►
From YouTube: OKD Working Group Meeting 08-02-2022
Description
The OKD Working Group's purpose is to discuss, give guidance to, and enable collaboration on current development efforts for OKD, Kubernetes, and related CNCF projects. The OKD Working Group includes the discussion of shared community goals for OKD 4 and beyond. Additionally, the Working Group produces supporting materials and best practices for end-users and provides guidance and coordination for CNCF projects working within the SIG's scope.
A
B
To
the
okd
working
group
meeting
for
August
2nd
of
the
year
2022.
we've
got
a
link
to
the
agenda
in
the
chat
and
feel
free
to
take,
maybe
the
next
30
seconds
or
so
to
take
a
look
and
make
sure
that
you
have
everything
in
in
there
that
you
want
to
talk
about,
or
if
you
want
to
move
something
around.
Let
us
know
we
want
to
be
mindful
of
time
constraints.
If
someone
has
to
take
off
early
and
you
have
an
item,
we
can
kind
of
shift
it
around.
B
Let's
see
I
think
we
have
everyone
here.
That's
so
far
has
an
agenda
item
on
so
take
a
few
seconds,
and
let
me
know
if
there's
anything
you
need
to
change.
A
B
So,
let's
put
that
under
the
okay,
DCI,
CD
updates,
so
we'll
say:
prowl
oops.
B
All
right,
I
think
we're
good
to
go
then.
So,
let's
start
out
with
okd
release,
updates
with
the
meme
and
or
Christian.
So
take
it
away.
E
Thank
you,
Jamie
I,
don't
see
running
here
today,
so
I'll
I'll
take
it
away.
We
have
a
new
release,
actually
our
first
4.11
release
and
we
are
actually
we
haven't
released
the
product
yet.
So
there
is
one
remaining
issue:
that's
blocking
ocp
4.11
from
getting
released,
which
is
an
issue
with
ovn
under
high
load,
so
you
might
hold
want
to
hold
from
upgrading
to
that
other
than
that.
I
think
that
that
only
happens
with
really
high
load
using
ovn,
which
is
the
default
in
in
okd.
E
So
beware
of
that,
but
otherwise
I
haven't
seen
any
big
issues
because
it
was
only
cut
three
days
ago.
So
I
haven't
followed
up
too
much,
but
there
is
the
first
4.11
base
out
well
three
days
ago.
D
Yeah
Christian,
it
shows
as
failing
from
the
July
9th
version
acceptance.
D
B
D
Okay,
it's
it's
on
my
pad!
So
oh
yeah.
E
D
E
Yeah
we
currently
only
have
AWS
I
think
as
blocking
as
required
since
I'm
not
currently
involved
with
cutting
the
releases.
I
can't
tell
you
too
much
about
it:
I'll
I'll,
Circle
back
with
Vadim
and
yeah,
maybe
we'll
we'll
upgrade
all
of
those
platforms
blocking
in
the
future,
and
certainly
once
we,
we
have
a
rework
of
the
whole
build
pack
of
in
in
the
work.
So
when
that
lands,
we
will
definitely
have
more
testing
and
require
more
testing
for
release.
Oh
no
I
think
yeah.
E
I
would
kind
of
take
this
first
release
with
caution,
since
there
is
one
known
back
now
there
anyways,
hopefully
by
by
our
next
meeting
in
two
weeks,
we'll
have
a
a
version,
a
4.11
version
that
works
on
all
these
platforms.
D
I
guess
the
only
thing
is
that
there
should
be
when
you,
when
you
upgrade
to
11.
It
would
be
nice
if
the
upgrade
worked
from
the
most
recent
10
version,
because
everybody
might
have
updated
to
that.
E
E
Unfortunately,
I
can't
change
any
anything
about
that
right
now,
but
we'll
we'll
try
to
make
sure
that
for
the
next
release
we
have
upgradability
from
the
last
PowerPoint
and
definitely
on
all
the
platforms.
Excellent.
D
F
Yeah,
you
know
whenever
we
go
through
these
things,
I'm
always
thinking
back
to
like
what
we've
been
talking
about
as
like
themes
for
okd
and
whatnot,
and
one
of
the
one
of
the
big
themes
that
I
always
see
is
like.
How
do
we
help
our
community
become
more
involved
in
these
things?
And
so,
when
you
know,
when
Bruce
was
kind
of
pointing
this
out
immediately,
my
mind
went
to
like
okay.
F
F
You
know
where
we
could
like
take
a
release
and
show
these
failures
coming
out
and
start
the
video
that
way
and
then
say:
okay,
here's
what
I'm
going
to
do
I'm
going
to
show
you
how
I
go
from
this
screen
and
how
I
drill
down
into
where
the
failure
was
what
the
test
was
that
failed
and
where
the
actual
breakdown
happened.
Right
and
I
and
I
wouldn't
know
ahead
of
time
what
it
was
going
to
do
right.
So
when
we
got
there,
maybe
we
would
find
a
flake.
F
Video
and
I
know
Bruce,
like
video,
is
not
the
perfect
way
to
cover
these
things,
but
it's
just
my
thought
because
it's
something
I
can
like
turn
on
and
just
do
like
talk
while
I'm
doing
it
or
whatever
and
then
maybe
later
we
transpose
it
into
some
text
to
go
with
it
or
something
but
anyways.
That's
just
what
I
thought,
while
y'all
were
talking
well.
A
I'm,
like
while
you
were
talking
I,
was
thinking
if
you
want
to
take
this
example
that
we
have
today
I
have
time
this
week.
