►
From YouTube: Introducing OKD Streams - Michelle Krejci Christian Glombek and Diane Mueller OKD Working Group
Description
Introducing OKD Streams
Guest Speakers:
Michelle Krejci
Christian Glombek
Diane Mueller
https://okd.io
OKD Working Group
A
Welcome
everybody
back
again
to
another
okd
working
group
briefing.
Today
we
have
with
us
michelle
and
christian
and
myself
and
we're
going
to
talk
and
try
and
introduce
some
new
concepts
around
something
called
okd
streams,
and
so
I'm
going
to
let
michelle
and
christian
introduce
themselves.
I'm
diane
mueller,
I'm
the
chair,
co-chair
of
the
okd
working
group
and
the
director
of
community
development
for
red
hat
in
the
hybrid
cloud
platform
and
michelle.
B
Yeah
so
today
we're
excited
to
be
talking
about
a
concept
that
has
been
really
filling
our
imaginations.
We'd
like
to
talk
about
okd
streams.
We've
been
thinking
about.
How
do
we
unlock
innovation?
How
do
we
get
better
release
signal?
How
do
we
enable
folks
to
get
involved
with
okd,
and
we
want
to
make
this
easier
and
easier,
and
the
concept
that
has
really
connected
with
us
as
we
explore
integrating
with
other
openshift
or
open
source
technologies?
B
So
previously,
when
we
referred
to,
okd
has
been
a
traveling
version
of
openshift
running
on
a
leading
edge
version
of
rel,
which
is
f
cos,
and
that's
that's
been
the
sort
of
singular
pipeline
that
we
think
about
when
we
think
about
okd
and
that's
been
useful
for
a
lot
of
you
know
for
a
lot
of
folks,
it's
been
useful
for
us
to
get
signal
early
on,
to
make
contributions
up
into
rel
folks,
have
really
enjoyed
making
use
of
okd
and
exploring
its
functionality,
and
some
of
some
folks
have
even
used
in
production,
but
we
want
to
create
sort
of
more
varieties
and
and
make
it
easier
for
folks
to
to
play
around
with
all
the
different
configurations
and
the
many
many
upstreams
that
come
in
from
to
openshift.
B
So
we're
looking
to
our
near
future,
of
enabling
more
and
more
of
these
builds
and
pipelines
to
explore
some
some
ideas
that
we
have
and
get
them
out
to
the
community,
so
see
what
folks
think
about
it
to
let
other
community
members
build
their
own
pipelines.
So
we're
really
excited
about
all
the
things
that
we
can
do
and
before
we
go
a
little
bit
more
into
what
we
can
do.
Let's
just
explore
this
present
a
little
bit
more
christian.
Could
you
explain
a
little
bit
more
about
tecton.
C
Absolutely
yes,
so
techtron
is
a
a
cd
system
essentially,
and
it
enables
you
to
run
tasks
and
pipelines
in
a
in
a
very
composable
and
customizable
manner,
and
this
really
fits
in
very
very
nicely
with
our
okd
streams
effort,
because
really
the
the
main
thing
about
okd
streams
isn't
necessarily
the
artifacts
that
are
produced,
but
it's
really
the
the
pipeline
that
builds
it
and
that
anybody
can
really
build
their
own
version.
Their
own
stream
of
okd.
C
With
this
pipeline
we've
been
working
on
on
a
tecton
pipeline
to
enable
this
and
yeah
they
are
tacked
on
enable
customizable
builds.
So
you
have
tasks
and
pipelines
that
you
can
customize
with
parameters
or
even
the
even
the
tasks
within
the
pipeline.
You
you
can
exchange
or
move
around
and
add
your
own
and
customize
them
that
way.
They
are
reusable.
They
run
essentially
anywhere.
You
can
run
them
on
a
on
an
open
shift
or
okd
cluster.
C
You
can
run
them
locally
in
kubernetes
in
podman,
for
example,
or
you
can
run
them
on
a
vanilla,
kubernetes
cluster.
The
only
thing
you
really
need
is
a
cluster,
a
kubernetes
cluster.
You
install
tecton
on
it
and
then
you
can
run
and
run
your
build
pipeline
pipelines
there.
It's
expandable
it's
essentially,
you
can
add
tasks
from
from
other
projects
to.
B
C
Pipeline
ex
expanding
them
and
really
customizing
them
that
way
they
are
standardized.
