►
From YouTube: 20191217 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
Hello
today
is
December
17th
2019.
This
is
the
standard
sing
cluster
lexical
call,
it
seems
like
we
are
low
in
attendance
because
I'm
guessing
a
lot
of
people
have
checked
up
for
the
holidays
already,
but
we
will
march
on
and
record
and
post
to
the
introverts.
In
your
absence
we
have
a
pretty
sparse
agenda,
we'll
see
what
we
can
get
through.
A
Lukas
has
went
back
to
school,
but
he
did
a
lot
of
good
great
work
for
the
community
so
on
the
pull
request,
I
talked
with
Justin
privately,
but
I
wanted
to
let
everyone
else
have
their
say
too,
as
well.
If
they
have
opinions
or
thoughts
about
the
idea
of
promoting
Lumiere
to
help
to
be
a
sigelei,
the
mayor
has
done
a
ton
of
work
across
the
whole
project.
Is
testing
from
testing
to
component
config,
to
automation,
to
curb
ATM
to
you
name,
it
little
mirrors
pretty
much
been
in
the
thick
of
it.
A
I
think
he
has
definitely
earned
his
stripes
there,
and
with
this
fancy
new
community-based
title,
he
gets
even
more
work.
So
next
up
was
for
Bo
Furi.
She
has
been
doing
a
lot
of
work,
helping
to
design
a
lot
of
the
new
features
in
sequester,
live
simple
and
as
well
as
integrate
them
across
some
projects.
A
So
I
thought,
given
all
his
work
across
the
many
areas,
as
well
as
his
mindfulness
of
how
he
approaches
coming
to
solutions
that
we
wanted
to
help
recognize
that
both
of
these
folks
have
been
active,
thoroughly
active
for
the
last
one
to
two
years
me.
So
if
you
have
any
questions,
come
as
complaints
concerns,
if
you
don't
feel
comfortable
talking
openly,
please
feel
free
to
ping
me
privately,
and
the
issue
is
currently
on
hold,
so
I
figured
I'd,
give
it
a
week
time
out
and
if
you
have
any
questions
comes
companies
concerns.
B
Yes,
thank
you.
So,
as
we
are
sparse,
this
might
be
a
little
trickier,
but
I
have
volunteered
myself
to
sort
of
organize
our
representation
at
coop
con
EU
next
year,
which
is
actually
I,
believe
April
early
in
April,
April
2nd.
So
it's
pretty
early,
it's
in
Amsterdam.
We
established
a
little
couple
minutes
ago
and
we
traditionally
do
a
Sig,
intro
and
deep
dive
and
then
also
in
addition,
we
do
some
please
correct
me
if
I
see
anything
correct
incorrectly.
We.
In
addition,
we
do
sub
project
updates
for
at
least
the
bigger
sub
projects.
B
B
It
is
obviously
a
some
work
to
prepare,
but
then
you
are
a
speaker.
Some
thing
I'm
planning
on
going.
I
but
I
also
think
we
should
sort
of
spread
the
spread,
the
representation
around
as
appropriate.
So
I,
don't
know
the
right
way.
I
guess
the
right
way
to
do
this:
I
open
a
call-out
and
slack
I
guess
as
well,
given
we
are
relatively
light
here,
but
I
wanted
to
bring
that
up,
and
hopefully
people
can
comment
as
well.
C
I
think
that
for
the
coop
I
mean,
if
someone
has
I
did
the
last
tool
and
I
don't
have
proper
tool
to
do
also
this
one,
but
considering
that
in
in
Europe
we
have
almost
on
all
the
other
contributors.
I
I
opened
that
someone
we
take
charge
of
the
Google
media
drive,
but
if
necessary,
I
can
step
in
on
Oricon
our
elbows
in
some
others
a
project
if
there
are
lack
of
contributor
that
they
are
not
coming
to
Europe
so
I'm
here
to
help.
A
Yeah
I
think
what
we
might
want
to
do,
too
is
talk
with
the
sub-project
owners,
if
there's
a
lot
of
contributors
to
make
sure
that
we
get
fair
table
opportunity
for
folks
to
do
the
call.
My
only
ask
is
that
typically
a
saintly
or
sick
chair
do
the
entry
six
year,
something
that's
been
around
for
a
really
long
time.