You
want
to
walk
through
it
and
do
that,
if
you're
the
right
person
to
record
that
and
use
blue
jeans
to
do
it
happy
if
you
want
to
book
an
hour
with
me
on
I,
don't
know
Thursday.
What's
today.
What's
today,
today's
Tuesday.
F
A
Pick
a
release
and
do
one
and
just
book
an
hour
and
I
will
I
will
come
on
and
record
it,
and
then
we
could
just
edit
it
and
put
it
up
there.
I'm
gonna
get
one
game
for
that.
B
This
is
fantastic
because
this
actually
syncs
up
with
the
discussion
before
about
creating
sort
of
intro
to
okd
video,
that
sort
of
explains
things
and
I
think
you
know
I
agree
a
video.
Is
it's
not
ideal,
but
it's
something
that
people
can
put
on
and
just
sort
of
sit
and
watch
or
have
in
the
background
or
whatever,
and
it's
a
passive
way
to
sort
of
soak
in
stuff
about
lkd
that
we
wanted.
F
So,
like
my
main
thought
here,
is
that,
like
this
style,
I
think
this
style
of
content
could
work
really
well
for
what
we're
doing
and
and
what?
What
I'm
kind
of
referring
to?
Are
these
ride-alongs
right,
where
it's
like
I'm
doing
this
stuff
all
day,
long
anyways
right?
But
if
I
could
somehow
set
aside
an
hour
or
something
to
do
a
specific
task
with
the
with
you
know
the
expectation
of
making
some
sort
of
content
out
of
it.
F
These
ride-along
videos,
then,
will
help
others,
because
it's
like
this
is
the
kind
of
day-to-day
activities
that
we're
doing.
You
know
maintaining
openshift
as
it
is
now,
but
as
we
start
to
transition
into
this
okd
ownership
maintainership,
it
would
be
great
to
be
able
to
say,
okay
well.
This
might
be
something
that
would
become
part
of
your
daily
activities
as
well.
D
A
A
Sure
the
Ride
Along
theme
so
now
we've
diverged
from
there,
but
let's
let's
place
a
bookmark
and
I
know
schools
pull
out
Bruce
for
you.
Time
is
now.
We
start
Labor
Day.
B
Excellent
well
then,
that's
let's
put
that
as
a
task
for
the
two
or
three
of
you
to
take
care
of
in
the
next
couple
of
weeks,
and
then
maybe
you
can
report
back
by
the
next
main
meeting
all
right.
Let's
move
on
to.
G
B
One
thing,
I
would
say
is
for
those
of
us
that
are
familiar
with
okd
and
and
our
varying
levels
of
familiarity
with
the
innards
look
for
stuff
in
the
documentation,
as
the
411
becomes
available
that
may
not
be
applicable
so
that
we
can
get
stuff
submitted
for
Michael
to
get
it
pulled
out
like
we're
still
finding
stuff.
Someone
found
something
like
a
week
ago,
that
was
the
the
tuning
operator
I
think
it
is
that
provides
the
fine-tuning
of
for
CPU
and
stuff,
like
that.
B
That's
not
applicable
for
okd,
that's
something
that
we
should
have
spotted,
but
it
just
it
just
never
came
up
right,
so
we're
always
finding
stuff
that
needs
to
get
pulled
out
of
the
documentation
that
gets
sent
down
all
right.
Moving
on
now
to
okd's
cicd
updates
with
Christian
and
John,
and
anyone
else
that
has
stuff
to
contribute
so
Christian.
E
All
right,
I
think
we
have
some
pretty
awesome
updates
here.
We
have
internally
started
on
on
creating,
as
we've
talked
about
earlier,
another
version
of
okd
based
on
Central
Street,
core
OS.
E
As
part
of
this
work,
we
are
building
out
text
on
pipelines
to
essentially
do
everything,
so
we
have
a
tecton
pipeline
that
is
kind
of
outside
of
the
pure
s,
cause
stream
core
OS
effort,
that
is
in
the
okg
project,
the
pipelines
repository,
which
is
kind
of
an
agnostic
pipeline
to
build
optional
operators.
For
that
we
have
provided
a
a
base
image
that
you
can
use,
that
is
also
Centos
based
and
that
can
essentially
be
used
to
build
and
to
ship
those
operators
in
the
operator
Hub
format.
E
Eventually,
we
want
to
include
you,
the
community
in
in
the
list
of
Administrators,
for
essentially
a
namespace
on
a
cluster
that
we
get
that
we
get
to
use.
Currently
we
have,
we
are
in
touch
with
the
operator
first
Cloud
to
to
get
those
resources.
We've
actually
created
a
group
there
already
a
couple
of
you
at
Jamie,
Brian
I've
included
you
in
in
that
in
that
list.
E
If
you
are
interested
in
also
getting
access
to
that
cluster,
where
we'll
essentially
have
the
ability
to
deploy
that
pipeline,
but
also
to
really
experiment
with
things,
you
know,
as
long
as
you
don't
do
anything
cheeky
there,
everything
should
be
fine.
You
should
really
be
able
to
do
things
yourself.
E
We
want
to
really
yeah,
allow
that
and
and
facilitate
that,
so,
if
you're
interested
in
helping
build
optional
operators,
please
reach
out
if
you're
not
on
the
list
already
and
I
will
see
to
it
that
you
get
admin
access
on
that
namespace
on
operate
first
as
well,
so
we'll
be
building
up
that
Tech
from
pipeline
for
optional
operators.
