►
From YouTube: 20190618 sig cluster lifecycle
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
Hopefully,
folks
can
see
my
screen
Yes
No,
maybe
so
alright.
So
if
you
could
add
your
details
to
the
attending
list,
that
would
be
super.
Helpful.
First
up
is
a
PSA
just
a
PSA.
For
a
long
time,
the
sig
we
have
bantered
about
the
creation
of
a
separate
sub
project
specifically
for
building
images,
it's
kind
of
an
overlap
between
sig
release
and
say
cluster
lifecycle.
We
would
probably
be
doing
most
of
the
maintance
work,
so
it
kind
of
would
live
here.
A
It
wouldn't
necessarily
have
to
live
with
the
release
team
either
because
we're
not
actually
publishing
artifacts
as
part
of
the
release
output,
it
would
allow
them
to
publish
artifacts
at
any
snapshot
point
in
time,
so
it's
primarily
for
end
consumers
and
those
who
are
working
on
cluster
API.
So
if
you
have
a
chance,
please
comment
on
it.
I
think
we'll
do
the
standard
process
of
letting
it
curate
in
the
community
so
give
folks
enough
time
to
read
through
the
cap
and
maybe
in
next
week.
We'll
have
one
more
discussion.
A
B
A
C
A
Next
up,
I
wanted
to
discuss
this
I
logged
an
issue.
I
talked
with
I
talked
with
Jordan
and
DIMMs
on
this
topic
is
that
it
overlaps
with
another
area
that
folks
would
like
to
do
so.
I'll
give
I'll
give
a
brief
synopsis
to
the
background.
Here.
We,
the
slash
cluster
directory
or
kuba,
has
been
quote.
Deprecated
quote
for
almost.
E
A
Reason
why
slash
clusters
on
in
perpetuity,
whoever
is
not
talking
or
any
major
it
lives
on
a
perpetuity,
and
it
constantly
gets
updated
and
we're
always
pained
almost
every
single
release
cycle
to
at
least
find
or
identify
issues
that
are
discrepancies
across
deployments
or
defined
or
fix
bugs
that
exist
inside
of
it.
So
what
I
would
like
to
do
is
to
at
least
start
to
enumerate
the
list
of
items
as
a
sig
of
what
would
be
required
in
order
for
us
to
remove
the
clustered
directory
in
in
the
ideal
world.
A
What
I
would
hope
for
would
be
to
use
the
tools
that
the
sig
has
created.
You
know,
given
that
a
lot
of
the
testing
automation
is
still
on
Google
Cloud
I
would
really
like
to
do
something
like
build
up
the
GCP
provider
for
cluster
API
and
use
a
lot
of
the
standard
patterns
that
we
sort
of
curated
in
other
on
straight
bi
providers.
Try
to
get
that
the
speed.
It's
like
a
longer-term
vision
but
I
think
in
the
short
term.
D
A
D
A
C
F
A
D
H
B
A
F
Yes,
I
think
that
if
we
so
what
what
is
our
now
that
we
have
kind
and
kinder
to
didn't,
we
at
some
point
say
that
we
would
like
to
move
so.
The
pre
submits
are
tests
that
are
really
fast,
stable
and
require
minimal
in
front
Google.
Kindly
hosts
like
90%,
at
least
of
these,
if
probably
more
thanks
to
the
to
the
grant.
F
But
is
there
any
way
we
could,
as
a
first
step,
use
some
of
the
other
cluster
API
providers
that
we
have
that
are
like
working
that
are
that
I've
received
the
the
the
maintenance
they
need?
Is
there
anything
blocking
us
from
from
setting
up
sayin
iew?
Yes,
they
they
donated
some
some
money
too
right.
So.
A
I
think
the
biggest
blocker
there
is
just
hooking
up
the
tooling
and
time
I
I
do
think
that
we
could
probably
hook
that
up
and
should
probably
hook
that
up
and
that
could
for
a
good
number
of
the
test
scenarios.
They
could
be
a
cohort
of
GCP
and
AWS
in
the
fullness
of
time
to
get
better
cross
coverage
across
cloud.
