conference
okay
screen
sharing
my
session
to
catacota.
G
A
Yeah,
that's.
We
are
pondering
on
the
same,
the
same
items,
but
we
weren't
able
to
show
a
console
to
the
vm.
We
were
able
to
start
it
luckily
by
editing
the
bmi,
but
you
couldn't
show
the
console.
Okay.
G
G
F
F
Like
all
things,
okay,
this
was
yeah
this
issue,
you
know
about
already
yeah.
We
spent
a
lot
of
time
on
that
yeah.
A
A
So
thank
you
for
joining
us.
See
you
next
week.
Of
course,
if
you
have
any
questions,
feel
free
to
reach
out
in
the
mailing
list
and
slack.