So
it's
it's
all
once
you
know
one
pipeline
and
one
task.
You
already
know
the
know.
The
concept
of
of
tactone,
which
has
a
tecton
pipeline,
essentially
consists
of
tasks,
and
the
tasks
consist
consists
of
steps.
So
the
the
step
is
really
the
the
primitive
here
and
whether
you
define
multiple
steps
within
a
task
or
multiple
steps
within
a
pipeline
is
is,
is
up
to
you
in
the
end.
C
But
there
is
this
very,
very
standardized
manner
in
which
these
tasks
are
created
and
written.
So
it's
it's
really.
The
the
learning
curve
is
really
great
there.
It's
really
easy
to
get,
get
it
up
and
running
and
it
has
quickly
become,
or
it
is
quickly
becoming
the
standard
for
running
pipelines
or
for
for
doing
build
pipelines,
and
it
is
scalable,
of
course,
it's
a
it's
a
kubernetes
native
project.
C
You
can
run
it
on
on
a
single
node
cluster
or
even
in
a
in
containers,
or
you
can
run
it
on
clusters
with
hundreds
or
thousands
of
nodes.
B
Yeah,
I
I
think
I
really
like
this
phrase
building
blocks,
because
this
is
what
gets
you
know.
A
lot
of
us
excited
sort
of
thinking
about
what
could
be
possible,
because
really
this
space
is
innovating
so
quickly
that
we
are
there's,
there's
no
way
that
we
can
keep
up
without
and
get
the
kind
of
feedback
that
we
want
from
folks.
With
our
you
know,
current
release
cycle.
B
So
an
idea
that
we've
been
exploring
is
is
what,
if
we
did
all
of
our
work
in
okd
streams
we
within
openshift
and
and
from
from
other
folks
who
are
looking
to
build,
maybe
a
small
small
version
of
like
a
you,
know:
small
core,
okd
or
a
strip,
a
a
a
paired
down
version
and
with
all
these
building
blocks,
you
know
we
can
really
sort
of
think
about
playing
with
and
reconfiguring
all
kinds
of
different
versions
of
features
and
feature
configurations
and
components,
and
it
allows
us
to
be
a
lot
more
flexible
and
composable,
and
this
is
a
this
is
a
a
sample.
B
This
is
not
something
we're
actually
exploring,
but
this
is
this
is
something
or
we
are
exploring
we're
not
committed
to
it.
You
know
what
would
it
look
like
for
us
to
to
change
our
our
release
river
and
in
invite
more
innovation
make
it?
You
know
really
invite
folks
to
to
play
with
these
concepts
along
with
us.
B
This
is
this
is
an
example
of
what
that
might
look
like
it's
important
to
note
why
we
would
consider
doing
this.
There
are
really
three
problems
that
at
least
I
I
spent
a
lot
of
time,
thinking
about
and
other
folks.
B
My
colleagues
spend
a
lot
of
time
thinking
about,
and
that
is
how
do
we
improve
our
release
signal?
How
do
we
know
when
something
you
know
will
is
ready
to
ship
and
techton?
If
we
can
imagine
running
all
of
our
okd
streams
through
tekton,
build,
for
example,
that
that
would
give
us
a
dashboard
of
you
know.
Imagine
if
every
pipeline
had
an
end-to-end
test.
That
said,
you
know
when
these
features
pass.
You
know
this
is.
B
That
means
that
we
have
achieved
what
we
wanted
to
achieve
and
imagine
if
we
could
look
at
that
dashboard
for
release
readiness
signal,
then,
from
from
you
know
the
standpoint
of
folks
who
are
wanting
to
sort
of
cut
a
stable
release,
they
could
sort
of
look
at
what's
what's
ready
or
what's
in
the
literally
what's
in
the
pipeline.
B
That's
a
really
exciting
concept
of
literally
what
is
in
the
pipeline,
and
you
could.
You
could
see
that
view
at
any
time
and
that
that
would
be
a
true
release
signal
for
us
and
and
it's
really
important
that
we
get
features
into
the
hands
of
our
customers
and
our
partners
much
faster
than
we
have
been
able
to
in
the
past.
This
is
something
we
think
about
a
lot.
How
do
we
get
and
act?
B
You
know
folks
to
really
touch
and
feel
what's
what's
happening
and
get
their
feedback
and
be
able
to
incorporate
it
in
and
with
with
okd
streams.
It's
all
out
there
in
the
open.
