►
From YouTube: CNCF TOC Meeting - 2019-08-06
Description
Join us for Kubernetes Forums Seoul, Sydney, Bengaluru and Delhi - learn more at kubecon.io
Don't miss KubeCon + CloudNativeCon 2020 events in Amsterdam March 30 - April 2, Shanghai July 28-30 and Boston November 17-20! Learn more at kubecon.io. The conference features presentations from developers and end users of Kubernetes, Prometheus, Envoy, and all of the other CNCF-hosted projects
C
C
So
I
guess
we
have
a
couple
of
new
sandbox
projects.
Okay,
we've
got
all
sorts
of
updates
on
like
community
events
and
meetings
coming
up.
I
think
there
might
be
a
bit
of
discussion
around
app
delivery
and
some
changes
of
state
for
some
projects
say
without
further
ado.
Let's
move
on
to
giving
a
giant
welcome
to
Clarkson's,
Tampa
and
sign
offs.
Congratulations
to
those
two
projects,
any
comments
or
questions
about
those.
C
A
A
A
We're
also
launching
something
called
kubernetes
communicated
community
days,
which
will
be
similar
to
kind
of
DevOps
days
in
spirits,
we'll
be
announcing
that,
hopefully,
within
a
month,
but
look
forward
to
that
news,
but
hopefully
this
will
enable
kind
of
more.
You
know
events
for
people
in
areas
that
haven't
necessarily
how
to
cube
con
or
have
been
able
to
attend
kind
of
a
kubernetes
or
CN
CF
event.
E
Question
Chris
and
I
can
ask
through
other
channels,
hopefully
folks,
don't
mind
where's
the
content
for
these
things
coming
from
is
this
gonna
be
seen,
see,
have
people
are
locally
sourced
or
you
know,
see
fps
and
try
and
get
people
to.
A
C
Speak
a
little
bit
more
to
this
as
well,
because
I'm
program
co-chair
for
this
first,
let's
not
let's
in
prepare.
So
yes,
as
Chris
says,
there's
going
to
be
a
local
committee
for
each
event
we're
trying
to
get
about
half
the
speakers
to
be
kind
of
local
for
some
definition
of
local
and
the
other
half
to
be
kind
of
international
speakers
and
try
to
get.
You
know
some
of
the
people
that
you
know.
Everyone
really
wants
to
see
them
speaking
and
to
be
able
to
get
them
to
travel.
I.
C
A
Their
worries,
you
know:
we've
have
a
lot
of
co-located
events,
part
of
Cuba
con
proper.
So
you
know
it's
a
contributor
summit
held
summit
prom
or
envoy
con.
Some
of
these
actually
still
have
open
CFPs.
So
people
you
know,
have
an
inkling
to
another
kind
of
up
event
that
get
a
talk.
Partic,
yukon
service
manager,
con
and
observability
practitioner
summit
had
their
CFP
open
until
the
end
of
this
month
on
august
30th
so
feel
free
to
submit
talks.
A
A
C
Okay,
so
we
see
organized
the
way
that
the
TSE
meetings
are
happening
every
month.
The
main
reason
for
this
is
to
introduce
a
private
TOC
meeting.
What
we
actually
are
finding
is
that,
when
the
committee
get
together,
we
find
it
much
easier
to
work
through
who's
going
to
do
what
on
which
you
know,
assessing
different
projects,
kind
of
speaking
our
minds
about
what
we
think
yeah
about
different
projects
and
so
on,
and
we
found
that
those
meetings
have
actually
enabled
us
to
kind
of
make
progress
on
projects
project.
C
More
effectively
than
we've
been
able
to
do
them
in
the
public
meeting,
so
we
did
try
to
record
them,
but
we
discovered
that
and
both
occasions
that
we
did
do
so
far.
We
had
some
kind
of
sensitive
discussion
that
meant
people
weren't,
comfortable
publishing,
the
recording,
so
I
think
we
will
just
stick
with
note-taking.
Everything
is,
you
know,
documented
as
we
come
out
these
part
of
TRC
meetings.
I
know
that
you
know
everybody.
C
Everybody
wants
the
spirit
of
everything
to
be
done
in
that
open,
but
I
think
the
reality
is
when
people
kind
of
want
to
you
know
raise
their
hand.
