►
From YouTube: Kubernetes SIG Cloud Provider 2018-09-19
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).
B
D
Han
is
helping
me
out.
He
will
hopefully
be
helping
with
some
of
the
extracting
Google
effort.
Oh,
fantastic.
B
So
actually
let
me
drop
the
link
to
that
into
into
the
thing
also.
So
if
you
have
any
notes
or
comments
on
that
I'll
work
on
finishing
these
slides
up
tonight.
So
if
there's
anything,
you
would
like
me
to
talk
about
leave
a
comment
in
there
and
hopefully
that's
the
that's.
The
share,
link
and
and
it'll
be
ready
by
the
community
meeting
before
morning.
Yeah.
A
E
B
B
F
I'm
basic
just
proxying
something
from
a
couple
of
hours
ago
at
the
there's
sort
of
this
parallel
offer
to
build
up
the
cluster
API
with
support
for
every
cloud.
The
technique
lives
under
six
cluster
lifecycle
and
they
there
is
a
team
from
kübra
Maddock
that
builds
up
a
an
implementation
of
that
for
digitalocean.
The
question
came
up
in
that
meeting
as
to
whether
or
not
they
were
gonna.
You
know
contribute
that
into
the
communities
project
in
some
way
or
move
for
the
ownership
into
that
into
the
q-
project.
F
A
A
E
F
A
A
A
So,
like
I'm
thinking,
we
should
just
spend
like
five
minutes
every
sick
meeting
actually
going
through
the
tracking
sheet
and
up
keeping
out
to
date.
Are
there
any
objections
to
that
or
do
we
all
think
we
should
do
it
offline
I'm
just
worried
that
if
we
do
it
offline
and
wasn't
actually
updated,
so
we
should
just
spend
the
five
minutes
now.
Do
it.
G
A
A
Okay,
so
I,
just
okay,
I,
don't
think
anyone
from
Baidu's
here
either
so
I'll
give
an
update.
So
we
are
out
of
tree
I
mean
we
have
one
more
checklist
waiting
for
getting
the
meeting
all
the
requirements
for
the
new
provider
repos
that
should
be
done
in
the
next
week
or
two
and
we
are
reporting
conformance
test
results
to
test
grid.
For
me,
111
so
far,
that's
been
pretty
easy
to
set
up
so
far
thanks
to
Ben
and
we
do
have
Docs
so
yay
tcp
seems
like
someone
already
updated.
A
A
A
B
I
Fan
you
can
give
an
update
on
each
spear
sure
well
we're
still
a
entry
and
how
to
tree
right
now,
so
we're
still
maintaining
both,
mostly
because
of
some
kind
of
split
personality.
Right
now,
even
though
we
have
a
team
that
is
maintaining
mostly
the
entry
and
a
team
that
is
maintaining
mostly
the
auditory,
we're
trying
to
make
it
better
on
that
front,
conformance
and
test
grade
is
actually
were
very
close
to
started
reporting.
We
actually
had
a
few
hiccups
and
that
in
the
past
two
days
we
resolve
both
of
them.
I
So
we
expect
to
have
something
coming
in
probably
tomorrow
or
Friday.
Latest
Doc's
were
a
little
behind
that
we
do
have
a
Doc's
folder
in
the
repo.
There
is
one
document
we're
marked
down
that
can
get.
You
started
from
like
start
to
finish,
but
it's
like
not
very
user
friendly,
so
we're
trying
to
make
it
better.
We
have
an
issue
open
and
we
have
somebody
assigned
to
it
so
we're
working
on
it.
Sweetie.
A
Yeah,
so
mainly
the
reason
I
wanted
to
do.
The
kind
of
sheep
was
because
there's
been
some
well
I.
Think
for
a
good
reason.
The
community
has
been
pretty
vocal
about
the
lack
of
dots
and
it's
been
about
like
three.
It's
been
about
like
two
months,
I
think
since
we
opened
the
cat
for
how
we
want
to
document
things,
and
so
I
think
this
is
just
a
good
way
for
us
to
keep
pushing
the
agenda
forward
and
making
sure
that
we're
actually
working
on
Doc's
on
a
bi-weekly
basis.
