►
From YouTube: Cluster API Provider Implementers Office Hours 20181218
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).
A
A
Don't
think
I'll
be
able
to
join
these
office
hours
today,
but
we're
working
on
that
and
we're
gonna
publish
a
draft
and
you
know
get
get
everybody's
feedback
and
comments
and
but
anyway,
it's
I,
don't
know
if
it's
something
that
you're
interested
in
or
or
you
know,
are
you
interested
in
one
of
the
other
providers
or
yet
you
have
any.
If
you
have
any
questions
or
anything,
anything
like
that,
yeah
feel
free
de.
B
Yeah
actually
I've
been
just
attending
these
meetings.
To
see
how
things
are,
things
are
doing
and
all
that,
but
really
having
tried
to
find
a
place,
we
were
to
contribute
so,
okay,
yeah
I,
guess
since
I.
Have
it
right
here,
I
don't
know.
Is
there
something
that
you
think
it'd
be
a
good
place
for
somebody
to
just
get
started,
something
something
boring
the
unit
I
didn't
want
to
do
and
just
wanna
push
it
aside.
That.
A
A
So
cluster
cuddle
validate
cluster
run
more
automated
on
Cluster
API,
repo
yeah.
One
of
those
might
be
one
of
those
issues
with
those
labels
would
probably
be
a
good
good
bed.
Let
me
let
me
take
a
look
at
them.
Some
of
them
are
a
little
a
little
old
and
then
I
don't
know
if
you've
seen
the
the
docs
the
the
get
book
documentation.
A
B
A
A
A
A
B
Actually,
care
about
the
project
from
the
people
at
the
concert.
Lifecycle
like
the
world
that
you
guys
are
it's
interesting,
a
provider.
Some
will
love
to
test
it
out
with
any
Google
or
AWS
specific
things,
especially
since
I
really
don't
have
access
to
a
sure,
babe
VMware,
anything
of
anything
else.
So
at
least
those
two
I.
If
there's
anything,
I
really
not
to
help.
A
So
so
yeah
so
yeah,
so
the
providers
that
are
being
currently
worked
on
are
targeting.
You
know:
different
public
clouds,
right
yeah
and
the
the
company
where
I
work.
We
we
have
an
on-premise
use
case.
You
know,
customers
that
that
one
be
able
to
deploy
kubernetes
where
there
is
no
infrastructure
API.
A
So
internally,
we've
been
working
on
a
on
a
on
a
provider
that
that
that
targets
that
kind
of
environment,
but
you
know
we,
we
had
some
pretty
tight
deadlines
and
we
had
to
deviate
from
where
a
clustered
API
was
at
the
time.
Does
this
work
for
us
started
back
in?
You
know
last
May
yeah,
but
now
we
want
to
come
back
and
and
and
yeah
collaborate
with
them
with
Samsung,
who
has
a
similar
use
case.
A
We
want
to
work
on
a
this
generic
provider
and
and
what
I
think
one
of
the
one
of
the
one
of
the
use
cases
for
this
generic
provider
is
that
you
know
if
somebody
wants
to
start
testing,
you
know
the
cluster
API,
building
the
cluster
API,
but
doesn't
have
access
to
like
virtualized
infrastructure,
nor
public
clouds.
There
are,
then,
this
generic
provider
would
be
good
to
have,
because
then
you
know
you
you,
the
idea.
B
A
You
know
it
actually
uses
SSH
to
to
invoke
commands
on
the
host
and
provision
kubernetes
on
the
machines
right
so
as
opposed
to
the
providers
that
target
the
public
clouds
where
they
actually
create
an
instance,
and
they
pass
in
something
like
a
cloud
init
script
and
there's
no
for
the
most
part,
there's
no
there's
no
sort
of
there's
no
invocation
of
of
commands
directly,
but
of
course
it
requires
right.
It
requires
that
you
have
this
infrastructure
API,
so
yeah
I
think
you
know.
If
you
are
interested
in.
A
B
A
So
we're
trying
to
so
because
David
and
I,
you
know
both
of
our
teams
have
have
kind
of
had,
you
know,
have
to
have
an
implementation,
but
that
you
know
it's
not
it's
not
really.
It's
not
really
actually
I
think
we
just
we
just
made
our
repos
public
just
now
and
but
but
because
you
know,
we've
deviated,
it's
not
really
a
provider
that
somebody
pick
up.
You
know
in
the
way
that
you
could
pick
up.
You
know
the
AWS
is
the
GCP
provider,
but
anyway,
so
we
have
this
internal
implementation.
A
So
we
what
we
sort
of
what
we
have
you
know,
are
the
things
that
we've
learned
from
it,
and
so
we
were
Dave
and
I.
We
want
to
work
on
a
you
know
on
a
proposal.
We're
gonna,
we're
gonna,
publish
that
this
time
next
week,
but
I'm
you
know
before
we
do
that.
So
actually
I
should
say
when,
when
we
do
that
right,
then
then
we'll
actually
you
know
once
we
nail
down
sore
the
requirement,
you
know,
what
is
it,
what
are
gonna
be
the
goals
right?
A
A
So
right
now,
though,
I
guess
the
most
is
the
most
important
thing
is
finding
those
requirements
and
those
blockers.