D
An
update
on
that
there
was
my
I
was
trying
to
so
we
can
do
that.
We
just
need
unique
email
addresses
for
each
and
I
was
struggling
to
create
I.
Guess
a
group
I
got
lost
in
the
hierarchy
or
the
the
org
of
who
creates
Google
Groups,
but
I
will
I
will
kick
it
again
and
see
if
I
can
figure
it
out,
but
we
basically
need
either
one
Google
group
or
any
google
groups
to
any
whether
they
support
suffixes
or
not,
or
whether
amaz.
A
B
A
A
Yeah
link
your
shoes
are
start
to.
You
should
be.
If
you
have
maintainer
rights,
you
should
be
able
to
modify
or
edit
this.
What
I'm
hoping
to
do
is
get
like
an
umbrella
list
of
the
work
items
we
need-
and
you
know,
start
to
actually
every
single
release
cycle
start
to
revisit
and
burn
away.
Some
of
these
items.
F
F
Yes,
give
it,
given
that
we
have
cluster
API
AWS
in
good
shape
and
and
kind
I
mean
I'm,
not
saying
we
should
test
the
cloud
provider
specific
things
that
is
nice
way
out
of
scope.
For
now,
like
can
I
create
a
aw,
slow
balancer,
which
service
type
load
blender.
No,
not
that
kind
of
stuff,
just
like
spin
up
get
compute
whatever,
wherever
it
is,
and
then
just
like
one
kind,
pretty.
H
A
We
start
this
you're
cutting
in
and
out
Lucas.
If
you
are
interested,
please
chime
in
on
that
chicken,
and
we
will
try
to
federate
out
the
work
appropriately
to
try
and
get
more
people
more
eyes
on
this
little
brew
visit
this
issue
across
these
meetings.
We
can
probably
move
on
because
we
could
wrap
it
hold
on
just
this
topic
if
we
wanted
to
and
and
not
get
to
the
other
agenda
items.
A
The
next
step
on
my
list
is
also
a
Soviet
associated
with
sort
of
like
project
maintenance
and
curation.
So
with
the
maturity
of
kind
and
kinder
I
was
talking
with,
then
the
other
and
I
don't
know
if
there's
a
that's
not
covered
in
those
two
tools.
That
is
that
dined
accommodates
today
and
the
question
I
had
is
whether
or
not
we
wanted
to
move
dined
to
retired
or
if
anyone
has
any
known
good
user
stories
that
aren't
covered
by
those
other
tools.
J
J
A
J
A
I'm
gonna
do
that
on
this
one
also
I
will
also
put
it
as
a
long-standing
issue,
but
it's
good
to
these
have
the
initial
PSA
here
and
then
as
we
as
we
can
revisit
it
again.
Next
time
and
I
have
more
detailed
information
and
also
try
to
sync
with
the
Moran
toast
folks
to
see
if
there
are
any
user
stories
that
aren't
being
met.
F
F
F
B
A
That
was
going
to
go
through
the
process
of
writing
up
an
email
and
asking
them
if
they,
if
they
want
to
continue
to
maintain
this
thing
or
if
we
can
move
it
into
a
deprecated
State
and
try
to
make
sure
make
sure
that
we
are
addressing
the
feature
parity
difference
across
the
different
implementations.
Okay,
yeah.
D
A
Yes,
right
yep,
so
there's
there's
a
well-defined
process
in
steering
for
retiring
projects,
so
right
cool,
and
for
those
of
you
who
don't
know,
because
this
is
fun
for
me,
we
actually
did
retire.
Kubernetes
contribu
have
been
around
the
project
forever
contribute
the
early
days,
dumping
bucket
for
everything
that
was
and
that.
F
A
Yeah
well,
every
now
and
then
wow.
You
know
it
was
one
of
those
things
that
it
was
a
compromise,
no
one
liked
the
compromise.
So
when
it,
finally,
you
know
when
his
finally
retired
retired
everyone
kind
of
you
know
there
was
no.
There
was
no
champagne
bottle
to
pop
at
that
time.
I
wish
there
was,
which
was
broader
notification,
but
we
can
move
on
I
digress,
Covidien
leave,
115
blog
posts,
Lukas
ratio.