Everyone
can
see
sort
of
what
we're
running,
what
we're
building
on
tecton
and
with
operate
burst,
and
so
we
could
really
get
that
sort
of
engagement
from
folks.
B
You
know
soon
as
soon
as
possible,
which
is
you
know,
enables
more
innovation,
and
currently
this
space,
like
I
said,
is,
is
moving
so
so
quickly.
We,
you
know,
we
need.
We
need
to
tighten
those
feedback
loops
as
much
as
possible
and
then
in
terms
of
how
we
enable
the
the
engineers
and
other
folks
around
us
to
do
their
best
work
and
to
feel
the
freedom
to
innovate
and
to
demo
their
work.
You
know
the
open
source
structure
gives
us
those
abilities.
This
is
a
sort
of
picture
of
a
guild.
B
C
If
I
may
jump
in
here
really
quick
on
the
slide
before
you,
you
had
a
a
bar
in
the
graph
that
symbolizes
prowl
and
pro
essentially
in
our
current
way
of
doing
things
mostly
has
the
functionality
of
of
a
potential
or
of
tacton
pipelines
and
proud
does.
Does
that
job
really?
Well
too?
It
is
just
that
in
pro,
the
learning
curve
is
very
steep.
It's
very
hard
to
get
things
going.
C
Proud
doesn't
have
a
big
community
like
tecton
and
we've
heard
over
and
over
again
from
our
okd
community
as
feedback
that
that
is
the
main
blocker
to
really
getting
involved
and
setting
up
in
your
own
prow
system
is
not
something
that
is
oftentimes
done
in
the
community
or
in
the
in
the
broader
kubernetes
community.
Even
it's
more
of
a
really
specialized
project.
C
It
it
pro
is
the
project
that
today
runs
kubernetes,
ci
and
openshift
ci,
but
from
our
feedback
we
have
heard
that
techtron
is
just
so
much
more
approachable
and
usable
and
in
our
the
way
we
do
things
today
they
essentially
have
to
land
in
the
code
base
and
then
they
are
tested
or
they
have
to
land.
In
a
pr
and
then
they
are
tested
in
pro-
and
you
can't
really
do
a
lot
of
testing
without
interfering
with
other
engineers
or
other
themes,
work
that
that
work
on
openshift.
C
So
what
using
cacton
would
open
us
up
to
is
really
enabling
anyone
whether
it
be-
and
we
have
a
lot
of
internal
potential
customers
for
that
as
well,
and
interest
really
enabling
teams
or
customers
or
anybody
to
build
their
own
thing,
their
own
version
stream
of
okd
with
a
specific
feature
or
a
a
specific
code
change.
C
They
want
to
test
and
essentially
put
that
in
a
in
an
okd
payload
that
they
can
then
deploy
and
test
in
in
yeah,
essentially
on
their
own
systems,
without
having
to
interfere
or
wait
for
retired
engineers
to
to
give
an
okay
to
test
for
their
code
change.
So
really
this.
This
would
enable
anybody
to
to
test
and
deploy
their
own
changes
and
integrate
that
with
the
rest
of
the
openshift
code,
phase.
B
Yeah
thanks
for
highlighting
that
christian,
the
the
you
know,
the
the
vision
here,
I
think,
would
would
be
to
take
prow
out
entirely
of
the
okd
stream
pipeline
process
and
enable
folks
to
be
able
to
run
their
own,
their
own
ci,
their
own
builds,
and
there
might
be
a
reason
that
openshift
keeps
proud
to
maybe
gate.
You
know
what
what
sort
of
pipelines
feed
into
openshift.
A
With
the
operate
first
cloud,
we
are
able
to
to
launch
a
pipeline
that
the
community
can
participate
in
help,
support
and
manage
and
then
duplicate
for
things
like
hpc
or
specific
to
openstack,
or
you
know
whatever
the
payload
customization
needs
to
be
for
theirs,
because
we
do
have
experiences
of
people
using
okd
with
fcos
or
fedora
core
os
in
production,
but
they
tweak
it
right
for
whether
it's
openstack
or
bare
metal
or
whatever
it
is,
and
but
they're
doing
it
all
on
their
own
and
they
they
really
have
asked-
and
that's
been
one
of
the
things.
A
So
this
okd
streams
is
really
a
wonderful
thing
in
terms
of
getting
more
community
input
and
feedback
quickly
into
the
project
without
interrupting
the
productize
builds
for
openshift
and
openshift's
customers.