B
B
B
A
If
we
find
that
this
needs
to
be
a
continuing
like
meeting
we'll,
probably
have
to
like
actually
form
a
working
group,
so
that
people
know
like
we're
to
find
meeting
recordings
and
then
agenda
and
whatnot.
So
this
is
kind
of
just
like
an
FYI.
If
you
want
and
invite
to
that
meeting
right
now,
ping
me
on
slack
and
I
can
get
you
invited
otherwise
expect
some
formal
working
for
the
cloud
provider
extraction
in
the
next
week
or
two
Walter.
Did
you
want
to
add
anything
to
that?
Yeah.
D
So
I
think
you
know
pretty
much.
Google
Microsoft
and
Amazon
have
very
deep
dependency
trees
and
are
going
to
take
a
lot
of
effort.
My
general
feeling
is
that
the
rest
of
our
entry
providers
can
probably
do
exactly
what
DIMMs
was
recommending
and
just
sort
of
build
outside
without
worrying
too
much
about
their
dependency
trail.
D
H
A
B
What
do
you
call
it?
The
cig,
the
sinc
cluster
lifecycle
code,
also
to
not
live
within
the
community's
repositories
anymore.
I
haven't
had
a
chance
to
respond
to
this,
but
you
know
it's
some
of
the.
As
many
of
you
are
aware,
some
of
the
history
of
the
of
the
project
is
of
this.
Cig
is,
to
you
know,
provide
advocacy
and
to
and
guidance
over
there
over
there
upon
the
creation
of
these
repositories
and
in
the
management
of
them.
B
I
can
also
understand
that
it's
work
like
in
every
time
a
new
provider
comes
along
and
you
want
to
add
one
two
three
repositories
and
that
that
can
create
work
and
confusion
and
an
explosion
that
the
community,
maybe
doesn't
want
so
I,
haven't,
haven't
had
an
opportunity
to
respond
to
this
yet
and
I
encourage
you.
If
you
have
an
opinion
on
this
to
chime
in
on
that
mailing
list
thread
and
I,
don't
know
how
Andrew
or
Django
feel
about
this,
but
I.
B
I
would
if,
if
we
were
moved
out,
I
would
I
would
hope
that
we
would
be
able
to
create
another
official
location
for
this,
so
that
we
can
maintain
at
level
of
community
both
in
terms
of
maintaining
the
the
quality
of
the
of
the
code
that's
coming
in,
but
also
as
a
way
to
provide
users,
a
kind
of
a
single
spot
to
discover
and
utilize
all
the
providers
that
are
available
so
I.
Don't
have
anyone
had
any
thoughts
about
that?
H
So
I've
been
telling
bad
line
both
before
this.
It
was
created
and
now,
after
with
Tim
I,
think
it's
actually
in
a
reasonable
spot
right
now,
so
the
action
was
that
we're
gonna
go
ahead
and
create
repositories
and
org,
as
until
the
extraction
is
complete
and
may
be
revisited
after
the
words,
my
hope
is
that
we
just
don't
because
I'm
not
I'm
pleased
personally
from
my
experience
and
other
projects
not
sold
on
the
multi
benefits
of
having
multiple
organizations
get
up
organizations
to
be
specific
here.
H
But
after
you
know,
we
have
some
some
mileage.
This
is
a
concrete
example
of
taking
a
different
route
than
the
communities
sakes
route.
So
we'll
have
some
actual
data
to
compare
so
I.
Think
no
one
is
going
to
do
anything
for
some
period
of
time,
so
I
think
we're
gonna
be
fine,
for
you
know
six
months,
two
years
so.
D
H
Into
the
discussion
it
it
touches
into
the
what
is
communities
too
much
for
us
to
have
I
think
to
make
any
reasonable
progress.
I,
don't
I
have
certainly
not
met
any
users
who
would
consider
the
functionality
provided
by
the
club
bus
to
be
optional
when
actually
running,
you
know
workloads.
So
the
idea
of
moving
them
out,
because
they're,
not
core
I,
just
fundamentally
agree
with
I've,
not
met
a
single
user
who
feels
that
way
and
so
I'm.