F
Right
so
I
I
got
tinged
by
by
George
recently
Castro
recently
and
he's
managing
the
new
blog
post
for
for
115,
and
he
said
they'd
liked
to
have
a
Cuban
and
post
so
well.
We
I
posted
in
in
cig,
12
life
cycle
leads
in
flack,
told
that
we
we
have
this
this
wish
and
now
on
yesterday,
and
today
we
quickly
wrote
up
something
that
that
should
should
service
the
link.
Is
there
I
created
a
quick
screencast
of
of
how
the
AJ
mode
works?
It
works
really.
Well
thanks!
Everybody
for
that
awesome
work.
F
It's
it's
super
smooth
just
to
guy
it's
eerily
too-tall
out
for,
like
the
reaction
could
actually
see
what's
happening,
joining
the
three
masters
and
it
took
less
than
three
minutes
I
think
so
that
is.
That
is
definitely
a
nice.
Nice
thing.
I'm
gonna
check
this
in
to
get
probably
after
this
meeting
yeah
tonight.
So
if
you
have
any
any
changes,
you'd
like
to
see
comment
comment
there,
please
and
we'll
try
to
address
them.
Otherwise
you
can
I'll
send
out
the
pr2.
F
Most
likely
it's
probably
gonna
land
on
Monday
next
week,
and
we
also
have,
in
the
main
convenience
1:15
release
blog
that
goes
out.
Hopefully,
tomorrow
we
also
have
a
section
called
cost
lifecycle,
stability
and
usability
improvements
that
were
just
in
a
few
sentences
talked
about,
but
we
we've
done
in
the
last
cycle
and
why
it
matters
right.
So
any
any
questions
to
that
comments.
Concerns.
F
Going
cool
right
and
then
we
have
the
logo
voting
that
little
mirror
credit
we
now
it's
been
up
for
a
week,
so
I
think
we
could
close
it
and,
as
part
of
this
announcement
now
that
we
gonna
I'm
gonna,
see
if
Alex
responded
there,
but
but
now
that
we
have
a
winner
from
from
the
poll
option,
two
which
is
see
see
reference
there.
We
can
actually
go
ahead
and
then
announce
this
during
during
the
blog
post
or
the
release
announcement
or
both.
So
so
that's
that's
nice.
F
Any
any
concerns
about
actually
accepting
this
particular
logo.
I
rode
up
to
the
three
most
popular,
the
stats
for
the
three
most
popular
logos
and
the
options
that
there
were
the
least
got
one
got,
got
67
votes,
17.4
percent
and
ii
got
12.4
percent,
so
it
was.
You
know
it
was
significantly
more
that
this
one
had.
F
F
F
Post
and
performance
results
we
actually
have
a
logo
is
this
all
came
from
that
we
submitted
the
conformance
test
for
114
and
Dan
said.
Well,
you
can't
you
shouldn't
use
company's
convenience,
logo
and
we're
like.
Well,
we
don't
have
any.
We
don't
have
our
own.
So
then
he
was
like
well.
Cn
CF
can
can
help
you,
we
have
designers,
so
there
we
go.
Thank
you
our
own.
F
K
Everyone
so
I'm
coming
here
today
as
a
co-chair
of
sing.
The
Emer,
as
some
folks
might
know,
we're
in
the
process
of
folding.
Some
of
the
club
provide
really
good
steaks,
including
the
Elmer,
and
we
always
hosted
the
cluster
API
provider
for
vSphere
undersea
VMware.
So
you
know
with
the
upcoming
folding,
we
wanted
to
propose
that
we
can
move.
K
A
F
A
D
K
G
K
Should
probably
address
that
and
just
to
give
another
point
of
information
here
in
sick
cloud,
cloud
provider
recently
proposed
a
change
and
the
Charter.
So
everything
that
is
not
everything
that
his
club
provider
dependent
and
isn't
it
owned
by
any
artist
states
will
end
up
being
owned
by
cloud
provider.
So,
like
things
like
CSI
and
are
things
that
are
not
currently
owned
by
any
other,
like
largest
sig
like
lost
lifecycle
would
be
on
the
cloud
provider.