It's
quite
difficult
to
do
in
a
very
large
public
call.
So
I
think
this
way
we
can
actually
be
more
effective
as
a
as
a
team
to
drive
things
forward,
but
if
anybody
does
have
any
kind
of
hesitations
or
concerns
about
that,
you
know
do
do
say
so
so
I'm
just
going
to
leave
that
for
a
moment
in
case
anyone
wants
to
type
in
a
question.
C
Okay,
Oh
interesting
point
from
Alex
saying
that
might
be
a
good
format
for
six
I.
Think
if
six
wanted
to
adopt
the
same
principle
in
order
to
kind
of
you
know
get
through
the
work
they
need
to
do.
You
know
so
long
as
they
are
transparently
documenting
all
the
decisions
they
make.
I.
Think
that's!
Okay!
Yes,
do
you
see
notes?
Yes,
absolutely
I
know
the
last
one
we
did
circulate
round
I
know
because
I
did
it
I
think
it
was
in
the
TOC
mailing
list
and
I
think
I've.
C
B
Under
review,
I
think
we've
reached
the
stage
where
the
people
who
are
contributing
to
the
setup
of
the
sink
have
pronounced
themselves
to
be
happy
with
it.
As
far
as
that
goes,
and
now
is
the
time
to
share
it,
with
the
tea
see
and
with
the
TOC
community,
so
I
believe
a
Louis
from
dynaTrace
posted
it
on
the
TOC
list
and
some
of
the
slacks
in
the
last
seven
days.
B
If
anyone
hasn't
seen
that
just
paying
a
new
week
and
there's
the
that
the
document
Chris
anyway,
you
can
get,
we
can
get
you
a
copy
of
the
doc
right
now
where
it
sits,
requests
a
comment
stage.
So
if
we
get
the
right
set
of
comments,
then
Michelle
and
I
will
work
with
the
the
editors
of
the
document
to
to
get
it
into
a
sort
of
cleaned
up
state,
and
then
perhaps
it
can
be
moved
to
a
vote.
Concurrently.
B
H
B
C
B
C
B
C
A
I
Something
on
the
C
gap:
Charter
some
people
are
asking
about
the
relationship
between
this
sick,
app,
the
the
cig,
app
sake
and
the
the
service
working
group
and
I
cannot
remember
what
the
TOC
decided
was.
The
service
working
group
gonna
be
pushed
under
the
sick,
up
delivery,
sig
or
it's
gonna
be
separate
a
camera.
What
the
decision
was
sorry
who
was
asking
that
question
Sarah
asked
it
on
there
and
there
was
somebody
else,
a
camera
gloss
on
the
dock.
There
a
couple
different
questions
about
it,
so
I,
probably.
B
Need
to
to
chat
about
this
more
with
Michele,
we
haven't
really
had
a
chance
to
sync
up
in
detail,
I'm
happy
to
declare
my
own
preference,
which
is
that,
at
least
for
for
the
next
couple
of
months,
we're
happy
to
put
the
serverless
WG
under
the
app
delivery,
sync
just
for
now
and
then
solve
the
problem
later.
I
think
that
you
know
it
needs
to
carry
on
its
work.
It
may
need
to
have
assets
so
in
home,
but
EOC
mate.
It
may
need
to
be
involved
in
that
decision.
B
I
B
I
mean
I,
don't
think,
there's
any
particularly
strong
desire
to
to
suspend
it
or
shut
it
down.
I
think
that
would
be
unfair
and
inappropriate.
Mm-Hm,
so
I
think
that
it
therefore
it
needs
to
continue.
Therefore
it
needs
a
home.
Therefore,
we
can
carry
on
in
this
way
for
the
time
being,
yep.
Okay.
Does
anybody
object
to
that.
C
C
C
D
C
F
C
C
F
C
C
C
C
J
A
I,
don't
think
so
like
they
would
just
have
to
agree
to
it
and
there
would
be
enough
TOC
sponsors.
So
if
they
agree
with
that,
we're
all
good,
if
they
don't
agree
with
it,
then
that's
their
decision
and
I
think
we're
kind
of
kind
of
in
the
middle
here.
With
this
situation,
where
I
think
in
toto
thinks
they
have
a
lot
more
usage
and
definitely
qualify
under
the
incubation
criteria,
where
not
everyone
is
sold
on
that.