A
But
getting
all
the
benefits
and
learning
of
you
know
how
to
tweak
things
out
and
customize
it
with
new
features
that
can
then
get
up
upstreamed
into
or
downstreamed
into
the
product.
B
This
is
we're
essentially
changing
the
course
of
a
river,
and
this
is
a
picture
of
the
chicago
river
in
illinois,
in
the
united
states
and
in
1900
the
chicago
river
was
reversed
and
it
was
it
is
possible.
B
We
have
done
it
before
and
it's
not
one
person
who
changes
the
course
of
a
river.
It's
it's
it's
a
lot
of
people
bought
in
to
yes,
we
are
going
to
change
the
course
of
a
river
and
it's
a
whole.
It's
a
whole
system
that
needs
to
sort
of
be
bought
in
to
changing
to
changing
its
course
and
it's
not
impossible,
but
it
it
does
require
that
everyone
feels
you
know
bought
in
to
what
we're
what
we're
doing
here.
B
It's
not
something
we
can
do
incrementally
you
either
change
it
or
you
don't
we
can't
partially
have
a
river
change
it.
It
all
changes
at
the
same
time,
there's
no
mvp
in
our
way
through
that.
Really,
although
there
are
there's
steps,
we
can
take
build
the
canal
here
build
the
dam
here.
That
sort
of
thing-
and
these
these
are
the
tools
that
we
would
use
the
the
locks
and
levees
and
channels
that
we
would
use
these
to
support
and
making
this
this
change.
B
In
the
course
of
this
river.
We
would
leverage
things
like
operate
first
and
techton
and,
of
course,
the
centos
stream
community
and
our
relationship
with
fedora
korres
would
all
be
part
of
this,
this
river
being
reversed.
B
So
we
wanna
know
what
folks
think
it
we
we
want
to
know
what
risks
folks
in
the
community
see
help
us
sort
of
identify,
maybe
our
things
that
are
blocking
our
view.
What
what
can't
we
see?
B
You
can
use
your
help,
identifying
what
those
risks
are,
and
what
do
you
think
are
milestones
if
there
are
when
I,
when
I've
shown
this
to
folks
to
a
few
folks,
just
to
get
get
their
feedback?
All
kinds
of
ideas
were
coming
up
for
people.
B
Oh
yeah,
I've
been
wanting
to
explore
rebasing
on
okd
stream,
kubernetes
on
opd
stream
first
and
see
what
that
looks
like,
or
I
wanna
you
know
test
this
on
some
hardware
or
I
want
to
I've,
been
wanting
to
work
on
making
it
easier
to
build
okd
locally.
So
those
would
be
really
good
milestones
that
if
folks
are
interested
in
naming
and
particularly
if
they're
interested
in
working
on
to
get
done,
we
want
it.
We
want
to
hear
what
those
milestones
are.
B
If
you
have
a
milestone
of
oh
yeah,
we've
we've
always
wanted
x
y
z.
You
know
if
we
can
do
that.
No
key
streams
that
helps
us
identify
things
that
we
can
set
up
as
milestones,
so
we
so
we
we
can
unlock
things
that
people
are
excited
to
see
as
we
go
along.
So
we're
we're
asking
for
your
for
your
thoughts
we
and
we.
We
love
particularly
help
identifying
these.
These
things
that
would
set
us
up
for
success.
A
Just
before
you
go
ahead,
there
are
a
couple
of
milestones
that
are
coming
up
already
and
christian,
and
a
number
of
the
folks
inside
of
red
hat
have
been
working
on
an
mvp
release
and
some
tech
tongue
pipelines
as
well.
So
maybe
christian.
If
you
could
talk
a
little
bit
and
we
have
recorded
videos
on
both
of
those
things,
but
just
a
little
quickie
on
the
mvp
release
and
when
that's
coming
and
about
the
tecton
pipelines.
C
Themselves,
absolutely
yes,
so
our
first
really
big
milestone
that
we
are
really
close
to
today
and
that
will
hopefully
hit
by
the
end
of
september
this
year
or
start
of
october,
is
the
the
mvp
minimum
viable
product
release
of
okd
on
centos
stream,
9
core
os.
But
that's
a
mouthful.
It's
sure
the
short
version
is
okay.
On
s-cos
and
s-cos
is
the
the
new
base
operating
system
that
we're
building
for
for
this
so
yeah?