So
if
you
have
you
know,
if
you
have
any
requirements
of
blockers
that
that
would
be
very
useful
to
know.
Otherwise
you
know,
when
we
publish
a
draft,
then
then
you
can
always
add
your
as
your
comments
there
yeah
and
then
once
I
guess
yeah.
Once
we
nail
down
the
the
proposal,
then
we'll
start
implementing
yeah,
and
at
that
point
you
know
as
as
much
help
as
we
can
get
really
would
would
be
would
be
great.
A
You
know
the
sort
of
they
will
be
implementing
the
actuators.
The
machine
actuator,
the
cluster
actuator
that'll
be
in
go
so
it'll
be
similar
to
their
to
the
other
cloud
provider
implementations,
but
we
will
have
some
additional
pieces,
you
know
so
so
the
other
the
AWS
cloud
provider
I,
think,
has
some
libraries
to
to
invoke
AWS
api's
and
then
the
unit
tests
for
those
so
we'll
have
we'll
have
something
similar.
You
know,
depending
on
how
we
end
up
communicating
with
machines
right
at
SSH.
A
You
know
we'll
have
some
library
that,
where
we're
invoking
commands
with
SSH,
maybe
we're
copying
overwrite.
If
we
want
to
write
configuration
files
to
the
to
the
to
disk,
we
might
be
using
SSH
for
that
or
you
know
perhaps
SCP
so
yeah
anything
if
you
have
any
kind
of
interests
or
you
know
there,
then
that
that
would
be
all
those
contributions
would
be
very
welcome.
B
A
So
let
me
let
me
head
to
the
tech
services.
You
know
we've
been
originally.
We
wanted
to
to
make
this
all
open
source
from
the
beginning,
but
it
didn't
happen
and
then
we
realized
we
had.
You
know
we
had
gone
source
or
so
far
from
the
upstream
implementation
that
it
really
wasn't.
You
know
we
weren't
it
wasn't
something
that
we
could
just
contribute.
A
So
so
then
we
we
didn't,
you
know,
make
the
repos
public,
but
we
thought
okay,
at
least
if
we
make
the
repos
public,
as
at
least
we
can
talk
about,
you
know
what
we've
done
for
better
or
worse
right,
though
there's
things
that
we've
done,
maybe
that
we
don't
like
but
anyway.
So
let
me
let
me
add,
that
to
the
dock
ssh
provider.
A
A
A
Yeah
and
so
then
the
yeah,
the
higher
the
high
level
design
of
this
SSH
providers.
You
know
we
try
to
keep
the
logic
and
the
sort
of
the
orchestration
the
sequencing
in
the
provider,
but
for
the
host
side
pieces.
You
know,
for
instance,
installing
the
container
runtime,
if
that's
necessary
or
pulling
down
the
dependencies
things
that
so
we
used
cube
ATM
but
cube
ATM
requires
that
you
pull
down
certain
dependencies
or
you
know,
configure
certain
things
that
is
out
of
scope
for
cube
ATM.
So
we
we
actually
wrote
a
separate
tool.
A
A
A
A
Anyway,
so
yeah
the
plan
is
to
keep
track
the
proposal,
get
feedback
and
then
start
start
implementing
and
yeah
I
mean
at
this
point.
We
you
know.
Obviously
we
don't
know
whether
whether
the
community
will
be
interested
in
in
something
like
node
ADM.
The
community
is
interested
in
STD
ATM.
We
open
source
that
a
couple
of
months
ago,
and-
and
you
know
it's
it's
seeing
some-
it
seems
some
adoption
so
yeah.
A
B
A
That's
a
good
question:
I
think
that
you
know
if
there,
if
there
is
an
issue
and
it
has
a
label,
you
know
Help
Wanted,
that's
a
pretty
clear
signal
that
you
know
nobody's
working
on
it
yeah.
This
is
the
office
hours
tomorrow
there
will
be
a
meeting
at.
Let
me
double-check,
ten
ten
o'clock,
the
same
time
zone,
so
two
hours
has
two
hours
earlier
than
this
one
started
and
there
we
typically
you
know
talk
about
the
the
the
cluster
API.
B
A
Know
maybe
take
our
take
a
small
one
and
and
look
at
it
you
know
see
if
there's
I
mean
yeah
I,
guess
it's
it's
you.
You
may
not
have
the
the
context
on
on
everything,
but
like
there's
some
somewhere,
you
might
be
able
to
sort
of
pick
up
the
context
just
from
the
PR
yeah
and
then
yeah
I,
see
if
you,
if
you
see
some.
If
you
see
some
issue
and
you
you
like
to
work
on
that,
then
yeah
just
just
add
a
add,
a
comment
and.
A
B
A
Great
yeah
and
if
you
haven't
seen
them
there
are
there's
a
recording
of
a
deep
dive
for
the
cluster
API
that
was
recorded
at
last
week's
coop.
Con
it'll
be
out
on
YouTube
Oh
awesome
in
that
case,
I'll.
Look
for
it,
yeah,
okay,
we're
good
great
yeah!
We'll
see
you
next
time
in
C,
all
right
think
yeah
I'll
wrap
up,
since
this
is
just
the
two
of
us
all
right.
Take
care.