►
From YouTube: CNCF TOC Meeting 2020-09-08
Description
CNCF TOC Meeting 2020-09-08
A
All
right,
we
are
now
recording
liz,
go
ahead,
okay!
So,
let's
start
at
the
top.
B
So
chekhov,
I
have
thoughts
and
I
think
this
might
be
one
of
the
ones
that
justin
also
sent
some
thoughts
on.
So
I
will
just
copy
for
the
record
justin.
B
Oh
sorry,
I've
sent
that
to
amy
privately
and
I
didn't
need
to.
I
meant
to
send
that.
B
And
I
must
say
I
had
similar
thoughts
that
it
currently
looks
very
bridge
crew
branded
which
isn't
a
blocker.
But
I
just
wondered
how
much
they
had
really
thought
about
that
and
whether.
C
C
To
be
fair,
the
timing
on
this
cloud
custodian
had
just
barely
come
in.
I
believe.
B
E
E
Forever,
you
know,
I
don't
like,
I
feel
like
at
the
sandbox
level,
yeah
I'd
I'd
like
a
summary
of
like
the
road
map,
but
I
don't
know
if
we
in
the
form
do
we
describe
like
how
what
we
want
the
road
map
in
the
form
of.
B
B
E
I
think
the
questions
you
asked
are
reasonable
and
we
should
just
go
back
and
ask
them
before
voting.
D
Not
quite
because
it
has
a
very
clear
scope
and
it's
yeah
it's
it
looks
like
it's
something
that
could
be.
It
could
be
useful,
but
yeah
getting
back
to
the
to
the
owners
makes
sense
at
this
point.
D
I
think
it's
a
really
cool
project.
I
think
so
too.
I
don't
know
one
thing
that
I
don't
want
to
be
the
first
one
to
say
it.
I
think
it's
very
cool.
I
just
didn't
understand
why
linker
d
was
brought
up
as
a
similar
project,
but
that's
just
such
a
minor
comment.
D
D
Feature,
I
don't
know
about
or
some
some
item
on,
the
road
map
that
kind
of
like
aligns
it
with
linker
d.
But
overall
I
think
it's
a
great.
B
Project
all
right,
any
other
comments
or
reservations
or
questions.
B
C
B
Feelings,
well
I
mean
he
sent
an
email
with
plus
one
on
the
others,
which
is
all
but
two.
I
don't
know
whether
we
can
do
remote
votes
like
that.
B
G
Concerns
I
have
some
reservation
about
this
project.
I
was
looking.
I
was
looking
at
the
roadmap.
Actually,
there
is
no
roadmap.
Well,
at
least
the
link
is
not
working
for
me.
I
was
trying
to
look
for
that.
I
didn't
find
anything,
and
then
I
was
looking
at
the
contribution.
It
seems
that
there
are
only
two
people
actually
working
on
it
and
there
is
no
like
noticeable
adoption
by
other
companies.
G
B
I
I
don't
know
if
that's
a
reasonable
thing
to
suggest
that
maybe
their
next
step
should
be
to
because
I
guess,
if
there
is
alignment
between
what
they're
doing
in
grpc,
then
that
would
be
great.
And
if
there
isn't
you
know,
and
maybe
it
could
become
a
sub
project
of
grpc,
and
if
there
isn't
alignment,
then
I
suspect
it's
a
problem.
G
I
tend
to
agree
as
well.
It's
very
I
mean
I'm
not
sure
if
the
project
maintainers
talk
between
themselves,
but
I
think
if
there
is
a
very
close
similarities,
or
at
least
it
makes
sense
to
make
the
ban
under
the
same
umbrella,
maybe
they
could
collaborate
but
yeah
at
this
stage.
I
don't
know
like
for
me
at
this
stage.
It
seems
just
like
a
very
standalone
project.
B
B
C
Rather
than
proton,
do
we
want
to
note
that
there's
no
road
map
in
here
or
the
road
map
was
not
available?