On
that
opinion,.
C
J
C
J
C
Yeah,
so
it
would
be
nice
to
get
you
know.
The
thoughts
from
people
like
Justin
who's,
a
maintainer
of
Mowgli
I,
knew
thinking
it
widely-used
I.
Think
there
are,
you
know
significant
number
of
cloud
native
doctors
by
using
basically
either
as
part.
You
know,
doctor
trusted
registry
or
truth
I
believe
there
are
significant
numbers
of
uses
of
it,
but
I
think
it's
one
of
these
difficult
to
get
documentation
on
how
many
there
are
the.
E
Would
also
say
that
you
know
and
correct
me
if
I'm
wrong
with
with
rocket
you
know
it
was.
It
was
clear
even
for
the
folks
who
were
on
the
hook
as
maintainer
for
the
project
that
they
were
moving
on
and
the
project
itself
was
sort
of
you
know
opting
into
this
archiving
process.
I
think
we're
sort
of
on
new
ground
here.
When
we
see
somebody
from
the
outside,
essentially
coming
and
saying
hey,
we
should
archive
this
project.
I
think
that's!
H
G
So
hi
I'm
Brandon
for
IBM
of
Saudi,
we
use
notary
or
spell,
but
we've
also
ran
into
issues
with
being
trying
to
upstream
certain
patches.
So
what
we
do
is
actually
we
have
a
different
set
of
factors
that
we
we
apply
from
production.
G
H
C
C
This
is
the
next
community
presentation.
It's
the
tough
graduation
review,
because
I
think
this
is
another
case
where
we
have
struggled
in
the
past
around
what
the
definition
of
graduation
means
respect
project
so
TLC
members,
let's
all
kind
of
get
our
thoughts
together
about
that
ready
for
this
graduation
review.
C
C
C
A
C
H
I'm
interested
in
that
operator
for
frame
work
personally,
because
I've
been
working
with
them
more
recently
and
also
because
I've
been
hearing
more
about
more
questions
around
how
to
operator
like
weird
whopper
eaters
go.
Should
they
be
in
the
CAF
as
part
of
the
home
community
I
also
get
questions
around
how
operators
can
be
packaged
and
how
they
relate
so
that
homework.
H
E
H
E
One
of
the
things
that
it
might
be
useful
to
actually
do
with
this
backlog
is
to
if
there
is
an
appropriate
cig
to
actually
you
know,
have
you
know
us
mark
which
cig
we
think
actually
should
should
be
helping
to
facilitate
guide.
Prep
sort
of
you
know
standard
set
of
information
around
these
things.
C
H
C
C
H
B
I
think
it's
implied
in
the
discussions
which
are
in
the
Charter
around
the
use
of
the
very
large
document
that
Gareth
rush,
Grove
and
Brian
grant
created,
which
has
about
95
different
projects
which
all
aspire
to
solve
some
a
tiny
little
problem
in
the
app
deaf
space.
Now,
obviously,
most
of
those
are
not
used
by
many
people,
and
probably
you
know
not
going
to
be
just
by
many
people.
However,
I
do
think
one
of
the
jobs
of
the
app
delivery
sync
should
be
so.
C
That
would
actually
be
really
helpful
and
I
I
slightly
fear
that
we
might
end
up.
On
the
one
hand
we've
talked
about,
you
know
the
Cambrian
explosion
of
new
projects
coming
into
the
sandbox,
but
is
there
a
risk
that
we,
you
know
the
first
end
projects
that
come
in
that
make
their
proposal
we
kind
of
go
yeah?
Why
not?
And
then,
after
a
while,
we
get
project,
fatigue
and
kind
of
go.
We've
got
enough
of
those,
but
maybe
we
haven't
hit
the
right
ones.
Yeah.
B
B
L
You
know
at
CD,
Prometheus
and
vault,
you
know
is
its
first
three
operators,
so
I
think
it's
important
to
know
that
it's
really
rooted
in
fundamental
operation
of
deploying
things
that
are
complex
within
the
kubernetes
landscape
and
not
something
that
Red
Hat
went
and
cooked
up
by
themselves.
Just
one.