But
since
cluster
API
is
owned
by
cluster
lifecycle,
I
think
it
makes
sense
for
just
consolidate.
A
Plus
it'd
be
nice
at
King
Khan.
If,
when
we
do
the
cluster
API
talks-
and
this
is
going
to
be
later
on,
I
think
we
have
a
topic
on
this
when
we
do
the
coop
context,
it'd
be
nice.
If
the
provider
of
limitations
could
a
brief
update,
along
with
the
the
main
capitis
to
be
able
to
actually
say
this,
is
the
state
of
things
to
give
an
overall
update,
as
part
of
it
could
come
discussion.
F
K
F
Right
right
so,
but
yeah
right,
the
folks
will
be
the
same
that
if
you're,
if
I'm
interested
in
vSphere
and
want
to
do
stuff
on
that
I
work
in
the
cloud
provider.
I
work
in
the
this
VMware
sub-project.
Obviously
cloud
provider
and
I
maybe
work
these
cluster
lifecycles
vSphere
provided
for
cluster
idea
and
things
like
that.
So
it's
it
spread
around
different
owners,
but
those
people
that
are
in
cloud
provide
specific
SIG's
like
GCP
and
I
WS
and
vSphere.
We
ever
and
similar
will
then
be
just
spread
around
different
places.
F
C
F
F
A
D
I
just
suggest
that
we
we
consider
old,
old
agape
providers,
although
other
cluster
of
the
providers
together,
rather
than
that
was
only
my
only
concern
like
whether
we're
doing
something
different
for
this
year
than
we're
doing
for
your
OpenStack
right.
Just
just
that.
We're
actually
homogenizing
everything
I'd.
A
F
K
E
E
A
You
so
typically
sub-project
owners
made
sub-project
owners
request
the
leads
to
have
a
submit
a
talk
under
cyclist
or
lifecycle
umbrella,
most
of
the
time
they're
just
status
updates
of
the
projects,
as
well
as
like
the
call
to
it's
a
it's
a
it's,
a
trifecta
of
status,
update
whatever
you
plant.
What's
the
what's
the
vision
status,
update,
what's
the
vision,
slash
plan
and
how
you
can
get
involved,
typically
so
I
discern
that
most
of
the
major
sub
projects
would
want
to
submit
a
talk.
B
F
To
mention
that
some
we
have
to
motivate
to
cstf
stuff
or
cube
conserve.
Why
why
we
should
have
like
five
different
different,
deep
dives,
but
that's
that's
something
we
definitely
will
and
can
do
as
long
as
you
have
the
we
think
we
have
the
right
content
represented
and,
as
Tim
said,
I
think
makes
sense
to
formalize
the
other
templates
or
just
like
requirements
for
four
slides.
A
A
A
A
I
like
that
idea,
so
if
you
have
other
thoughts,
please
add
them
to
the
list,
but
we
can
probably
because
really
have
twenty
four
or
three
minutes
left
kind
of
want
to
make
sure
we
go
through
our
our
round-robin
of
updates.
Here
it
would
be
helpful
to
if
someone
else
can
help
me
will
take
notes
with
our
relevant
updates.
Are
there
any
other
Lancer.
F
F
E
Didn't
go
ahead,
closer
up
yeah,
we
have
a
full
walkthrough
scheduled
for
next
week,
June
at
will.
This
is
gonna,
be
in
kind
of
superseding
the
office
hours,
and
else
we
will
go
through.
Like
kind
of
like
the
closer
you
get
good
Bayes
rule
I
could
do
some
some
quick
demos.
This
should
take
like
around.
Like
one
hour.
E
The
next
item
is
machines.
They
strapped
in
proposal.
This
is
still
like
a
work
in
progress.
We
haven't
talked
about
like
like.
If
we
want
to
plan
to
merge
this
like
this
week
or
next
week,
it's
like
kind
of
ongoing.
We
will
start
opening
issues
to
kind
of
like
start
splitting
up
the
work
and
the
first
one
that
we
actually
tackled
and
kind
of
readapted
for
be
one
out
for
one
is
the
note.
Rest
controller.