C
A
G
I
Yeah,
I
wonder
you
know
this
is
not
specifically
for
this
project,
but
I
wonder
if
there's
any
kind
of
I'm
very
sensitive
to
this,
just
the
general
maturity
and
traction
you
know
is
there
in
in
any
of
your
mind,
is
there
like
a
bar
like
that
to
be
crossed
before
it
could
be
considered,
even
as
a
cncf
sandbox
project,
because
if
someone
could
just
come
in
and
just
check
all
the
other
boxes,
you
know
I,
I
start
a
project
today
I
mean
there's
an
extreme
case
and
then
and
then
you
know,
have
a
road
map
and
have
a
little
bit
of
code,
and
that
is
that
I
mean
that
probably
wouldn't
wouldn't
pass
the
the
the
gut
check
right.
I
It's
it's
a
it's.
It's
got
to
be.
I
mean
this.
This
project
is,
is
probably
amongst
the
the
least
mature
that
I've
ever
seen.
So
I
wonder
if
that
plays
into
you
know,
folks
thinking.
I
B
B
B
B
D
C
Well,
actually,
I
feel
like
we
can
work
with
I'm
looking
at
column,
j
j1,
and
it
sounds
like
that-
would
just
need
to
be
reworded
slightly.
B
I
think
there's
two
different
points
here:
the
alignment
I
people
answer
this
in
quite
a
useful
way
around
like
how
how
their
project
relates
to
other
cloud
native
projects,
but
justin
we're
just
talking
about
how
for
a
few
of
these,
we
we
feel
like.
We
want
to
understand
why
they
want
to
join
the
cncf,
and
perhaps
we
should
considering
ask.
Should
we
ask
a
question:
why
do
you
want
to
donate
this
project
to
the
cncf
to
the
fullest.
J
B
To
speed
we
are,
we've
just
been
talking
about
proto-p,
so
we
decided
that
for
chekhov
and
well
for
chekhov,
we
weren't
confident
about
the
road
map
and
shared
your
concerns
about
the
the
amount
of
branding.
So
I
think
we
recommended
having
a
conversation
with
them.
First
about.
G
B
Like
what
they
want
out
of
it,
and
why
they're
doing
the
donation
and
do
they
really
understand
that
they
will
have
to
give
up
all
the
branding
and
they
also
don't
have
a
great
roadmap.
They
just
have
a
list
of
issues
so
backstage
past
protocol
also
shared
concerns
and
recommended
that
they
should
speak
to
grpc,
but
that
led
shane
to
asking
about.
B
J
E
B
B
I
actually
really
liked
they,
they
kind
of
document,
their
origin
story,
which
I
thought
was
really
really
useful.
You
know
that
this
is,
you
know
we
had
a
problem
at
wayfair
and
we
wrote
this
project
to
solve
it,
so
that
was
pretty
nice.
D
B
B
D
J
I
mean
I
think
that
there
are
there's
a
lot
of
experimentation
in
the
edge
cube
space
and
I
think
encouraging.
That
is
a
good
thing.
I
think
that
no
one
really
at
this
point
knows
what
architectures
are
going
to
be
successful,
and
so
I
think,
but
there's
definitely
interest.
So
I
think
I'm
I'm
quite
positive
that
we
should
be
encouraging
experimentation
in
this
space
because
it's
it's
it's
a
real
space.
G
A
great
agreement
with
that
yeah,
I'm
actually
quite
surprised
and
please
pleasantly
surprised
that
they
are
actively
collaborating
with
cluster
api
and
they
providing
their
own
integration
with
it,
and
it
will
enable
the
creation
of
clusters
in
bare
metal
or
at
least
having
that
interface
around.
So
I
think
that's
actually
because
we
had
a
lot
of
questions
about.
How
can
I
create
my
cluster
but
not
be
in
a
cloud
vendor
but
still
be
bare
metal,
so
I
think
that's
going
to
be
a
very
good
addition
to
the
ecosystem.