L
E
Maybe
you
know
yeah
I
think
each
operate
like
the
question
is:
is
that
each
each
operator
is
a
project
in
and
of
itself
it
has
its
own
life
cycle,
it
has
its
own
releases,
it
has
its
own
community,
it
has
its
own
database,
and
so
we
should
view
those
as
individual
in
you
know
independent
projects
right.
So
if
we
end
up
with,
like
a
you
know,
a
fubar
operator
for
a
project-
and
you
know,
project
fubar
doesn't
want
to
own
that
operator,
for
whatever
reason
then
I
think
it's
reasonable
for
us
to.
E
You
know,
assuming
that
it
meets
a
certain
bar
for
us
to
have
that
be
at
CN
CF
project,
but
that's
different
than
the
operator
framework
right,
which
is
really
a
framework
in
general
and
I.
Think
we
need
to
be
view
that
within
the
set
of
comparable
tools
and
then
also
to
Alexis's,
point
I
think
having
the
app
delivery
sig
help
map
out
these
different
options
is
great,
because
I
think
that's
a
good
first
step
to
actually
sort
of
like
how
do
we
actually
track
these
things?
Maybe
we
add
these
to
the
landscape.
I,
don't
know.
E
Maybe
we
create
a
new
landscape
around
these
types
of
tools,
but
there
are
also
different
audiences
here.
I
think
one
of
the
things
that
I'd
love
to
see
get
started
in
the
kubernetes
project
in
general
is
the
idea
of
an
SDK
right.
We
don't
actually
have
sig
SDK
or
something
that's
developer,
focused
in
terms
of
actually
building
up
an
experience
for
developers
and
I.
Think
the
operator
framework
and
cube
builders
start
to
get
at
that.
That's
a
different
issue
than
the
tools
that
are
completely
config
related,
so
I
think
teasing.
L
E
C
Think
I
have
similar
concerns
that
you
know
there
will
be
some
operators
where
it
feels
like
a
good
fit
for
CNC
f4.
It
feels
like
a
good
fit
from
an
exploring
where
the
operators
even
work
in
that
space
and
I
think
Brooke
might
be
a
good
example
of
that
you
know,
is
an
operator
appropriate
for
storage.
Presumably
that's
a
kind
of
experiment
that
rookies
you
know
has
undertaken,
but
I
wouldn't
like
to
see
us
having
like
500
operators
for
every
single
possible
vendor
yeah.
B
I
think
I
think
there's
two
different
ways
of
looking
at
this.
So
one
of
them
is.
There
are
functional
extensions
to
Cuba
Nettie's
clusters,
which
an
operator
happens
to
be
the
correct
idiom
to
use
for
making
them
work
with
Cuba
Nettie's,
that's
and
that
was
actually
pioneered,
as
Erin
said
by
Brandon
and
the
folks
at
core
OS
and
then
a
few
other
people
too,
and
then
there's
the
other
phenomenon
Mojo's
akima,
which
is
that
there
is
a
systematic
and
growing
proliferation
of
operator
for
X
for
any
X
that
you
can
think
of
which
could
cause.
B
You
know
tremendous
confusion
as
well
and
I
think
it
could
should
be
ideally
handled
in
a
different
way,
perhaps
in
the
Linux,
Foundation
or
somewhere
else.
Finally,
I
think
there's
an
element,
you
know,
I,
don't
personally
feel
comfortable,
saying
hey
the
Kafka
operator
should
be
done
by
confluent.
I
think
that
that
is
gonna
cause,
other
kinds
of
problems
which
I
probably
don't
need
to
explain
anyway.
Michelle
was
raising
her
hand
I.
H
Yeah
I
just
discussion,
hopefully
in
saqqaq
delivery
as
well,
because
there's
some
prior
art
here
so
home
also
would
fall
under
Sagat
delivery
and
we
have
a
lot
of
experience
from
helm,
hub
and
maintaining
charts
and
so
charts
each
artist.
You
know
application
specific
software
that
runs
in
kubernetes
and
at
first
we
thought:
hey.
Let's
put
these
charts
in
a
community
repo,
so
they're
easy,
easily
accessible,
easy
for
people
to
just
grab
something
off
the
shelf
and
run
that.
H
But
there's
a
lot
of
issues
around
that
one
being
just
maintaining
these
charts
and
having
them
in
one
central
location
is
really
difficult
and-
and
it's
there's
a
huge
scale
issue
there
and
with
operators
it
will
with
helm.