That's
one
of
the
efforts
then
more
currently
focus
on
the
s-course
effort
is
building
a
tecton
pipeline
to
build
s-cost
itself,
which
is
also
kind
of
agnostic.
E
If
you,
if
you
see
it
as
a
as
a
core
OS
assembler
pipeline
in
tactile
and
that's
another
repository,
we've
just
created
in
the
okd
project,
which
is
the
osup
pipeline,
we'll
be
pushing
the
core
OS
assembler
pipeline
the
actual
code
there
very
soon,
and
with
that
you
do
need
the
KVM
device
exposed
to
the
pods.
So
you
need
you
need
that
running
with
with
privileges
on
the
cluster.
That
is
something
we're
apparently
working
out
how
to
deploy
it
to
operate
first.
E
So
we
can
then
leverage
that
pipeline
there
for
our
s-cos
development.
We
will
have
internal
okd
namespace,
but
you
will
ask
the
community.
You
will
be
free
to
deploy
that
pipeline
to
the
okd
working
group
namespace
as
well,
to
build
your
own
things
if
you
like.
So
since
it's
kind
of
an
agnostic
or
a
similar
pipeline,
you
really
just
need
a
RPMs
as
inputs
and
then
you'll
get
the
reboot
images
and
the
OS
tree
in
various
formats
as
output
of
that
pipeline.
E
So
that
is
very
cool.
And
then
we
have
a
third
effort
in
the
making,
which
is
a
third
repository
in
that
okd
project
organization
on
GitHub,
which
is
how
I
think
it's
called
something
s
cause
9
rebuild,
and
that
is
essentially
that
is
kind
of
the
most
experimental
thing,
but
you
must
set
it
up
and
I.
Think
Alessandro
from
our
multi-arch
team
is
also
looking
at
it,
and
that
is
something
we
want
to
use
as
a
replacement
for
for
the
release
controller.
E
We
currently
run
in
our
prow
CI
environment,
so
that
would
allow
us
to
build
a
release
payload
without
a
problem.
That's
essentially
the
part
bundling
all
the
things
together.
The
container
images
for
the
cluster
side,
as
well
as
the
the
container
that
includes
the
OS
tree,
as
well
as
references
to
the
boot
images,
create
in
order
to
create
a
payload
for
that
from
that.
So
really
that
would
allow
us
to
run
all
of
essentially
building
all
of
openshift
without
prow
within
Tech
talk
and
I.
E
Think
that
is
really
exciting
and
I'm
really
looking
forward
to
it.
So
internally
we
are
focusing
on
getting
s-cos
built
and
we'll
be
doing
that
on
operate
first.
At
least
until
the
end
of
the
year
building
both
the
the
boot
images
on
operate
first
and
possibly
also
the
payloads.
We
might
do
that
in
prowl.
E
It
just
depends
on
on
what's
easier,
really
in
the
end,
I
think
our
developers
internally
would
love
to
use
tecton
for
for
these
things,
so
there's
a
good
chance
that
we
actually
do
the
payload
Creation
with
tacton
as
well,
eventually
yeah
so
I
did
see,
and
that's
why
I
put
it
on
the
agenda.
I
did
see
that
discussion
1288
that
Brian
opened
and
I,
because
there's
also
a
link
to
the
okd
project.
E
I
wanted
to
kind
of
reiterate
that
here
that
yeah
now
that
we
have
or
will
soon
have
resources
at
our
hands.
E
We
should
definitely
start
organizing
that
ourselves
as
a
group
a
bit
more
possibly
with
with
a
kanban
project
on
on
the
GitHub
project,
on
the
okd,
project.org
or
or
in
any
other
manner,
really
but
I'm.
All
for
for
tracking
our
tasks
and
organizers.
E
B
All
right
any
questions
or
feedback
for
Christian
on
anything
that
he
referenced
there
in
terms
of
getting
things
built
with
cicd,
pretty
exciting.
C
Well,
you
know
I'm
always
a
day
late
and
a
dollar
short,
but
I
I
have
a
feeling
that
prowl
is
not
going
to
be
going
away
anytime,
real
soon,
because
they
got
all
that
all
that
other
stuff
that's
built
into
it,
and
so
I
get
a
little
bit
of
background.
I
mean
we've
talked
for
a
long
time
about
how
we
have
not
really
had
the
ability
to
do
building
on
our
own
and
push.
You
know
our
own
source
changes
and
stuff
like
that,
because
you
know
that
openshift
repository
is
rightfully
owned.
C
You
know
by
all
the
red
hat
folks,
so
finally
about
well
I,
guess
almost
three
weeks
ago
or
whatever,
if
I
said,
you
know
what
I'm
just
really
curious,
how
all
of
this
works.
I
ran
into
a
couple
of
documents,
I
kind
of
piqued,
my
interest
a
little
bit,
and
it's
hard
to
really
over
the
last
two
weeks,
two
weeks,
Deep
dive
into
prow
and
into
how
CI
operator
works
and
how
a
lot
of
these
jobs
are
being
kicked
off
with.
C
You
know
changes
being
made
to
the
repositories
and
stuff
like
that,
and
although
it
doesn't
show
it
right
now,
those
links
I
have
a
prow
that
works,
probably
85.
Right
now,
I
can
do
like
oh
see.
I
was
able
to
build
it
and
push
new
images
and
stuff
like
that.
What
I
can't
do
yet
is
build
okb,
yet
because
of
the
whole
release
controller,
and
also
trying
to
figure
out
how
to
configure
my
cluster.