B
F
E
So
I
named
it
actually
yeah,
so
we
wanted
to
do
a
throwback
to
like
just
like
kind
of
the
vibe
that
we
had
at
deus
and
then
include
the
new
folks
at
microsoft
and
then
include,
like
you,
know,
community
people
who
we
wanted
to
work
with
on
open
source
stuff.
J
J
Oops
the
scene
of
spec
is
in
the
other
org.
Isn't
it
the
other
letters
foundation.
B
E
J
J
D
K
Yeah,
if
I
recall
correctly,
this
project
presented
to
sig
storage
and
we
invited
them
to
collaborate
with
the
the
container
object,
storage
interface
initiative-
that's
already
going
on
and
they
started
participating
in
that.
I'm
not
sure
whether
it's
worthwhile
to
continue
to
to
accept
this
as
a
separate
project
or
not.
I
I
I
know
it's
it's
kind
of
early
stage
and
things
will
evolve,
but
you
know
that
I
think
the
better
you
make
it
crisp,
the
more
likelihood
that
you
get
more
users
and
more
developers-
and
you
know
more
stars
so
part
of
the
problem-
is
it's
not
clear
what
it
is.
B
B
J
I
think
maybe
we
should
talk
to
sig
storage
about
this
a
bit.
I
would
like
to
have
feedback
from
them
if
they've
spoken
to
them
already.
K
I
think
that
wouldn't
hurt
definitely
worth
talking
to
to
the
the
leads
there.
I
think
my
personal
opinion
is.
I
would
pass.
K
I
think
there's
a
lot
of
overlap
with
the
project
with
what's
already
being
developed
in
the
cap.
That's
being
pointed
out
in
kubernetes,
and
so
what
I'd
like
to
see
is
the
project
kind
of
differentiate
itself?
From
that
and
say,
okay,
you
know:
is
it
complementary
to
that
and
adding
additional
functionality
on
top,
which
it
potentially
could
do
if
it
wanted
to
or
or
is
it
going
to
kind
of,
tackle
the
same
space
in
a
different
way,
which
I
would
caution
against.
I
I
mean
even
if
it
tackles
the
same
thing,
a
different
way
if
it
points
out
the
reason
for
it
like
I
I'm
just
kind
of
I
guess
I
I
I
just
find
it
difficult.
It
looks
like
it's
adding
a
lot
of
metadata
to
you
know
to
the
data
that
seems
like
what
its
purpose
is,
but
but
but
I
I'm
not
sure
I
I
fully
comprehend
the
significance
of
it.
Is
that
like
automatically
generated,
is
that,
like
you
know
it's
it's
it's
it's
hard
it
just
I'm
kind
of
lacking
understanding
a
little
bit.
K
Yeah,
I
think
my
understanding
is
that
they
take
various
data
sources
like
a
bucket,
an
object
bucket
and
make
it
easy
to
surface
it
up
into
a
container
they'll
mount
it
into
that
container
and
just
kind
of
give
you
various
ways
to
be
able
to
keep
track
of
different
data
sets
very
similar
to
the
cozy
proposal.
Right
now,
awesome.
K
Yes,
it's
similar,
so
the
the
proposal
that
is
a
cap
is
called
a
container
object,
storage
interface
and
effectively
they're,
looking
to
standardize
how
object
storage
is
surfaced
into
containers
and
kubernetes
such
that
we
could
kind
of
have
a
portable
interface.
Much
like
we
have
for
file
and
block
today
with
csi.
I
K
They
listed
it
as
a
link
to
the
cap,
because
that
project
is
still
in
design
phase.
B
B
B
D
B
B
J
J
B
And
I
guess,
as
as
a
separate
thing
for
us
to
think
about,
is
you
know
how,
when
I
I'm
just
sort
of
flagging
up
that
at
some
point
we're
going
to
have
to
think
about
which
of
these
different
competing
products
projects
when
they
apply
for
incubation?