E
If
you
don't
know
like
cluster
api,
has
a
concept
of
machine
which
are
like
then
become
kubernetes
nodes
and
we
try
to
link
a
machine
to
kubernetes
node.
Sometimes
the
cluster
controller
runs
in
management
cluster,
which
is
a
cluster
that
runs
outside
of
the
cluster
that
you're
trying
to
spin
up,
and
in
that
case,
close
Rafi
I
didn't
have
any
support
for
accessing
the
targeted
cluster.
E
E
A
Cool
thanks,
yes,
so
for
the
broader
update
that
was
a
there,
was
an
update
to
the
main
repository
that
it
makes
sense
to
offset
cluster
API
by
six
weeks.
So
that
way,
it's
easier
for
them
to
consume
the
kubernetes
release
and
to
make
sure
like
the
comedian
types
are
solid.
It's
gone
through
some
betting.
It's
got
at
least
like
a
dot
one
or
dot.
Two,
your
release
underneath
it's
about
and
then
two
vendor
in
the
types
information
and
go
from
there.
A
F
See
if
this
sounds
good,
oh
another,
another,
a
meta
question
that
sprang
to
mind:
how
do
do
we
have
any
weight
so
to
say
manage
the
difference?
Oh
now
that
we're
accepting
we
Spherion,
we
do
OpenStack
GCP
at
AWS.
All
of
these
do
we
have
any
process
of
tracking
dashboard
or
some
kind
of
place.
I
could
go
to
and
know
that.
F
A
This
is
a
long-standing
problem.
I
I
have
strong
opinions
on
this
strong
opinions
loosely
held,
which
is
basically
a
way
of
stating
that,
like,
let's
get
an
argument
about
nothing
so
the
we
needs
to
be
addressed,
it's
not
been
addressed.
We
keep
on
kind
of
punting
that
issue
around
in
a
circle.
I
got
none
of
the
answers,
anything
that
I
think
that
anybody
likes
so
I
think
I'm
gonna
defer
to
that
group
and
that
until
we
get
consensus
there
and
I'll
help
moderate.
E
I
want
to
add,
like
we're
kind
of
trying
to
move
forward
I'm
like
reduce
fragmentation,
that
you
know
between
providers
that
we
have
today.
As
you
mentioned
like,
we
need
something
like
how
like
I'd
like
a
dashboard
some
source
I
like
that,
we
can
see
which
providers
like
are
supported
and
etc.
So
we're
trying
to
move
forward
in
that
direction.
It's
long
road
right.
F
I
I
think
give
a
quick
update,
so
could
me
now
is:
quite
there
are
no
much
activity.
We
are
monitoring
the
tests
agreed
for
release.
Everything
looks
good.
We
had
a
strange
problem
with
IP
I
serve
answering
unauthorized
and
we
implemented
a
workaround,
but
the
problem
should
be
better
investigating
in
order
to
find
at
the
root
and
six
the
reality
of
the
problem,
and
that's
it
for
the
till
tea
that
are
going
on
now,
the
next
office,
our
meeting
tomorrow,
we
are
planning
to
where
we
want
to
have
a
planning
for
116.
A
J
Next
step,
add-ons
call
so
Daniel
and
I
have
done
some
project
management
and
if
you
take
look
at
the
issues
pages
they're
there
for
reference,
we've
got
some
concrete
items
of
work
now
that
are
stepping
towards
the
proposal
of
moving
forward
with
implementing
more
kind
of
customized
patching.
We've
got
the
project
board
as
well
yeah,
so
using
customized
not
only
for
its
patching,
an
overlay
support,
but
also
then
wrapping
it
in
some
packaging.
J
If
anyone
is
looking
to
contribute
into
interesting,
space
is
obviously
something
that's
been
a
problem
space.
That's
need
to
be
solved
in
user
space.
You
know
out
in
the
community
and
now,
if
we've
got
something,
that's
really
doing
some
more
sophisticated
work
in
our
upstream
project.
So
contributions
welcome,
please
come
and
help
us
work
on
stuff,
I.
Think.
