►
From YouTube: KubeVirt Community Meeting 2019-08-28
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
Welcome
everybody:
this
is
a
Kubrick
community
meeting,
August
28
2019.
First
up
on
the
agenda
is
the
per
a
theist
service
monitor.
This
is
a
PR
that's
been
pending
for
a
while
I
think
we
worked
through
the
last
of
the
bugs
there.
We
had
a
bit
of
a
Heisenberg
situation
going
on
where
what
we
tried
to
identify
what
the
problem
was.
It
became
hard
to
see.
B
A
C
Yeah,
actually,
the
flick,
finer
tunas
to
find
flaky
test
results,
we're
nearing
completion
to
have
some
reporting's
for
daily
wiki
and
for
weekly
reports
for
viewing
test
results
where
tests
single
tests
flip
phone
from
green
to
red
so
which
is
effectively
a
flaky
test,
so
I'm
working
or
I'm
wrapping
my
head
around
color
correlation
problem,
so
I
think
if
that's
done
this
might
be
merged
today
or
tomorrow.
So
yeah.
D
Small
update
last
week,
we
understood
that
it
was
in
review
now
has
been
published
and
the
guys
that
cut
the
cola
has
had
added
that
lab
to
the
list
of
the
application
registration
definition
yeah.
We
got
some
feedback
from
Politico
our
day
right,
metals,
just
defender,
but
is
still
listed
match.
We
have
from
the
documentation,
so
we'll
keep
doing
some
variation.
But
if
there
are
no
questions
next,
okay.
E
A
E
Pure
number,
two
five
nine
seven
has
been
merged,
and
with
this
PR
you
can
now
annotate
your
virtual
machine
and
the
annotations
will
be
carried
down
to
the
pot
level.
This
actually
enables
guys
using
Calico
to
add
the
IP
groups
and
manage
the
pots
and
virtual
machine
inside
the
Calico,
Network
and
I'm
very
sure.
There
will
be
other
helpful
use
cases
of
enabled
by
this
small
PR
now.
So
that's
all
for
me.
A
F
Support
and
it's
been
there
for
some
time
there
have
been
some
concerns
about.
You
know
how
generic
is
it,
but
what
I
would
like
to
think
of
it
is
that
you
know
we
can
if
there
are
not
a
lot
of
concerns
about
the
design
or
stuff
like
that,
we
can
just
can
we
just
push
it,
and
then
you
know
if
there
are
other
vendors
who
come
up,
we
are
always
open
to
you
know,
modifying
the
implementation
and
actually
I
have
changed
it
to
make
it
generate
a
certain
level.
A
A
G
A
G
G
A
F
So
I
haven't
made
it
again,
I
think
repeat
the
same
thing:
I
haven't
made
it
a
bit
generic,
oh,
but
if
there's
a
decision
that
needs
to
be
made,
you
know
is
GPU,
be
considered
a
generic
as
a
generic
PCI
device,
or
should
it
be
considered
as
a
separate
PCI
device.
So
currently
the
way
it
will
work.
Is
you
know
if
you
are?
If,
if
you
want
to
add
some
other
vendors
GPU,
they
might
have
to
do
certain
things
in
their
in
their
device
plugin.
F
G
Absolutely
yeah
there's
a
the
main
concern
that
I
had
this.
The
is
the
generic
GPU
variable
that
that
we're
using,
if
you
could
make
it
more
generic
for
OPC
ID,
then
you're
right.
There
is
a
correlation
part
at
the
moment.
What
we're
pushing
into
this
variable
is
just
a
list
of
of
PCI
addresses
that
needs
to
be
configured,
and
we
don't
recognize
this
PCI
addresses
and
don't
correlate
them
to
a
specific
device
that
needs
to
be
configured.
G
F
I'm,
sorry,
but
I
think
you
know
as
far
as
I
see
it
we
can,
if
we
let
this
go
through,
and
we
at
least
start
supporting
the
GPUs
case.
After
that
we
can
always.
You
know
if
there
are
other
vendors
who
start
coming
in.
We
can
always
week
this
week.
This
function
ID
based
on
how
they,
how
they
want
to
support
o
how
they
want
to
support
their
PCI
devices.
We
can
always
tweak
this
or
add
more
functionality
to
it
and
I.
G
G
H
As
it
was
mentioned
before,
so
there
was
some
some
work
looking
at
work
done
in
parallel
to
address
the
issue
of
when
the
VMI
record
referred
refers
to
a
data
volume
that
includes
a
block
mode
volume.
Then
a
the
VMI
didn't
start
the
walk
around
towards
to
assure
to
the
PC
the
the
data
volume
is
later
on.
So
we
talked
to
the
two
separate
peers
and
a
winner,
the
into
one
PR,
which
is
now
in.