Do
we
want
to
consider
them
as
as
competitors
that
that
kind
of
thing,
I
think
it's
not
a,
not
a
decision?
We
need
to.
B
J
D
H
B
I
am
guessing
that
this
has
all
the
requirements
in
place.
B
Does
anybody
want
to
make
any
comments
or.
J
B
L
Wow,
this
is
actually
a
little
bit
different
than
the
kubernetes
itself
right.
So
it's
actually
a
set
of
controllers
that
enhance
kubernetes,
but
the
controllers
are
all
about
application.
Workload
management,
for
example,
sidecar
management-
there
is
a
thing
called
a
clone
set-
is
similar
to
a
hybrid
version
of
stable
side
and
replica
sites.
So
basically,
kubernetes
upstream
only
support
a
key
set
of
workload
controllers
and
it
this
product
complements
that
and
the
us
actually
has
a
lot
of
users
in
china
and
I
think
lyft
is
also
using
urban
crews.
L
Yeah
for
the
at
the
one
step
side,
we
are
working
with
the
upstream
to
actually
move
most
of
it
to
upstream,
but
acceptance
for
the
other
things.
If
upstream
wants
to
do
similar
things,
we
just
move
it
there,
but
right
now,
it's
a
place
for
us
to
experimenting
new
ideas
and
new
workloads.
Besides
core
kubernetes
and
I
think
for
kubernetes.
L
Actually,
we
are
trying
to
limit
the
scope
of
all
these
key
workloads
and
for
some
other,
like
companies
who
wants
more
like
workloads,
they
can
collaborate
on
this
product
and
I
think
this
is
the
goal
we
create
this
product
that
we
want
donated
to.
Since
you
have.
J
L
Would
it
would
this
if
you're
familiar
with
kubernetes,
there's
thing
called
controller
manager,
it's
actually
similar
to
the
control
manager?
The
control
manager
actually
manage
a
set
of
things
right,
yeah
yeah
this.
L
If
it's
related
to
workload
management,
it's
in
the
boundary,
so,
for
example,
if
you
want
to
manage
your
own
like
application
and
your
application
has
like
different
requirements,
upstream
controller
provides-
and
you
can
experiment
your
controllers
here
and
if
it's
like
something
related
to
like
operators
and
other
like
non-controller
related
stuff.
It's
not
the
scope
in
this
project.
L
Well,
I
I
think
we
can
talk
to
the
upstream,
but
my
understanding
is
upstream
is
actually
trying
to
limit
the
scope
of
the
core
controllers
and
we
developed
crd
in
that
kubernetes,
so
that
users
can
actually
extend
kubernetes
outside
the
kubernetes
code
like
part
itself,
and
I
think
that's
the
goal
of
crds
and
that's
why
we
are
experimenting
things
off
at
kubernetes
and
for
things
like
at
once,
stable
side
right.
We
add
some
little
things
here
and
there
we
will
try
to
like
move
those
to
kubernetes.
K
I
think
kubernetes
is
designed
for
exactly
this.
This
type
of
thing,
where
there
are
effectively
kubernetes
sub-projects
that
are
sponsored
by
kubernetes
cigs,
so
I
think
the
advanced
stateful
set
would
be
a
good
example
of
a
project
that
could
find
a
home.
There
is
the
intention
eventually
to
do
that
for
all
of
these
controllers,
or
is
the
intention
to
support
them
independent
of
kubernetes.
L
So
I
think
it
depends
for
at
one
stable
cell,
we
actually
want
want
them
to
eventually
goes
into
kubernetes.
For
the
other,
like
workload
types
we
have
to
like.
Do
the
experiments
and
see
what
is
the
adoption
rate
and
the
weather
is
common
to
other
users,
and
then
we
can
decide
whether
we
should
push
this
into
kubernetes
afternoons
or
we
can
just
host
it
in
another
like
product,
because
I
think,
like
kubernetes
kind
of
the
base
of
this
product,
and
if
it's
a
controller,
the
controller
types
are
coming
enough.
