►
From YouTube: KubeVirt community meeting 2021-03-24
Description
KubeVirt community meeting 2021-03-24
A
Okay,
hello,
everybody
welcome
to
the
coobert
weekly
community
meeting,
I'm
your
host
chris
caligari
and
I
have
a
co-host
of
stu
got
today
and
let's
begin
our
agenda
here,
we
usually
start
off
with
introductions
for
new
new
folks
who
join
the
project.
But
as
I
look
over
the
list,
I
think
everybody
here
has
has
been
here
a
while.
B
I
was
just
asking
whether
we,
whether
we
should
now
ask
everyone
who
is
trying
to
do
that,
to
point
him
to
the
not
yet
merged
a
questionnaire
that
he
has
to
fill
out
and
check
all
the
marks
and
and
do
everything
that
is
requested
on
there
and
then
then
he
can,
for
example,
get
get
back
somehow
yeah
as
the
pr
is
not
yet
merged,
and
I
saw
your
answer,
but
I
guess
you
want
to
formulate
the
answer
yourself.
Chris.
A
C
A
A
E
Let's
pretend
for
a
minute
that
the
the
pr
was
merged.
What
would
be
the
difficulty
with
following
the
formalized
process
that
we've
laid
out,
even
though
it's
it's
not
yet
formally
merged
and
canonized
in
our
system?
I
think
we
can
all
reasonably
agree
that
it's
a
good
system
and
there's
no
objections,
so
I
would
play
devil's
advocate
and
say
yeah.
Let's,
let's
start
using
it,
even
if
we
don't
have
the
document
completely.
B
Yeah
chris
actually
to
chime.
You
know
that
I've
actually
asked
how
whether
he
would
be
willing
to
do
that,
and
he
agreed
on
that
and
yeah.
I
I'm
not.
I
I
think
he
told
me
that
he
wanted
to
share
a
somehow.
Is
this.
This
check
a
list
of
check
marks
with
with
the
community
somehow
and
yeah
get
get
everything
in
what
was
requested
in
the
list
and
yeah.
Then
we
should
be
good
with
him.
B
E
B
C
Failed
following
following
a
pr
created
from
you
guys
and
try
to
build
out
and
stand
to
the
mailing
list.
That
was
the
process
I'm
just
about
to
send
to
the
mailing
list.
Basically,
in
there
there's
like
10
checkpoints
and
the
two
sponsors
I've
filled
in
and
then
the
prs,
I
need
to
write
here
so
right
now,
basically,
google,
google
doc
submit
to
the
mailing
list
right.
This
is
with
the
github
interaction.
C
B
B
A
A
A
D
I
just
wanted
to
give
a
quick
heads
up
about
a
change
which
you
may
see
in
the
cia
system
when
you're
coding
in
the
past,
bazel
was
also
caching.
Our
unit
test
results
and
vasily
created
the
pull
request
which
will
stop
doing
that
and
on
the
and
do
the
actual
opposite.
It
will
run
every
unit
test
multiple
times
and
it
also
shortened
the
execution
times
a
little
bit
in
the
unit
tests.
D
G
D
D
D
D
D
A
F
A
F
F
F
For
example,
if
you
just
make
a
small
change
and
you
need
to
sync
a
lot
of
things
and
we
were,
we
wanted
to
ask
if
we
can
think
of
a
way
that
it
allows
to
just
update
a
single
component
in
kubert
in
general,
like
I
was
working
on
inverted
api,
I
made
a
small
change,
but
I
had
to
remove
all
my
vms
reapply
all
components,
but
I
just
wanted
to
update
with
api
and
it's
a
general
starter
of
the
conversation.
I
guess.
D
D
F
E
Clusters
have
state,
and
so
let's
say,
you're
changing
out
the
vert
handler
daemon
set
that's
great,
but
there
would
already
be
pre-existing
caches
and
what
have
you
on
the
vert
controller
on
vert,
api,
etc?
So
it
it
looks
good
it
briefs.
Well,
you
know
in
your
mind,
it's
kind
of
a
do
what
I
say
not
what
I
mean
sort
of
thing,
but
but
there
are
devils
in
the
details
and
that's
what
roman
is
into
that.
So
I'm
not
in
general.
E
D
My
two
senses
so
what's
interesting
is
maybe
we
can
speed
up
the
deployment
as
such.
So
when
you
do
make
cluster
sync
and
you
just
change,
I
mean
well,
I
mean,
what's
that,
what
you're
probably
experiencing
a
lot
right
now
is
you're
changing
something
in
the
code
base
and
it's
shared
across
multiple
containers.
F
D
C
D
G
A
workflow
for
you
kevin
if
we
had
a
way
of
building
the
new
images
that
they
change,
pushing
them
to
your
test
cluster
and
then
you
manually
know
which
ones
were
the
ones
you
want
to
update.
So
let's
say
you're
working
on
vert
handler,
and
you
know
that
your
change
only
impacts
pert
handler
use.
My
quickly
push
vert
handler.
D
I'm
not
sure,
but
it
can
be.
We
still
take
everything
with
devil
too,
so
it
may
be
good
enough
to
just
go
after
the
deployment
to
the
operator
and
set.
The
devil.
Tag
is
the
target
tag,
and
then
you
can
do
this
because
you
wouldn't
just
run,
make
push
images
and
be
pushed
into
images,
and
then
you
would
have
on
the
nodes,
the
new
devil,
tagged
images
and
just
killing
the
deployment
they
would
use
the
new
one.
Something
like
this.
F
D
C
F
A
A
A
The
other
advanced
features
such
as
instant
messaging
and
video
conferencing
that
are
embedded
into
the
platform
are
going
to
be
turned
off
if
attendees
want
to
do
something
more
advanced
and
more
one-on-one
with
the
community
just
send
them
my
way
and
we'll
schedule
time
and
open
up
open
up
a
zoom
conference
generally.
I
just
want
to.
I
want
to
use
the
zoom
con
conference
as
much
as
possible.
A
A
I've
already
talked
about
the
booth
features
kubecon
boot,
the
kubecon
booth
is
going
to
be
office
hours
only
and
that's
similar
format.
At
a
summit.
It's
just
going
to
be
a
chat
during
blocks
of
hours.
I
have
a
document.
That's
basically
got
everybody.
Who's
gonna,
be
at
the
at
the
booth
during
one
hour.
One
hour
blocks
of
time,
I'd
like
to
have
two
people
per
per
block
of
time.
A
That's
that's
gone
on
with
the
with
between
cooper
and
ospo.
I
think
I'm
gonna
change
that
this
this
next
cycle
and
and
like
I
emailed
the
cougar
develop
list.
These
meeting
notes,
I'm
gonna
do
the
same
with
the
the
ospo
meeting
notes,
so
that
everybody
is
on
the
same
page
as
as
me
with,
what's
been
getting
discussed.