C
The
first
time
with
all
the
different
configuration
files
has
been
a
challenge,
because
the
automated
way
of
doing
it
can't
handle
a
huge
amount
of
changes
all
at
once.
So
there's
all
these
config
maps
that
have
to
be
created
that
are
created
from
all
the
job
specs
and
the
CI
specs
and
stuff,
like
that
and
I'm
still
trying
to
figure
out
how
to
do
that.
Initial
configuration,
but
I
can
do
one
of
the
twosies
with
no
issue.
C
So
what
what
this
may
allow
us
to
do?
One
it'll
give
us
you
know
much
more
visibility
into
how
we
do
testing.
Do
we
want
to
do
our
own
testing,
create
our
new
jobs
that
are
outside
of
red
hat
and
also
take
advantage
of
all
the
stuff
that
redhead's
already
giving
us
I
mean?
All
of
this
is
available.
You
know
in
the
GitHub
repositories,
including
the
source
for
how
to
build
these
build
images
and
stuff
99
of
them.
C
It's
been
really
it's
been
fun,
frustrating
annoying
and
exciting
at
the
same
time,
because
every
time
I
get
a
little
bit
closer.
There's
like
oh
well.
How
do
I
do
this
now
and
then
you
have
to
go
through
and
dig
through
everything
and
it's
not
well
documented
I.
Think
the
people
who
know
it
know
it
really.
Well,
it's
like
well,
it's
all
you
know
we
all,
or
we
already
know
how
it
works.
C
So
it
just
works,
but
yeah
I
think
in
a
in
a
couple
of
weeks,
I'm
hoping
to
have
a
hoping
to
be
able
to
build
at
least
okd10
4.10.
You
know
or
come
close
to
it
then
figuring
out.
How
do
we
pull
in
changes
from
you
know
the
openshift
repository
into
while
I'm,
calling
it
The
OCD
experience,
repository
in
rebuild
and
stuff,
but
I'm
doing
it
for
my
own
benefit.
You
know
just
from
my
own
knowledge,
I'm,
hoping
that
it
can
be
a
benefit.
You
know
as
we
go
forward.
C
You
know
with
the
new
build
process
that
Christian
is
talking
about.
Maybe
there's
some
way
that
we
can
integrate
things
so
that
we
can
do
automated
builds
based
on
you
know
a
push
going
into
the
repository
or
something
so
maybe
we
figure
out
how
to
do
a
new
job
that
goes
to
the
you
know
to
the
new
build
process,
as
opposed
to
the
current
build
process.
So,
but
it
gives
us
something
that
you
know
well
I
control
right
now,
but
ultimately
you
know
we
may
be
able
to
control.
C
Currently,
it
runs
in
my
open
shift,
cluster
downstairs
or
okd
cluster
downstairs
in
my
basement.
So
if
this
turns
into
something
real
would
have
to
find
someplace
else
to
conceivably
put
it
but
yeah.
That's
it
and
I
put
the
two
links.
I
mean
there's
not
a
whole
lot
in
in
either
of
them
right
now,
but
you
know
prowl's
up
and
working.
C
That
I
mean
the
only
thing
is
right
now,
that's
what's
stopping
me
from
testing
I
can't
do
ede
testing
because
I
don't
have
you
know
the
resources
to
do.
You
know
a
bunch
of
AWS
bills,
but
we
could
also
put
this
on
AWS
or
you
know
the
the
actual
cluster
in
AWS
or
put
it
in
Google
or
I.
C
Don't
think
we
could
do
it
and
I
just
went
rolled
out
of
my
head
where
we're
putting
the
current
stuff,
because
prow
really
requires
cluster
admin,
access
and
so
does
CI
operator,
so
they
probably
aren't
going
to
want
to
run
prow
and
stuff
in.
E
So
on
on
operate
first,
they
actually
run
Pro
already,
so
we
might
be
able
to
integrate
with
their
instance
or
eventually
we
will
right
now
we'll
we'll
just
get
namespaces
on
existing
clusters,
but
eventually,
once
we
have
the
the
s-cost
version
of
okd
the
folks
behind
operate
first,
which
is
our
office
of
the
CTO,
they
are
interested
in
spinning
up
an
okd
cluster
for
us
that
we
will
then
for
their
control
and
own
full
name
so
yeah.
E
That
is
a
few
months
out
still,
but
if,
if
until
then,
we
can
deploy
that
into
just
the
namespace
because
they
already
have
Pro
on
that
cluster.
So
I'm
not
sure
what
exactly
you
need
in
terms
of
rbac
and
and
so
on,
but
we
might
be
able
to
get
you
to
get
that
thing
deployed
on
operate
first
already
well
like
by
let's
say
next
week
when
we
have
all
of
our.
C
E
It's
very
valuable
that
we
tackle
this
from
just
you
know
different
perspectives
here,
because
eventually
we
will
for
the
s-cost
thing.
We
will
probably
do
it
internally
with
a
in
Pro,
with
a
release,
controller
and
PCA
operator
for
for
the
new
okd
on
s-cos
builds,
but
we
will
still
and
and
that
even
more
with
the
new
operate
first
infrastructure
we
still
want
to
do
more
experiments
and
custom,
builds
and
sure
I.
Think
elmico
is
very
interested
in
that
too.
E
So
we'll
have
that
ability
and
the
more
we
try
it
the
more.
We
learn
the
different
ways
that
get
us
to
to
a
payload,
I
think
yeah,
the
better
we're
all
off,
and
we
all
benefit
from
that.