A
B
A
B
Okay,
great
well
I
have
a
thank
you
for
everyone
that
already
stepped
up
I
will
probably
I,
guess
actually
create
another
doc.
The
solution
is
always
another
doc.
I
hope
when
we
create
another
doctor
to
collaborate
more
on
this
and
break
this
out
and
reach
out
and
slack
and
reach
out
to
the
product
leads
for
the
sub-project
I.
B
A
A
E
I
was
gonna,
say:
I
could
probably
give
a
brief
overview
of
the
current
state
of
the
conformance
testing
right
now
at
least
the
AWS
job
on
master,
and
we
have
conformance
job
against
the
latest
release
branch
we've
been
keeping
those
pretty
good.
Andy
has
been
staying
on
top
of
those
that,
as
we
received
breakage,
making
sure
that
those
get
resolved
pretty
quickly.
E
E
A
E
Yeah
I
think
the
biggest
thing
that
I
worry
about
right
now
with
it
is
that
they
are
not
necessarily
overly
reliable.
Yet
you
know
some
of
it's
just
normal
test
flakiness,
but
there's
also
aspects
we've
seen
where,
because
we're
building
images
and
deploying
cluster
API
and
then
running
the
tests,
sometimes
in
the
set
up
steps,
we
do
encounter
some
issues,
though
okay.
A
They
don't
need
to
be
purple
requests
they.
The
release,
blocking
jobs
are
not
all
purple
regress,
similar
periodic
s--,
but
there's
the
release
blocking
signal
basically
mean
that
if
something
were
to
fail
immediately,
the
release
team
would
be
poking.
You,
with
their
iPoker,
stick
to
try
to
figure
out
what
went
wrong.
Okay,.
A
So
my
long-term
goal
would
be
that
we
kill
slash
cluster
director
with
fire
because
it's
like
woefully
under
maintained
and
the
state
doesn't
even
really
maintain
it
or
ruin
it,
and
but
we
still
kind
of
like
limp
along
in
perpetuity,
so
I
think
the
if,
as
soon
as
we
can
get
or
feel
comfortable
to
flip
the
switch
to
make
cluster
API
release
blocking.
That's
basically
testing
all
the
thing.
A
lot
of
componentry
that
the
cig
action
was
maintains,
as
well
as
the
kubernetes
bits.
I.
B
I
think
I
said
last
time:
I
put
up
in
absentia,
we
ve
got
our
moreless
caught
up.
In
other
words,
we
have
a
master's
in
out
targeting
118,
which
is
nice,
so
we're
sort
of
try
and
land
some
bigger
features.
Looking
at
things
like
add-ons
and
container
the
and
cluster
API,
all
that
good
stuff
I,
don't
know
they
make
it
all
in
three
months,
but
yeah.
It's
sort
of
in
a
reasonable
place.
B
C
Yeah
I
didn't
finish
to
write,
but
so
the
main
point
is
that
the
last
week
we
did
planning
and
t0
for
this
cycle
is
to
move
but
mean
out
of
train.
We
are
the
first
project
doing
this,
so
there
are
a
lot
of
work
to
do
some
question
mark
here
and
there
lupa
will
create
a
great
document
trying
to
define
now
to
do
this.
This
word,
and
this
will
be
the
most
important
job
for
decipher.
C
A
G
F
A
H
A
B
H
A
F
A
A
E
A
A
A
F
Yeah
I'm
gonna
start
work
on
the
API
types,
so
we
have
a
configuration
file
rather
than
flags,
which
is
the
interface.
Now
there's
been
a
recent
uptick
in
PRS
for
small
features,
some
which
have
been
flags
and
then
bug
fixes
so
I've
been
keeping
on
top
of
the
reviews
there
and
last
but
not
least,
I
want
a
shout
out
to
Lumiere
for
her
decoupling
stdm
from
kubernetes
kubernetes.
We
had
a
an
issue
with
go,
get
with
go
modules
and
he
got
to
the
bottom
of
that
and
yeah.
D
We're
working
towards
B
1
for
3.
We
have
been
doing
like
a
lot
of
work
on
the
control
plane
proposal.
There
is
still
one
proposal
out
which
I
think
it
has
a
time
a
lot
like
for
tomorrow
and
that's
the
North
remediation
machine
flow
is
also
coming
along
and
we
also
like
to
been
working
on
poster
coffee.