L
B
I'm
looking
on
the
face
of
it,
I
think
there's
some
really
great
experiments
going
on
here
that
are,
you
know,
I'm
not
in
any
way
questioning
the
work.
I
think
the
work
isn't
good.
I
am
wondering
whether
this
is
something
we
should
talk
to
the
kubernetes
steering
committee
about,
because
I'm
wondering
why
you
know
would
the
kubernetes
project
prefer
this
kind
of
experiment
to
be
happening
within
their
project
or
outside
of
their
project,
and
you
know
I.
B
G
B
Having
kind
of
different
kubernetes
resource
experiments
kind
of
competing
with
each
other,
it
doesn't
feel
quite
the.
As
I
say,
although
I
think
it's
good
work,
I
do.
I
don't
think
this
is.
J
B
I
don't
think
the
sandbox
should
be
a
shortcut
that
sort
of
circumvents
the
way
kubernetes
are
running
projects
and
equally,
if
a
shortcut
is
needed,
because
somehow
it's
difficult
to
get
enhancements
working
through
kubernetes,
we
need
to
unblock
that.
I
feel
like
this
is
something
we
should
be.
B
E
Yeah,
I
think
looping
and
sig
apps
would
be
really
good
to
just
to
get
an
idea
of
whether
it
makes
sense
for
this
to
be
a
sub
project
under
under
them,
and
some
subprojects
really
start
from
within
the
sake,
and
this
is
already
pretty
mature.
So
getting
that
also
view
from
steering
on
how
to
handle
the
situation
and
what
their
perspective
is
would
be
great
as
well.
B
B
E
And
then
I'd
love
to
like
just
to
also
add
on
love,
to
see
this
get
somewhere.
Bender
neutral
and
you
know
be
part
of
the
community,
because
I
really
do
appreciate
this
work
and
think
it's
very
useful.
E
B
B
B
B
B
D
I
would
love
that
too,
especially,
given
that
there
was
no,
that
the
project
owners
haven't
completed
the
alignment
with
other
cncf
projects
section.
I
believe.
B
E
B
Yes,
so.
B
E
I'd
like
more
information,
I
I
really
like
that
these
projects
are
coming
in
from
end
users,
but
I'm
seeing
that
maybe
we
need
we
need
to
go
back
and
ask
them
for
more.
So
I
just
want
to
encourage
encourage
these
kinds
of
projects,
but
just
with
a
little
more
information,
yeah.
B
B
E
Even
though
it's
not
a
question
on
the
forum,
can
we
ask
them,
like
you
know
what
they're
looking
for
from
the
cncf
just
so
we
get
a
better
idea
of
that
too.
J
I
mean
it's
got,
I'm
again:
15
forks,
155
stars,
it's
a
it's
a
I
mean
I
like
the.
I
think
the
idea
is
interesting,
but
it's
the
sort
of
thing
that
I
would
like
to
see
usage
to
know
that
it's
that
they're
approaching
it
right
in
that,
so
that
people
find
it
are
finding
it
helpful,
because
I
think.
J
I
I
think
supporting
database
migrations
is
something
that
people
have
a
problem
with
in
this
in
cloud
environments,
but
there
are
also
difficulties
with
it
like
rollbacks
and
a
b
tests.
You
know
sorry
green
blue
deployment,
all
sorts
of
kind
of
issues
with
that.
I
don't
know
if
they've
really,
if
they've
solved
enough
of
the
users
problems
in
this
space,
to
make
it
plausible.
B
J
Yeah,
that
was
that
was
my
view,
that
it's
a
good
idea,
and
maybe
we
should
just
talk
to
them
dependency.
I
mean
again,
it's
mostly
one
person.
E
B
B
D
C
B
B
B
B
C
Closed
meeting,
so
the
tricky
part
about
that
is
that
we
have
a
two-month
cycle
right
now
and
right
now
we
still
have
quorum.