So
it's
really
cool
to
hear
that
you
got
proudeployed
and
that
you
got
it
85
working,
that's
pretty
amazing,
already
so
I'm
looking
forward
to
to
seeing
that
and
yeah
taking
a
look
at
what's
happening
there.
B
Hey
John,
could
you
do
a
little
like
you
know,
five
minute,
video
walkthrough
or
something
like
that
of
what
you
have
I
think
that
would
be
really
educational
for
folks.
C
Once
they
get
a
little
further
along
yeah,
the
the
actual
build
of
prow
itself
is
is
pretty
easy.
You
don't
want
to
have
a
cluster,
it's
the
integration.
Now
of
all
the
of
all
the
openshift
release
stuff,
you
know
because
pretty
much
almost
everything
is
needed
is
in
the
open
shift,
release
repository
on
GitHub
and
and
then
that
is
where
all
the
all
the
meat
of
the
product
and
I'm
still
fighting
my
way
through
parts
of
it.
So.
A
A
Yeah,
ultimately,
yes,
well
anything
you
can
parcel
off
or
you
need
help
with
for
the
rest
of
the
community
that
you
know,
or
is
it
still
pretty
much
an
individual
active
value.
C
Right
now,
right
now
it
is
I
mean
I've,
pinged,
Vadim
I
haven't
had
the
pink
Christian
yet,
but
I've
come
close
a
couple
of
times,
but
I'm
able
to
I've
been
able
to
work
through
it,
but
yeah
as
soon
as
I
as
soon
as
I
have
enough
knowledge
to
share
that's
beyond
the
basics.
Right
now,
I'll
be
happy
to
okay.
A
And
and
as
I
said
with
Mike
I
would
be
happy
to
record
that
with
you
and
ask
silly
questions
that
move
us
forward
a
little
bit
and
then
get
that
available.
So
when
you're
ready
just
let
me
know
and
I'm
getting
back
into
the
video
creation
mode
here.
So
as.
C
G
F
John
I
will
I
got
a
couple
thoughts,
just
a
feedback
here
and
I.
You
know
first
I'll
Echo
what
Shri
and
Christian
are
saying
like
100
worthwhile,
like
it's
awesome
to
hear
you
made
it
so
far,
so
first
thought
about
tests
that
we
should
probably
cover
or,
if
you're
looking
for
like
a
Target.
F
There
is
a
series
of
tests
that
originated
in
kubernetes
that
are
like
kind
of
the
the
base
conformance
tests.
We
also
run
those
on
openshift
container
platform.
That's
probably
like
a
really
safe
place.
To
start
in
terms
of
like
an
initial
payload
is
those
like
basic
conformance
tests
and
whatnot
yeah.
Just
because
we'd
have
coverage
with
like
open
shift
and
with
you
know,
Upstream,
then
the
second
thought
about,
like
the
the
mirrored
repositories
that
you
have
like
from
you
know
other
locations.
F
We
have
a
few
Bots
that
we
use
internally
to
sync
from
kubernetes
repos
to
open
shift
repos.
We
could
totally
share
those
and
and
repurpose
them
and
figure
out.
If
there's
a
way,
we
could
use
them
in
okd
too.
Oh,
that
would
be
cool
yeah
and
they
can
run
off
GitHub
actions
or
they
could
run
off
of
prow
like
there's
a
lot
of
different
ways
to
trigger
them,
and
then
the
light.
F
The
last
thought-
and
this
might
be
something
that
would
be
useful
for
okd
in
the
future
here-
is
that
we're
currently
developing-
and
this
is
it's
open
source
right
now,
so
you
could
get
it
we're
developing
a
certification
tool
for
openshift,
and
this
is
something
that
you
could
run
on
like
an
open
shift.
You've
deployed
to
determine,
if,
like
it,
meets
Red,
Hat's,
conformance
and
whatnot,
and
this
tool
is
it's
open
source
and
everything
and
it
bundles
some
of
the
conformance
tests
from
Upstream.
F
So
there's
the
idea
that
wants
this
and
I
think
it's
getting
released
this
week
or
next
week.
This
is
a
tool
that
we
could
actually
use
outside
of
prow.
If
we
wanted
to
once
we
deploy
an
okd
cluster,
you
could
run
this
certification
tool
to
see
if
it
passes
like.
Is
this
cluster
configured
in
the
proper
way,
and
it
would
give
you
a
report
of
like
all
the
tests
that
ran
on
it
and
everything-
and
you
know
I-
can
share
more
about
that
if
people
are
interested
but
anyways,
those
are
just
the
thoughts.
C
I
think
I
think
there's
one
that
that
I'm
not
sure
about
okd
machine
OS
I
think
that
one
has
an
issue
because
of
the
requirement
for
KVM,
but
I
think
we
shipped
that
out
for
build,
so
I
haven't
really
played
with
that.
Yet
so
anyway,
that's
my
that's
my
update
in
my
my
my
play
thing
for
the
moment.
B
That's
fantastic
all
right.
Let's
move
on
now
to
f-class
updates
and
we
appreciate
your
patience
Timothy.
While
we
are
going
through
that,
I
probably
should
have
put
you
earlier
in
the
agenda.
H
Oh
good,
oh
good,
can
you
hear
me?
Yes
all
right?
Yes,
all
good,
and
it's
kind
of
related.
So
right
now
like
if
we're
talking
of
pure
Federal
questions,
they
aren't
that
much
of
interest.
For
this
week
we
are
looking
already
at
Fedora
37
changes
which
is
upcoming
release
during
a
couple
of
months,
although
if
federal
court
reside,
but
it's
probably
not
going
to
land
right
now
in
okg
in
a
release.