That
is
the
first
milestone
that
we're
very
close
to
to
hitting
today
so
stay
tuned.
C
For
that
and
after
that,
we
yeah,
we
already
have
internal
customers
for
the
pipeline
that
want
to
try
out
things
like
replacing
machine
api
with
cluster
api,
and
there
is,
I
think,
endless
opportunities
here
for
also
books
from
the
community.
If
they
have
a
new
component
or
a
new
cni
or
csi
provider,
they
want
to
put
that
into
okd.
They
can
essentially
build
their
own
stream.
With
with
that
feature
and
test
it
out,
the
first
big
one
will
be
the
mvp
coming
very
soon.
A
Okay
and
the
the
even
what
I
think
is
even
more
interesting
news,
not
that
that
is
an
interesting
news,
but
we
are
then
going
to
take
those
first
pipelines
and
put
them
into
tecton
hub,
so
we're
won't.
A
We
really
want
to
engage
with
the
tecton
community
because
we
are,
we
are
learning
and
I'm
sure,
there's
other
tecton
exports
experts
out
there
who
can
help
us
make
more
efficient
pipelines
and
show
us
our
mistakes
and
maybe
give
it
a
run
on
their
hardware
and
their
cloud
platforms
too,
and
that's
really
one
of
the
things
that
we're
hoping
out
of
this
is
not
just
your
feedback,
but
for
you
to
take
these
things
and
test
them
and
tell
us
and
give
some
pull
requests
against
it,
and
that's
the
link
that
you'll
see
at
the
bottom
of
that
next
slide.
A
C
A
Yep,
we
walked
through
that
in
another
video
just
the
other
day,
with
luigi
zuccarelli
and
christian,
so
we'll
make
that
video
available
too.
So
there's
there's
lots
of
ways
you
can
test
and
give
us
feedback
and
we're
really
looking
forward
to
hearing
from
you
all
michelle.
A
B
So
I
think
everyone
that
I've
shown
this
to
is
excited
and,
at
this
point,
we're
building
a
road
map.
So
what's
useful
in
building
a
road
map
are
for
folks
to
ask
the
hard
questions.
Have
you
thought
about
how
you're
going
to
acts
is
useful?
Well,
the
thing
that
I
would
worry
about
the
most
is
why
those
are
really
useful
things
for
folks
to
call
out
and
also
oh,
I've
always
wanted
to
do
abc.
B
Give
us
a
picture
of
what's
in
your
head.
This
helps
us,
you
know
going
to
the
moon
is
a
there's
a
long
way
from
from
here
to
the
moon,
and
you
know
we're
going
to
point
our
rocket
ship
in
that
direction,
but
it's
nice
to
have
some
stops
along
the
way.
So,
if
there's
like
a
space
station,
we
can
visit
or
satellites
we
can
launch
or
if
there's
projects
we
can
do
on
the
way
to
really
unblocking
folks
from
innovating
and
playing
with
these
streams.
B
We
we
are
really
open
to.
You,
know
diversity
of
viewpoints
here
if
there
are
other
ideas
or
things
that
we
could
be,
you
know
technologies,
we
could
be
leveraging
or
ways.
We
could
be
thinking
about
this.
It's
good
to
get
to
have
those
ideas
in
now.
B
A
All
right,
well
christian
and
michelle,
thank
you
so
much
for
your
enthusiasm
and
all
your
support
of
the
community
efforts
and
for
helping
out
drive
some
faster
innovation
into
both
ocp
okd
rail
and
a
bazillion
other
open
source
projects
that
are
part
of
the
openshift
ecosystem
and
the
cloud
native
ecosystem.
I
can
just
this
is
one
of
those
projects
that
really
is
a
game
changer.
I
think
for
lots
of
communities
internally
and
externally.
A
I
know
there's
folks
out
there
in
the
okd
working
group
and
in
the
periphery
that
haven't
spoken
up
and
we'd
love
to
hear
from
you,
especially
if
you're
currently
doing
bespoke
okd
builds.
We
want
to
know
if
this
will
help
you
and
you
know,
and
if
not
why
and
maybe
come
back
to
us
with.
You
know
how
we
can
make
this
easier
for
you,
so
that
you're
not
doing
bespoke
stuff
on
your
own
out
there
and
can
take
advantage
of
some
of
these
tecton
pipelines
we're
building.