D
There
is
a
lot
of
care
about
her
go
to
and
address
in
advance
if
I
didn't
get
them
like
before
the
last
few
days,
but
I
will
today
and
we're
also
updating
all
our
dependencies
and
the
mean
required
person
for
the
management
plus.
There
will
be
good
news
for
116
and
we
are
planning
to
move
to
C
at
the
big
one
as
well
within
this
week.
Oh.
D
A
E
You
so
I,
don't
think.
We've
made
much
progress
in
the
past
couple
of
weeks
for
the
image
builder
we've
had
a
couple
of
fixes
come
in
for
the
existing
packer
and
ansible
scripts,
but
other
than
that
I
know
that
there's
been
work.
Actually,
there's
work
been
done
to
actually
get
the
documentation
published
similar
to
how
we
have
for
cluster
API
went
net,
laugh
hi,
so
that
has
happened
and
I
know.
Moshe
was
looking
at.
How
do
we
enable
automated
builds,
but
I
don't
think
there's
been
any
progress
made
on
that
yet.
A
One
of
the
things
I
want
to
think
about
doing
too
is
if
some
projects
have
like
a
call
to
arms
thing,
maybe
in
our
status
updates.
If
there's
certain
work
items
that
folks
would
like
to
get
to
you,
but
just
don't
have
enough
resources
if
they
want
to
highlight
them
here.
I
can
probably
try
to
take
this
back
towards
the
broader
community
and
say
from
folks
who
are
interested
here.
A
A
Wish
Ted
ass
automation
was
easy,
but
sadly
it's
like
a
weird
Rube
Goldberg
machine.
G
Here
so
we
really
for
mini
cube.
We
released
one
six
last
week
to
coincide
with
the
kubernetes
to
be
117
release.
We
managed
to
get
about
a
40%
speed,
improvement
for
new
vm
and
startup.
We
also
have
a
handful
of
updated
components
and
complaints
addressed,
so
we
did
have
some
regressions,
though
file
locking
on
Windows.
We
also
broke
off
blind
support.
These
were
signs
of
systemic
issues
when
I
can
comes
to
our
integration
testing
it.
G
If
you
have
integration
tests,
don't
ignore
the
results
if
they're
flaky
so
after
the
holidays
will
be
merging
a
bunch
of
the
things
that
we
demoed
at
Q
on
the
multi,
node
VM,
free
and
pause,
as
well
as
cleaning
up
our
integration
tests
and
hopefully
forcing
it
so
that
people
can't
merge
when
on.
If
the
integration
tests
are
failed,
at
least
for
our
reliable
ones.
Oh,
we
got
any
questions.
H
Great
yeah,
so
in
add-ons
as
far
as
work,
that's
hoping
to
be
merged
at
some
point
got
some
updates
into
the
cat
made
some
documentation,
proposals
that
are
still
working
progress
for
the
Covidien
integration.
The
kuba
TMP
are
now
has
an
upgrade
implementation
whittling
down
the
to
do's
on
that
it
still
needs
to
rebase
for
a
component
config
refactor
that
rusty
merged
into
cube
idiom.
H
Cool
moving
on
to
components,
standard
I
just
wanted
to
do
some
call-outs
for
a
bunch
of
work.
That's
been
happening.
We've
been
very
active
since
coop
con
had
a
very
successful
coop.
Con
Mike
and
I
are
just
really
glad
to
be
working
with
new
contributors
and
a
few
seasoned
people
who
have
shifted
what
they've
been
working
on
from
groups
like
swig
testing,
to
focus
on
components,
standard
component,
config
isn't
the
only
thing
that
we
do,
but
it
is
the
primary
effort.
H
Currently
OB
tech
and
la
latinum
are
rebooting
coop
controller
manager
component
config
there's
been
a
lot
of
design
discussion
and
just
they
are
championing
pushing
that
through.
So
we
really
want
to
focus
on
getting
an
API
that
is
somewhat
usable,
but
really
just
focusing
producing
something
that
people
can
use
and
then,
if
we
need
to
change
it,
that's
why
we
have
component
config
can
do
stuff
in
a
future
version,
but
we
really
want
to
get
to
an
alpha
of
that
and
that
people
can
use.