So
it's
a
little
bit
for
the
future.
H
Oh
yes,
there's
one
thing:
we
are
changing
the
compression
for
the
initial
form
FS
to
that
STG
that
it's
so
that
might
impact
you
if
you're
doing
that
for
PC
boot,
but
I,
don't
think
okay
G
does
that
much
pixie
boot,
so
that
should
probably
be
transparent
for
okay
default
and
yes,
apart
from
that,
most
of
my
time
has
been
working
on
escorts
changes
so
making
sure
that
exos
actually
works
and
bills
and
that
we
have
all
the
packages
available
for
you
perfectly
in
the
open
so
cryo
and
things
like
that.
E
Timothy
since
I
have
you
here
earlier,
the
question
came
up
where
the
the
Centos
stream
RPM
repositories
live,
because
we
are
now
rebuilding
escorts,
as
you
yeah
well
we'll
read
about
that
tomorrow.
But
if
you
have
a
link
to
those
repository
the
jam
repos
and
could
share
those,
if
you
have
it
at
hand,
that
is
that
would
be
awesome.
Otherwise,
I
think
it
can
also
wait
until
tomorrow.
H
Yeah,
so
let's
keep
some
context
around
that,
so
the
main
thing
for
building
stream
Center
stream
forests
like
95
of
if
it's
not
there.
Well
all
the
packages
are
just
directly
pulling
both
from
the
central
stream
read
books
like
the
central
stream
report
that
already
exists
that
are
part
of
the
saturated
stream
project.
H
Then
there's
like
a
couple
of
packages
that
we
need
on
the
side,
which
is
essentially
cryo,
openshift
OC
and
the
couplet
itself.
So
among
those
packages
like
openshift
is
already
no
openshift
open.
V
switch
is
already
part
of
a
special
interest
group
innocentris
project,
so
they
have
their
own
seek
and
they
publish
their
own
packages
there.
H
So
we
use
that
and
then
we've
got
a
couple
of
missing
packages,
so
couplet
cry
you
OC
and
zoos
right
now
are
just
not
available
anywhere
because
we're
setting
this
up
so
they
are
not
ready
yet,
but
have
been
talking
with
the
folks
me
and
doing
the
work
on
those
to
try
and
set
up
something
and
to
have
them
available
publicly.
So
right
now
it's
likely
that
we're
going
to
reuse
corporate
Federal
corporate
projects
to
host
them
and
as
soon
as
they
are
available,
they
will
be
used
by
listen
to
a
stream
manifest.
E
Okay,
so
if
I
understand
you
correctly,
we'll
we'll
pull
from
copper
eventually
or
was
it
the
suse
prompt?
What
is
it
built
open,
build
thing.
H
E
Okay,
very
cool,
but
that
sounds
very
promising
once
you
have
an
update
on
that,
could
you
ping
ping
me
so
we
get
awesome.
Thank
you
very
much.
B
Awesome
very
cool,
any
questions
or
feedback
on
that.
A
I'm
just
going
to
mention
again
at
the
Central
West
Dojo
at
confus,
we're
gonna
have
a
bit
of
a
presentation
on
S
class
and
okd
and
and
start
socializing
that
so,
if
anybody's
going
to
Dev
Comp
US,
which
is
in
August
August
I,
think
Wednesday
is
this
sent
to
West
Dojo
at
Dev,
Comp
US.
B
All
right,
let's
move
on
now
to
documentation
subgroup
updates
with
Brian.
G
So
promoting
last
Tuesday,
so
we're
still
waiting
on
a
couple
of
updates
from
the
CERN
folks
and
also
for
the
Sno
update.
G
G
G
So
there
is
an
effort
about
to
start
and
to
look
at
that
front,
page
experience
what
needs
to
be
on
there
and
then
also
going
into
for
different
users,
you
user
groups
and
for
just
someone
who
wants
to
use
it,
install
it,
but
then
also
as
aniko
mentioned
as
people
want
to
admin
and
begin
to
experiment
with
it
some
more
guidance
on
how
people
can
troubleshoot
and
how
they
can
start
and
get
to
understand
a
little
bit
more
around
there.
So
that
is
a
project.
That's
that
we're
about
to
start.
G
We
are
still
trying
to
move
everything
to
our
new
okd
project,
GitHub
repo,
it's
a
bit
like
herding
cats
trying
to
get
everyone
lined
up
because
we
need
to
do
the
the
moves
from
the
open
ship.
So
we
need
the
people
who
can
actually
transfer
and
then
we
need
to
do
all
of
the
updates
around
the
website,
because
obviously
we
now
host
okd.io
in
GitHub
pages.
So
we
need
to
make
sure
that
as
we
move,
the
Repose,
the
the
DNS
entries
are
all
updated
as
well.
G
So
that
will
happen
when
we
get
everyone
lined
up,
and
that
will
happen,
but
yeah
we're
still
trying
to
do
that.
G
G
So
we
are
looking
to
revamp
that
guide,
section
and
I'm
going
to
take
the
lead
on
that
and
to
actually
get
that
some
of
it
will
be
moving
to
blogs.
But
then
the
question
is
what
we
want
to
do
with
that.
So
that's
an
ongoing
project
and
then
I
I
think
the
other
thing
that
was
started
looking
at
is
now
that
we're
beginning
to
sort
of
get
the
community
enabled
a
bit
more.
G
As
Christian
mentioned,
we
want
to
start
looking
at
how
we
actually
govern
the
community,
how
we're
going
to
look
after
the
okd
project
repo
as
a
community,
and
things
like
that,
so
what
I've
done
is
I've.