A
My
only
comment
might
be
that
if
we
we
try
to
adhere
to
the
policy,
you've
had
for
a
long
time
that
we
are
going
to
try
to
deprecate
slash
cluster.
Please
don't
add
this
to
clusters.
We
can
avoid
to
because
what
happens
in
perpetuity
is
all
features
of
major
modifications,
see
how
somehow
get
added
to
that
thing.
Yet.
J
F
Right
so
we
just
had
the
meeting
before
this,
and
we
discussed
a
few
few
things.
One
of
them
is
that
I've
decided
to
step
down
as
a
lead,
basically
just
because
of
lack
of
time,
which
is
really
unfortunate,
but
I
will
and
also
I'm,
going
to
university
and
a
couple
of
weeks
now.
So
that
is
also
a
preparation
in
in.
F
F
We're
obviously
I
won't
have
the
same
amount
of
time
I
had
now,
but
I'm,
hoping
like
that
Lee
or
any
similar
person
would
would
stay
a
step
up
instead
and
spoil
people
up
to
speed.
As
part
of
this,
we
just
discussed
that
we're
hoping
or
we're
planning
to
do
a
new
contributor
onboarding
session,
like
we
did
for
this
group
and
by
that
be
able
to.
F
A
E
J
D
Yeah,
no
businesses
use
business
as
usual.
No
real
significant
updates,
so
I
will
happily
yield
the
time
to
people
is
enough.
I.
A
H
I'm
here
from
mini
cube,
okay,
so
I
mean
this
is
basically
it,
but
basically
one
point
one
making
Windows
less
terrible
experience
and
updating
the
latest
versions
of
like
cryo
and
docker
and
kubernetes,
and
we're
working
on
a
command
line
globalization.
So
anything
that
gets
picked
up
spit
out
to
the
command
line
is
gonna,
be
translated
into
the
language
of
your
system.
Locale
on
the
fly
and
he's
gonna
Thomas
is
actually
in
China,
he's
gonna
demo
that
a
cube
con
and
our
dock
site
is
sort
of
up
and
running.
H
L
Hi,
can
you
guys
hear
me
yeah,
oh
yeah,
this
is
mithya
from
mini
cube,
also
we're
also
working
on
the
VM
free
mini
cube.
We
sign
our
milestone
for
next
release.
I
just
want
like
an
on-set,
so
we
are
working
on
that.
So
you'll
have
a
mini
cube
in
doctor
VM
freezers,
maybe
as
a
faster
experience.
That's
it.
M
Yeah,
the
the
first
step
project
meeting
is
happening
next
Monday.
So,
if
you're
interested
please
join,
there's
an
invite
already
to
the
google
group.
We
would
like
additional
approvers
I.
Think
at
the
moment,
I'm
I'm
able
to
approve
I,
think
Justin,
you
I
think
no
cigars
are
busy
with
with
a
lot
of
other
things,
so
I
I
would
like
to
get
additional
approvers.
I
I
have
a
couple
of
people
in
mind:
I
just
wanted
to
I
I,
don't
I,
don't
know
what
they
you
know.
What
the
protocol
is
so
first
I
wanted
to.
M
A
It's
like
it's
based
on
kinetic
contributions
right
if
there
are
people
who
are
gonna,
maintain
the
code
and
who
have
been
active
in
the
in
the
project.
I
hope
helping
sort
of
build
them
up
to
speed
to
be
a
good
approver
is
the
typical
mode
of
operation
so
like
in
coop
ADM.
We
regularly
try
to
recycle
people
around
right,
select
you
kind
of
build
them
up
and
every
cycle
we
try
to
promote
people
who
have
been
very
active.
A
M
M
D
M
D
C
D
M
F
F
As
Tim
said
in
the
at
the
end
of
cycle
way,
we
try
to
graduate
promote
the
people
that
have
been
doing
good
work
and
we
would
encourage
you
to
do
the
same.
Some
project
owners
for
each
sub
project
would
do
the
same
for
for
that
specific
project,
and
if
you
have
questions
you
can
reach
out,
tough
leads
in
slack
for
exam.
We
have
it.
We
have
a
channel
for
such
meta
discussion,
link
there
or
just
to
anyone
directly.