B
I
B
J
Hi
so
I'm
the
guy
who's
actually
in
charge
of
implementing
that
organization.
The
reposing
kubernetes
sakes
get
the
exact
same
level
of
automation
as
the
repos
in
kubernetes
right
you.
So
you
get
all
the
benefits
of
the
same
things
working
the
same
way.
Additionally,
if
your
repos
in
community
six,
it
gets
tagged
with
a
topic
for
your
six.
So,
theoretically
everything
that's
tagged
with
Kate's,
a
cloud
provider.
J
My
first
concern
is
to
make
sure
we
don't
live
in
a
world
where
there
are
two
tiers
or
there
are
blessed
cloud
providers
and
there
are
not
blessed
cloud
providers
and
right
now,
the
state
of
today,
where
we
have
some
things
in
kubernetes
and
some
things
and
communities
sakes
could
in
the
minds
of
some
make
that
distinction,
and
we
want
to
avoid
that.
And
so
that's
why
Tim
and
myself
were
of
the
opinion.
J
Nothing
should
be
in
kubernetes
so
that
everything
is
on
an
equal
footing
and
then
selfishly
speaking
as
somebody
who
has
to
manage
github
organizations
I
want
fewer
of
them,
not
more
of
them,
but
maybe
kubernetes
CSI
as
an
organization
to
host
everything
related
to
CSI
could
be
used
as
an
example
of
prior
art
to
have
kubernetes
cloud
providers.
Organization
I
do
have
concerns,
though,
whether
then
that
means
like
ultimately,
every
snake
is
going
to
want
to
know
and
github
organization.
I'm
squatting
on
some
27
to
31
different
github
works
for
that
purpose.
J
If
we
have
to
go
down
that
road,
but
I
don't
want
us
to,
but
like
you
know,
if
we
have
all
the
cloud
providers
considered
part
of
core
kubernetes,
why
wouldn't
you
have
all
the
network
plugins
and
all
the
storage
plugins
and
all
the
container
runtimes
as
part
of
core
kubernetes
as
well?
There
does
need
to
be
a
line
drawn
somewhere.
I
agree:
this
ain't,
the
group
to
hash
it
out,
I
think
that's
a
an
architecture
discussion,
but
just
to
alleviate
the
concerns
of
like
automation,
workflow,
there's,
no
problem
swapping
these
things
back
and
forth.
H
We
there
we
want
that
I,
don't
think
we
want
in
a
unconcious,
multiple
class
of
cloud
provider.
I
do
think
there
is
explicit
value
in
having
two
classes
of
cloud
provider,
one
being
the
cloud
providers
that
are
interacting
and
supporting
the
project
from
a
project
health
standpoint.
So
as
part
of
this
SIG
have,
except
they,
they
adhere
to
technical
standards
for
documentation
for
foreign
to
test
grade,
for
conformance.
That
I
think
is
the
bar
that
this
body
here
is
is
is
fully
capable.
J
Deferred
coming
on,
but
that's
so
that's
what
we
have
in
place
for
queue.
Brunetti
sakes
like
you,
can't
just
do
whatever
you
want.
You
do
actually
have
to
follow
the
standard
processes
and
automation
that
we
have
throughout
the
rest
of
the
project.
The
distinction
to
get
into
kubernetes
is
whether
or
not
sake
architecture
decides.
You
are
considered
core,
but
like
that
level
of
maintenance
and
upkeep
and
expectation
and
good
community
citizenship
is
why
cuber
net
e-cigs
exist,
because
that's
intended
to
be
the
cig,
sponsored
say,
get
owned
projects.
I
mean
I.
H
Competition
with
him
about
this,
specifically,
I
don't
agree
that
this
is
actually
sig
architectures
purview
until
it
relates
to
how
code
is
imported.
There
are
positive
structure,
I
think,
belongs
either
to
the
steering
committee
or
a
contributor
experience.
But
this
what
you're
saying
here-
and
this
is
not
the
form
taps
and
I-
just
don't
buy
it.
That's
their
responsibility.
I
think.