Actually,
if
you
go
into
that
repo
you'll
see
that
there
is
actually
now
a
project
in
there
I'll
just
post
that
into
the
chat,
so
I've
started
sort
of
experimenting
with
creating
a
board
on
there
and
what
we're
looking
to
do
is
we're
looking
to
repurpose
the
documentation
working
group.
G
So
that's
the
working
group
that
meets
the
every
other
Tuesday.
This
is
the
main
working
group
where
we
actually
sort
of
bring
everything
together
and
on
the
alternate
Tuesday.
The
documentation
work
group
needs
and
we're
going
to
repurpose
that
really
to
focus
on
managing
the
community
and
the
GitHub
projects.
The
boards
and
the
pull
requests
issues
so
that
that
working
group
is
going
to
become
more
the
community
working
group
and
will
manage
the
community
assets
and
moving
things
in
the
project,
so
we're
I'm
going
through
trying
to
create
that
process.
G
G
So
I
am
looking
for
any
input
into
how
we
want
to
do
this.
Who
should
have
access
to
sort
of
pull
pull
request
and
do
we
want
to
adopt
standard
licensing,
standard
copyright
across
all
our
repos
and
do
we
want
to
configure
the
repost
and
not
allow
direct
commits
and
pushes
to
the
main
branch,
all
of
these
sort
of
things
that
we
certainly
have
to
start?
G
Looking
at
now
that
we've
got
our
own
on
guitar
bog,
so
we
need
to
start
defining
all
that
and,
as
you
can
see,
it's
in
the
new
column
on
the
project
board
so
and
that
will
be
moving
across
the
board
as
we
actually
pull
that
in
and
that'll
all
be
written
up
on
okd.io.
So
that's
where
all
the
process
will
be
written
and
available
for
everyone
to
see
and
understand
how
it
all
works
and
Jamie
and
I
now
have
access
to
Twitter.
G
So
we
can
tweet
I'm
going
to
start
once
we
get
things
around
the
the
community
board.
I'll
start
tweeting
it
there,
as
we
open
up
I,
still
want
to
do
a
community
event.
Once
we
get
the
pipeline
and
The
Operators
a
little
bit
more
established
to
try
and
get
some
participation
there,
so
we
will
start
tweeting
sort
of
community
activity
and
events
that
are
coming
up.
G
I
think
that's
everything
that
we
talked
about.
I.
A
Think
on
the
Twitter
side
of
things
Christian
you
mentioned,
we
have
a
411
release.
Should
we
be
tweeting
about
it
and
sending
out
a
note
that
it's
there
or
would
you
prefer
that
we
wait
until
ocp
comes?
Does
it
matter
to
you.
E
It's
not
really
that
I'd
like
ocp
to
be
there
first,
it's
more
like
we
have
a
known
issue
with
the
current
with
that
release,
so
I
would
rather
wait
until
ocp
comes
out,
because
that
means
the
the
issue
has
been
resolved
and
then
we'll
get
a
rebuild
after
that
yeah
I'd,
rather
wait.
I
wouldn't
want
anybody
to
upgrade
not
be
disappointed
when
they
run
this
in
a
high
yeah.
We.
B
Will
be
tweeting
for
the
410
ones
now
that
I,
you
know,
I
got
the
credentials
last
week,
so
I'll
be
sure
to
tweet
something
out
for
all
the
410
of
them.
B
C
D
G
B
E
So
it's
important
to
note
that
this
isn't
an
upgrade
blocker,
so
it's
not
gonna
lock
you
in
at
least
not
as
far
as
we
know,
even
if
you
go
through
that
version,
eventually,
I
I'd,
like
for
the
next
release
to
still
we
tested,
be
upgradable
from
the
latest
4.10
release.
E
I
think
we
can
still
decide
what
upgrades
to
test.
Usually,
we
obviously
go
from
n
minus
1
to
n,
and
this
time
we
probably
want
to
wanna
do
both
from
the
and
minus
one,
which
is
definitely
profit
11,
one
now
and
also
n
minus
two,
which
is
the
latest
four
point,
I'll
see
what
we
can
do
about
that.
That
is
another
thing
that
eventually
in
second
half
of
well
in
the
last
quarter
of
this
year,
we
want
to.
E
We
want
to
also
revisit
once
we
have
the
core
OS
layering,
essentially
merged
into
into
our
code
base.
This
part
for
that
we'd
like
to
switch
over
to
making
the
F
cause
Okay
okayd
machine
OS
content
image
through
layering.
Instead
of
making
a
whole
new
RPM
os3
compose.
E
That
is
when
it's
officially
gonna
land,
but
the
support
for
it
well
fully
gonna
land,
which
includes
actually
doing
these
layering
builds
within
the
cluster.
That
is
something
we
don't
actually
require
for
for.
Switching
over
to
that
format,
we
just
need
support
for
consuming
the
new
cluster
added,
the
new
format
in
the
machine,
config
Daemon,
and
that
will
land
in
4.12
as
its
currently
plan.
E
So
I
I,
don't
think
we're
going
to
miss
that,
because
they're
already
pretty
far
along
just
that,
we
yeah
there's
kind
of
two
steps:
first
to
be
able
to
to
consume
the
new
format
and
then
also
be
able
to
build
the
new
the
new
or
build
your
own
derived
images.
That's
what
the
feature
is
for
on
your
cluster,
which
we
don't
actually
need,
because
we
will
be
building
those
images
on
on
our
built
in
front.