We
decided
to
go
with
home
hub
and
really
encourage
people
to
write
their
own
charts
and
maintain
their
own
charts.
So
as
a
vendor,
you
would
maintain
your
own
chart,
which
is
a
specific
way
for
you
to
bring
your
okay
and
I.
H
Think
we're
I
think
there
are
some
similarities
here
in
terms
of
how
we're
operators
should
live,
and
this
whole
conversation
around
the
operator
landscape.
So
I
think
that
and
I
would
encourage
you
know
the
vendor
that
is
most
associated
with
that
operator
to
maintain
and
host
and
host
that
operator
as
well.
But
again,
that's
my
personal
opinion
and
I
would
want
to
have
that
larger
discussion
in
say:
got
delivery.
That's
an
example.
The
type
of
discussion
I'd
like
to
see
you.
C
B
So
I
think
that
sounds
very
useful.
You
need
yep
the
other
one.
That's
come
up
a
fair
bit
and
I'm.
So
personally,
supportive
of
is
something
that
starts
to
talk
about.
What
are
the
quote?
Unquote.
Delivery
patterns
know
how
this
delivery
happened.
There's
lots
of
discussion
of
blue
green,
a
B
canary
staging
promotion
pipelines.
Yet
again
it
could
be
some
useful
work
done
around
that.
What
is
the
difference
between
deployments
and
delivery,
for
example,
etc,
etc?.
H
B
B
53
participants
on
the
tall
and
not
one
person,
is
volunteering,
but
what
should
be
a
really
exciting
opportunity
is
to
show
everybody
that
you're
really
exciting
to
work
with,
so
they
scream
a
little
bit.
No,
no,
no,
no
just
put
up
your
hand
and
say
I'll:
do
it
anyone
just
don't
have
to
be
today.
I
mean
you
could
come
back,
you
know
in
a
week
and
say
I've
thought
about
it
and
I
decided.
This
is
a
really
good
thing
to
offer.
B
C
C
E
Mean
no,
this
is
just
a.
This
is
a
recurring
topic
for
me
is
I,
think
you
know
one
of
the
things
as
we
think
about
a
platform
when
a
developer
approaches
a
platform.
What
they
want
to
do
is
is,
you
know,
have
the
tool
set
that
says:
here's
how
to
approach
this
thing
as
a
developer
and
I
think
we.
We
have
a
bunch
of
that
stuff.
That's
sort
of
organically
sprouted
around
the
cloud
native
ecosystem,
but
it's
not
something:
that's
cura,
curated
and
packaged
up
for
a
developer
facing
audience,
and
so
inside
kubernetes.
E
You
know,
I'd
love
to
see.
You
know.
Client
go
be
more
accessible
to
to
developers
because
right
now
it's
it's
hard,
but
I
think
that
also
starts
getting
into
the
operator
framework.
It
also
starts
talking
about
like,
for
you
know,
prod
that
actually
have
multiple
sides,
but
the
operators
side
in
the
developer
side.
How
do
we
actually
represent
the
developer,
face
and
stuff
on
that?
And
do
we
actually
start
building
a
cross
project
sort
of
SDK
across
this
stuff?
It's
a
bigger
discussion.
M
E
We
might
have
a
landscape
saying
well,
here's
those
sort
of
you
know
the
rust
language
bindings
for
ED
CD
right
so,
like
you
know,
there's
a
combinatorial
explosion
around
this
stuff.
How
do
we
actually
start
getting
a
handle
on
this
and
really
providing
the
rights?
The
right
input
there
I,
don't
know
it's
unformed
thoughts,
but
I
do
think
that
there's
a
missing
there's
a
missing
aspect
here,
which
is
the
developer
facing
stuff.
B
Mean
this,
this
area
is
one
which
we've
been
talking
about
for
a
couple
of
years
now
and
during
that
time
it's
been
consistently
agreed
that
it's
an
important
area,
but
it's
also
been
consistently
agreed
that
participants
in
the
community
don't
feel
they
have
enough
spare
bandwidth
to
start
putting
new
effort
into
this
area.
So
we
might
just
have
to
wait.
We
can't
do
everything
at
once
and
we
have
to
focus.