B
Excellent,
so
let's
move
into
any
last
thoughts
here,
there's
a
an
event:
there's
Devcon
for
USA
Defcon
us
and
the
Centos
Dojo
on
August
17th.
Folks
should
note
that
we've
got
our
tasks.
I
had
a
quick
question
of
oh
go
ahead.
B
I
had
a
quick
question:
does
anyone
have
any
thoughts
about
how
we
might
I
don't
know
what
the
timetable
is
for
incorporating
Gateway
API
into
ocp?
Is
there
a
way
that
okd
could
leverage
itself
as
being
ahead
of
the
game
in
terms
of
Gateway
API
and
I'll
probably
provide
a
link
for
folks
that
don't
know
what
we're
talking
about
so
maybe
Christian
or
someone
else
can
chime
in
on
this.
E
At
least
this
is
I,
while
I
haven't
discussed
this
with
with
anyone
on
the
team,
yet
I
do
think
this
is
the
kind
of
experimentation
where
we
want
to
enable,
specifically
with
with
this
new
kind
of
being
able
to
build
your
own
payloads
and
things
like
that.
We'll
hopefully
enable
exactly
this.
We
have
a
couple
of
projects
like
this
I
think
Gateway
API
is
one
of
the
the
big
ones
we
want
to
try
as
early
as
possible.
E
Definitely
there's
more
with
cluster
API,
which
is
the
team
Mike
is
on,
and
those
are
the
things
I'm
definitely
looking
forward
to
building
experimental
payloads
for
so
we
can
really
get
a
very,
very
early
testing
of
that.
Currently
it's
hard
because
we
don't
have
the
infrastructure
to
build
these
one-off
payloads,
but
we
are
kind
of
using
this.
This
tentos
core
OS
MVP
work
now
to
build
a
pipeline
that
is
agnostic
and
that
can
just
be
used
to
build
the
things
with
whatever
content.
E
B
Yeah
for
folks
that
are
that
are
watching
at
home
that
don't
have
access
to
the
link
right
away.
Gateway
API
is
is
sort
of
the
next
generation
of
well.
It
builds
around
Ingress
and
provides
even
more
functionality
than
the,
and
it
just
went
beta
I
think
last
week
or
the
week
before
and
real
interesting
stuff.
There
does
anyone
have
anything
else
before
we
end
the
meeting.
F
I
just
wanted
to
tack
a
comment
onto
the
end
of
that
discussion
and-
and
you
know
I
guess-
maybe
this
is
like
partially
for
the
okd
community,
but
also
partially
for
the
red
Hatters
who
are
here
I.
Think
we
probably
need
to
put
some
time
into
figuring
out
like
if
we
can
get
these
experimentations
working
in
the
okd
community.
F
What
does
it
look
like
to
have
a
Delta
that
we
take
back
to
Red
Hat
because,
like
I'm,
very
curious
to
see
like
if
we
could
experiment
in
this
community-
and
we
do
things
on
The
Operators
that
are
part
of
openshift,
then
like
okay?
There
has
to
be
a
way
for
us
to
kind
of
formalize,
bringing
that
back
into
openshift,
proposing
it
to
ocp
and
I.
F
Think
there's
another
way
to
go
about
this
too,
which
is
like
through
olm
and
everything,
so
I
think
I
I'm,
very
curious
to
see
if
people
in
the
okd
community
start
to
innovate
things
through
the
creation
of
new
operators
that
we've
never
seen
before,
and
that
maybe
starts
to
look
a
little
bit
more
like
in
Fedora
when
a
user
brings
a
new
RPM
for
a
new
type
of
you
know
functionality.
So
maybe
there's
some
forging
work.
We
can
do
there
there
but
anyways
a
topic
to
continue
talking
about.
A
E
Kind
of
tangential
to
the
operator
effort,
there's,
apparently
a
very
interesting
enhancement
if
yeah,
if
you'd
like
to
take
a
look
at
platform
operators
so
in
the
future,
I
think
what
what
openshift
is
going
to
do.
It's
going
to
be
more
modular,
more
optional
operators
and
you
pick
and
choose
your
own
kind
of
setup.
E
D
A
So
I
was
just
going
to
put
on
the
table:
I'm
I'm
booking
space
for
kubecon
North
America
for
the
openshift
commons
and
I've
got
a
room
that
I
can
use
for
an
okd
working
group
meeting
or
face-to-face
social.
If
people
are
coming
so
I
will
let
you
know
next
week
once
the
contract
is
signed.
A
But
I
was
thinking
of
doing
something
like
in
the
morning
hosting
an
operator
framework
kind
of
in
this
spare
room
meeting
and
then
an
okd
working
group
meeting
and
I
will
not
be
there
because
I'll
be
running
around
in
the
other
room.
Doing
the
commons
things,
but
I
think
you
guys
can
handle
that
and
then
in
the
evening,
coming
together
for
some
sort
of
social
thing
on,
and
that
would
be
the
25th
which
is
the
of
of
in
not
until
October.
A
So
that's
I'm
working
on
that,
because
I'm
trying
to
get
some
space
for
a
mini
operator
con
as
well
where's
it
going
to.
C
D
B
D
B
B
All
right
thanks
so
much
for
your
time,
folks,
be
sure
to
check
the
meeting
notes
and
check
the
new
board
that
Brian
has
because
that's
where
stuff's
going
to
get
put
up
now,
thanks
to
everyone
for
all
your
hard
work,
and
we
will
talk
to
you
next
time.
Take
care.